University of Southern California Dissertations and Theses

The Construction And Empirical Test Of A Theory Based On Selected Variables In Small-Group Interaction

(USC Thesis Other)

The Construction And Empirical Test Of A Theory Based On Selected Variables In Small-Group Interaction

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
THE CONSTRUCTION AND EMPIRICAL TEST
OF A THEORY BASED ON SELECTED VARIABLES
IN SMALL-GROUP INTERACTION
by
Joe R ichard Udry
A D is s e r t a tio n P re se n te d to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In P a r t i a l F u lf illm e n t of the
Requirem ents f o r the Degree
DOCTOR OF PHILOSOPHY
(Sociology)
June 1960
UNIVERSITY O F S O U T H E R N C A U FO R N IA
GRADUATE SC H O O L
U NIVERSITY PARK
LOS A N G ELES 7 . CA LIFO R N IA
This dissertation, written by
J .R ic h a r d .Udry
under the direction of hi-3.....Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by the Graduate
School, in partial fulfillment of requirements
for the degree of
DOCTOR OF PH ILO SO PH Y
Dean
Date................J .u n e jl.9 .6 0
DISSERTATION COMMITTEE
TABLE OF CONTENTS
C hapter Page
I. INTRODUCTION............................................................................ 1
In tro d u c tio n to the problem
Purpose of the Research
O rg an izatio n of the D is s e r ta tio n
I I . THE LOGIC OF THEORY CONSTRUCTION........................... 7
In tro d u c tio n
A G eneral Conception of S c i e n t i f i c Theory
The Logic of Concept Formation
The Logic of P ro p o s itio n s
Summary
I I I . A THEORY OF INTERACTION IN SMALL GROUPS . . . . 37
In tro d u c tio n
An In c lu siv e Theory of I n te r a c tio n
A R e s tr ic te d Theory of I n te r a c tio n
IV. AN EMPIRICAL TEST OF THE RESTRICTED THEORY
OF INTERACTION................................................................... 7b
In tro d u c tio n
P r a c t i c a l Mechanics
V. RESULTS OF THE EMPIRICAL STUDY................................... 87
In tro d u c tio n
E v alu a tio n of the Hypotheses
E v alu a tio n of the Theory
VI. SUMMARY........................................................................................... 113
BIBLIOGRAPHY............................................................................................... 122
APPENDIXES..................................................................................................... 129
i i
LIST OF TABLES
Table Page
1. D isp o sitio n of Cases of the O rig in a l Sanple * 79
2. D isp o sitio n of Hypotheses Derived from the
T h e o r y ............................................................................ 89
3. Summary of the P o ssib le and Obtained Ranges*
Means* and Standard D eviations of the
V ariab les .................................................... . . . . . 92
4. P earsonian I n te r c o r r e la t io n C o e f f ic ie n ts fo r
the Five V ariab les In te rp re te d fo r Whole
Groups (D .f. = 4 9 ) ................................................ 93
5. P earsonian I n te r c o r r e la t io n C o e ff ic ie n ts fo r
the V ariab les I n te r p re te d fo r P a irs of
I n te r a c to r s ............................................................. 94
6. M ultiple C o rr e la tio n A nalysis fo r the Five
V ariab les I n te rp re te d fo r Whole Groups
(D .f. = 4 6 ) .............................................................. 104
7. Comparison of M ultiple C o rre la tio n C o e ffi
c i e n ts of the V ariab les In te r p re te d fo r
Whole Groups and fo r P a irs of I n te r a c to r s . 110
i i i
CHAPTER X
INTRODUCTION
In tro d u c tio n to the Problem
Many s o c io lo g is t s assume t h a t sociology has emerged
in to i t s p e rio d of m a tu rity , where th e o ry and e m p iric a l r e
search are welded in to a sin g le s c i e n t i f i c endeavor.^ Al
though i t s t i l l seems re s p e c ta b le to " th e o riz e " w ithout
doing connected re s e a rc h or even in d ic a tin g how connected
re s e a rc h could be done, most re s e a r c h e r s h e s i t a t e to p u b lish
e m p iric a l re s e a rc h w ithout a "theory" of some s o rt to which
the re s e a rc h can be r e l a t e d .
The problems of r e l a t i n g th e o ry and re s e a rc h have
been d isc u sse d fre q u e n tly in such a way t h a t the re a d e r i s
l e f t w ith th e im pression t h a t th e o ry and re s e a rc h are two
r a th e r d i f f e r e n t th in g s which have a g re a t d e a l t o c o n trib
ute to one another i f only one ta k e s the tro u b le to r e l a t e
them. T his has le d , in a t l e a s t a few c a s e s , to the fo rc in g
of a r e l a t i o n s h i p between some p re v io u sly e x i s tin g th e o ry
and rem otely connected e m p iric a l work, or even the in v e n tio n
■^See, fo r example, Robbin M. W illiam s, J r . , "Contin
u ity and Change in S o c io lo g ic a l S tudy," American S o c io lo g i
c a l Review. XXIII (December, 1958), 619-33.
1
of th e o r i e s to j u s t i f y e m p iric a l s tu d ie s .
T reatm ents of the lo g ic of s c i e n t i f i c method have
long been a v a ila b le dem onstrating the c o n tin u ity between
n
th e o ry c o n s tru c tio n and e m p iric a l re s e a r c h . A good d e a l of
c u r re n t sociology, however, seems to r e f l e c t a lack of
awareness of t h i s s i t u a t i o n . T h is d i s s e r t a t i o n d e s c rib e s an
attem pt to e s t a b l i s h lo g ic a l c o n tin u ity between th eo ry and
re se a rc h in the s o lu tio n to a su b sta n tiv e s c i e n t i f i c prob
lem.
The su b sta n tiv e problem of t h i s re s e a rc h i s the
s p e c if ic a tio n and e m p iric a l v e r i f i c a t i o n of r e l a t i o n s h i p s
among a se t of fiv e v a r ia b le s in sm all-group i n t e r a c t i o n .
Many hundreds of s tu d ie s of sm ail-group in te r a c ti o n have
been p ublished in the p a s t decade, and "sm all groups" have
become a recognized f i e l d of sociology. So larg e a number
of v a r ia b le s has been measured in such a v a r ie ty of ways
t h a t th e re i s a s e rio u s need f o r r e p l i c a t i o n and sy stem a tiz
a tio n of work alre ad y done. This study i s intended to f u l
f i l l t h i s need p a r t i a l l y .
The f iv e concepts s e le c te d f o r t h i s re s e a rc h are as
fo llo w s:
1. Consensus.
2. I n te r p e rs o n a l u n d ersta n d in g .
^Otto Neurath ( e d .) . I n te r n a ti o n a l Encyclopedia of
U n ifie d S c ie n c e . Vols. I and I I : Foundations of the Unity""of
Science tChica'oo: The U n iv e rs ity of Chicago P re s s , s e r i e s of
20 monographs in d iv id u a lly d a te d ).
3. Democracy.
4. Frequency of in te r a c ti o n .
5. Size of groups.
These fiv e concepts were se le c te d on the b a s is of (1) t h e i r
freq u en t appearance in the l i t e r a t u r e on small groups and
(2) the p o s s i b i l i t y of specifying an unambiguous em p iric a l
r e f e r e n t f o r each concept, which w i l l a t the same time seem
to most s o c io lo g is ts to measure what i s u su a lly meant by the
concept in s o c io lo g ic a l usage. In so far as p o s s ib le , meas
urement techniques used by previous r e s e a r c h e rs were used,
so as to maximize the r e p l i c a t i v e value of t h i s re s e a rc h .
Purpose of the Research
The su b stan tiv e purpose of the re se a rc h is to de
velop a system atic th eo ry of s o c ia l i n te r a c ti o n in primary
groups. This theory i s based upon a small number of w ell-
studied v a r ia b le s in sm all-group in t e r a c t i o n , and c o n s is ts
of a p re c ise statem ent of the i n t e r r e l a t i o n s h i p s among th e se
few v a r ia b le s . The theory th u s re p re s e n ts an ordered sum
mary of p re se n t knowledge about th e se f a c e ts of in te r a c ti o n .
An e m p irical study was designed to t e s t the u t i l i t y of the
theory fo r p re d ic tin g in te r a c ti o n p a tte r n s in a sample of
adolescent same-sex frie n d s h ip groups.
The m ethodological purpose of the re se a rc h i s to de
velop lo g ic a l procedures f o r c o n s tru c tin g the theory of in
t e r a c t i o n , and to d escrib e the lo g ic a l problems unsolved or
4
inadequately solved by other so c io lo g ic a l t h e o r i s t s .
The value of a theory of
in te ra c tio n
Is small-group in te ra c tio n ready fo r development of
a theory? The following statem ent of Woodger, the th e o r e t i
ca l b io lo g is t, seems to set fo rth a sensible c r i t e r i o n fo r
the time at which theory becomes u se fu l in a f i e l d .
The question of the use of a technique of theory
con stru ctio n a r is e s as soon as a number of ( r e lia b le )
g en e ra liz a tio n s have been esta b lish e d and i t becomes
necessary to order them in to a system. There i s pre
sumably no question th a t ordering them in some way is
d e s ir a b le .3
I t i s assumed th a t such a time has a rriv ed in the study of
so cial in te ra c tio n in small groups.
What b e n e fits may accrue from a theory of in te ra c
tio n ? The follow ing possible b e n e fits are proposed:
1. The theory w ill b r ie f ly summarize d iverse em
p i r i c a l research es.
2. The co n stru c tio n of a f a i r l y p recise theory w ill
help uncover fa u lty logic which fre q u en tly pre
v a ils in deductions drawn from casual thought
about so cio lo g ic al g e n e ra liz a tio n s.
3. Such a theory may lead to the discovery of new
knowledge, or to new areas for fu tu re in v e s ti-
^ J . H. Woodger, "The Technique of Theory Construc
tio n , " J^nternatjlonal.Jiincy^OEedia,^^ Vol.
II: Foundations of the Unity of Science. No. 5 (Chicago:
The U n iv ersity of Chicago P re ss, 1939.
5
g a tio n .4
O rganization of the D is s e r ta tio n
The study i s presented according to the follow ing
plan.
Chapter I I s p e c if ie s the logic of theory co n stru c
tio n to be used in the development of a theory of in te ra c
t io n . This chapter c o n s is ts of th re e p a r ts . The f i r s t p a rt
is devoted to the g en eral conception of theory to be used.
The second p a rt s t a t e s the logic of concept form ation. The
t h i r d p a rt p re se n ts the logic of r e l a t i o n s between proposi
t io n s .
Chapter I I I p re se n ts the th eo ry of in te r a c tio n on
two le v e ls . The f i r s t p a rt of the ch ap ter s t a t e s the theory
in the language of speech, with nominal d e f in it io n s of i t s
concepts. (By nominal d e f in it io n is meant a s t i p u l a t i o n
th a t a s p e c ifie d ex p ressio n is to be synonymous with a ce r
t a i n o th er ex p ressio n whose meaning i s assumed to be d e te r
mined or agreed upon.) This is termed the in c lu siv e th e o ry ,
because i t s meaning is quite broad. Since in t h i s form the
theory i s im precise and su b ject to numerous in te r p r e t a ti o n s ,
the in c lu siv e th eo ry i s next in te rp r e te d in mathematical
term s f o r a r e s t r i c t e d range of phenomena: s p e c i f i c a l l y ,
4See a ls o l i s t s of the value of theory proposed by
Robert K. Merton, S o cia l Theory and S o cia l S tr u c tu r e ,
(Glencoe: Free P re ss, 1967), PP• 96- 97; and Woodger, op.
c i t . . pp. 74-78.
6
same-sex a d o le sc e n t groups. The th e o ry in t h i s form ap p ears
as a s e t of t e n l i n e a r e q u a tio n s , fo u r of which are d e s ig
nated axioms, and th e o th e r s ix of which can be deduced from
th e s e axioms.
C hapter IV r e p o r ts an e m p iric a l study conducted to
t e s t p a r t of t h i s r e s t r i c t e d th e o ry . A random sample of
same-sex a d o le sc e n t groups from a suburban Southern C a l i f o r
n ia high school c o n s t i t u t e d th e s u b je c ts f o r the e m p iric a l
study. C hapter IV d e s c rib e s th e a c tu a l conduct of th e r e
search te c h n iq u e s , and c h a r a c t e r i s t i c s of th e u n iv e rs e and
the sample.
C hapter V d e s c r ib e s th e f in d in g s of th e e m p iric a l
study as th ey apply to the r e s t r i c t e d th e o ry . The f i r s t
p a r t of the c h a p te r e v a lu a te s the hy p o th eses d e riv e d from
the th e o ry in term s of th e fin d in g s of th e e m p iric a l study.
The second p a r t of the c h a p te r e v a lu a te s th e t h e o r e t i c a l im
p o rtan c e of th e co n cep ts in r e l a t i o n to t h i s p a r t i c u l a r th e
ory. I t a ls o p r e s e n ts an e v a lu a tio n of th e g e n e ra l te c h
n iques of th e o ry c o n s tr u c tio n u t i l i z e d , and su g g ests c e r t a i n
m o d ific a tio n s of the th e o ry .
C hapter VI i s a summary of th e d i s s e r t a t i o n .
CHAPTER I I
THE LOGIC OF THEORY CONSTRUCTION
In tro d u ctio n
The purpose of t h i s chapter i s to develop the logic
necessary fo r the co n stru c tio n of a theory of in te r a c tio n .
This chapter i s presented in th re e p a r ts . The f i r s t p a rt
s e ts fo rth the general conception of s c i e n t i f i c theory which
w ill serve as a guide to the r e s t of the study. The second
section d escrib es the logic of concepts to be used in the
theory; ( l) the t h e o r e tic a l c r i t e r i a fo r c o n stru c tio n of
concepts; (2) the s ta tu s of concepts as " fre e c re a tio n s of
i n t e l l e c t ; " and (3) the logic of " p a r ti a l in te r p r e t a ti o n ."
The t h i r d p a rt of the chapter develops the logic of proposi
tio n s to be used in theory c o n s tru c tio n . In t h i s p a r t, the
form in which the p ro p o sitio n s w ill be presented i s f i r s t
discussed. Next, by using as an example a theory previously
presented by Z e tte rb e rg , an explanation i s presented of (1)
the methods of deduction to be used in theory c o n stru c tio n ,
and (2) the assumptions which must be made to make deduc
tio n s from such a theory and to determine the congruence be
tween em pirical data and the theory.
7
8
A General Conception of S c ie n tif ic Theory
Robert K. Merton described six ty p es of endeavors
which are commonly r e f e r r e d to as theory:
1. Methodology, or the logic of s c i e n t i f i c proce
dure .
2. General s o c io lo g ic a l o r ie n ta tio n s , i . e . , broad
p o s tu la te s which in d ic a te the ty p es of v a ria b le s
which are to be taken in to account.
3. A nalysis of s o c io lo g ic a l concepts (meaning anal
y s is or d e f i n i t i o n ) .
4. Post-factum so c io lo g ic a l in t e r p r e t a t i o n s , which
apply presumed "laws" to "explain" d a ta .
5. Em pirical g e n e ra liz a tio n s in sociology.
6. True so c io lo g ic a l th e o ry , which, although Merton
does not e x p l i c i t l y define i t , i s e v id e n tly a
number of p ro p o sitio n s on a h igher le v e l of
a b s tra c tio n th a t 5, specifying g en e ra l r e l a t i o n
ships among a se t of common concepts.^*
M erton's work i s a good beginning fo r our d e f in it io n
of th eo ry : a number of p ro p o s itio n s on a higher le v e l of ab
s tr a c tio n th a n em p irical g e n e ra liz a tio n s , specifying r e la -
2
tio n s among a s e t of common v a r ia b le s .
^Robert K. Merton, S o cial Theory and S o c ia l S truc
tu re (Glencoe: Free P re ss, 1957) j pp.""86-99.
*This i s not th e place to argue about what le v e l of
a b s tr a c tio n i s necessary in order to j u s t i f y the use of the
term, th e o ry . Campbell i n s i s t s th a t only those systems of
Herein, the term, th eo ry , w ill be applied to any set
of i n te r r e l a te d p ro p o sitio n s some of the terms of which are
def in e d .
A theory so conceived c o n s is ts of th re e d i s t i n c t
p a r ts : (1) a se t of p ro p o sitio n s s ta tin g r e la tio n s h ip s among
concepts; (2) a set of undefined term s, the r e f e r e n ts of
which are assumed to be agreed upon (c a lle d observation
term s); and (3) a se t of r u le s fo r connecting the concepts
of the p ro p o sitio n s to the undefined term s, i . e . , a se t of
d e f in it io n s . A completely e x p l i c i t theory co n tain s a set of
r u le s which specify perm issible lo g ic a l operations and com
b in a tio n s of the theory. These r u le s are u su a lly the r u le s
of lo g ic and/or mathematics, and are u su a lly not enumerated
fo r each th e o ry , although they form an im p lic it p a rt of i t .
In a u se fu l s c i e n t i f i c th e o ry , the set of a b s tra c t
p ro p o sitio n s when in te rp re te d by the set of d e f in it io n s must
imply confirmable em pirical consequences. As an example of
a u s e le s s theory, one might be co n stru cted using the ab
s t r a c t symbols of the mass-energy equation,
p ro p o s itio n s of high g e n e ra lity and a b s tra c tio n be so la
beled. (Norman R. Campbell, "The S tru c tu re of T h e o rie s ,”
Readings in the Philosophy of Science, ed. Herbert F eig l.a n d
May Broadbeclc t N e w York: A ppleton-C entury-C rofts, In c . ,
1953J, pp. 288-308.) Hempel m aintains t h i s i s an a r t i f i c i a l
d i s t i n c t i o n . (C arl G. Hempel, "The T h eo reticia n s Dilemma,"
Minnesota S tudies in the Philosophy of S cience. Vol. I I :
bbncepts. T heories and the Mind-Bodv Problem, ed. Herbert
t e i g l , Michael deriven, and drover Maxwell (.Minneapolis:
U n iv e rsity of Minnesota P re s s , 1958J .)
For an example of a theory in which even the lo g i
i d
e = me ^
in te rp re te d as so c io lo g ic a l v a r ia b le s , where e i s group
p ro d u c tiv ity , m i s the number of persons in the group, and
c i s the average in te llig e n c e of the group. I t seems doubt
f u l th a t t h i s would prove to be a confirmable p ro p o sitio n .
Yet i t i s a confirmable p ro p o sitio n in a u se fu l theory when
given i t s usual p h y sical in te r p r e ta tio n : i . e . , i t has con
firm able em pirical consequences.
This theory i s not presented as a p a rt of some a l l -
in c lu siv e th e o r e tic a l system. I t w ill i n i t i a l l y attem pt to
explain known f a c t s — f a c t s already demonstrated by previous
em pirical so c io lo g ic a l re se arc h . From t h i s theory w ill then
be deduced a d d itio n a l f a c t s to be v e r if ie d by em p irical in
v e s tig a tio n . But the s ta rtin g po in t w ill be the s e ttin g up
of p r in c ip le s from which already-known f a c t s can be derived,
and th e se known f a c t s are quite lim ite d in range. In t h i s
sense, i t might be considered a "theory of the lower
range. T his assumption i s im p lic it in a l l th a t follow s:
i t is p r a c t i c a l to develop th e o rie s of higher a b s tra c tio n
only a f t e r the lo w er-lev el th e o rie s which they include have
been developed. This i s an assumption the u t i l i t y of which
has been demonstrated in other scien ces, but whose value to
c a l r u le s used have been made completely e x p l i c i t , see J . H.
Woodger, "The Technique of Theory C o n stru c tio n ," jog. c i t .
4As co n tra ste d with " th e o rie s of the middle range"
discussed by Merton, ,og. c i t . . pp. 9, 280, 328.
11
to sociology rem ains to be d eterm in ed .^
The Logic of Concept Formation
In th e development of the lo g ic of concept forma
t i o n , th r e e to p ic s w i l l be co n sid ered :
1. The t h e o r e t i c a l c r i t e r i a fo r the c o n s tru c tio n of
co n cep ts.
2. The s t a t u s of concepts as " f re e c r e a tio n s of the
i n t e l l e c t . "
3. The lo g ic of " p a r t i a l i n t e r p r e t a t i o n . 1 1
In the d is c u s s io n which fo llo w s, the term , concept, w ill be
used to mean th e name of a n o n lin g u is tic e n t i t y such as a
p ro p e rty , c l a s s , r e l a t i o n , fu n c tio n , or th e l i k e . The meas
urement of a concept w ill be used in the sense of measure
ment of th e p ro p e rty , c l a s s , f u n c tio n , r e l a t i o n , e t c . , which
th e concept names.
T h e o re tic a l c r i t e r i o n f o r
c o n s tru c tio n of co n cep ts
The p o s it io n ta k en in t h i s d i s s e r t a t i o n i s t h a t in
^This s tr a te g y fo llo w s the su g g estio n of Sheldon:
"The h i s t o r y of science in d ic a te s t h a t the most f r u i t f u l
th e o r i e s have been developed to e x p la in known f a c t s . . . .
From th e o r i g i n a l th e o r ie s are then deduced a d d itio n a l
f a c t s to be d isc o v ered in n a tu re , but the s t a r t i n g p o in t
has always been th e s e t t i n g up of p r i n c i p l e s from which
a lre ad y known f a c t s could be d e riv e d , and th e known f a c t s
have always been q u ite lim ite d in ra n g e ." (Richard Sheldon,
"Some O b serv atio n s on Theory in S o c ia l S cien ce," Toward a
Q eneral Theory of A c tio n , e d . T a lc o tt P arsons and E
S h ils [Cambridge: Harvard U n iv e rs ity P re s s , 1951].)
Edward
order to make progress as a science, sociology w ill have to
be b u i l t around concepts constructed fo r th e i r th e o r e tic a l
or system atic importance, r a th e r than fo r t h e i r so c ial im
portance. A concept is said to have th e o re tic a l importance
if i t can be shown th a t i t has close re la tio n s h ip s with a
small number of other concepts which to g eth er with i t ex
p la in a major p a rt of the v a ria tio n in em pirical phenomena
of i n t e r e s t . ^ A concept w ill be said to have so cial im
portance i f i t re p re se n ts a focus of public thinking on a
so c ial problem, or if i t conceptualizes an important so cial
value. Many tim es the so c io lo g ist w ill want to begin h is
work with a concept of so c ia l importance, since he w ill be
seeking so lu tio n s to problems of so c ia l i n te r e s t. For ex
ample, the so c io lo g ist may have the object of determining
the conditions associated with delinquency. The concept,
delinquency, has so c ia l importance. Now when the sociolo
g is t s e ts out to fin d the causes of delinquency, he i s , in
e f f e c t , se ttin g out to co n stru c t a theory of delinquency:
he is searching fo r the r e la tio n s h ip between the concept
delinquency and other concepts. Should i t tu rn out th a t no
very clo se r e la tio n s h ip s e x is t between delinquency and other
concepts, we would say th a t delinquency has l i t t l e th e o r e ti-
or a more d e ta ile d treatm ent of t h i s topic see
C arl G. Hempel, "Fundamentals of Concept Formation in Em
p i r i c a l S c ie n c e ,’’ In te rn a tio n a l Encyclopedia of Unified
Science. Vol. I I ; Foundation si of the Unity of Science. ~ No.
7 (Chicago: U n iv ersity of Chicago Press, 1952).
13
c a l im portance. This would make a th eo ry of delinquency
d i f f i c u l t , cumbersome, and of low p re d ic tiv e power.
The s o c io lo g ic a l t h e o r i s t , th e n , i s in te re s te d in
b u ilding th e o r ie s with concepts which have high t h e o r e t i c a l
importance, whether or not they have any immediate s o c ia l
importance in them selves.
This i s not to say t h a t the s o c io lo g ic a l t h e o r i s t i s
not concerned with im portant s o c ia l is s u e s . As a m atter of
f a c t , the p o in ts of d e p a rtu re fo r s o c io lo g ic a l th e o ry w ill
probably continue to be im portant s o c ia l is s u e s . When the
concepts of s o c ia l importance prove to be of l i t t l e th e o r e t
i c a l im portance, however, the s o c io lo g is t who follow s t h i s
p o in t of view w ill attem pt to rep la ce them with other con
c e p ts , r e f e r r in g to th e same g en e ra l area of phenomena, but
of g re a te r t h e o r e t i c a l im portance.
Concepts as " fre e c r e a tio n s of
the i n t e l l e c t '1
W ithin the bounds of the c r i t e r i o n of t h e o r e t i c a l
im portance, concepts are to be considered " fre e c r e a tio n s of
the i n t e l l e c t . " By t h i s we mean only t h a t they are con
stru c te d according to the purposes of the s c i e n t i s t , and are
not imposed by the nature of the n o n lin g u is tic world of
ev e n ts. This i s a p o in t of view e v id e n tly not shared by
some o ther w r ite r s on concept f o rm a tio n .7 Gouldner, fo r
7 R einhardt Bendix and Bennett B erger, "Images of
Society and Problems of Concept Formation in Sociology,"
14
example, c r i t i c i z e s Parsons fo r not understanding th a t "the
elements of s o c ia l systems cannot be merely c o n s titu te d
a p r i o r i , but must be in d u c tiv e ly sought and em p iric ally
Q
v a lid a te d ." But i f the elements of so c ia l systems are un
derstood to be concepts, one does not in d u c tiv e ly seek a
concept. Concepts cannot be discovered; they must be in
vented.
L azarsfeld and Barton are s im ila rly m isleading when
they s ta te th a t one of the c r i t e r i a fo r the c o n s tru c tio n of
c a te g o rie s in which to c l a s s i f y data is the c a te g o ry 's
"adaptation to the stru c tu re of the s i t u a t i o n ." ^ Do s itu a
tio n s have a p a r tic u la r s tru c tu re ? Or does the s c i e n t i s t
impose a s tru c tu re on them? There i s an i n f i n i t e number of
ways to stru c tu re any s i t u a ti o n . The one chosen depends on
the purposes of the s t r u c t u r e . ^
Symposium on S o cio lo g ic a l Theory, ed. Llewellyn Gross
(Evanston: Row, P eterson and Company, 1959); Alvin W .
Gouldner, "R eciprocity and Autonomy in F unctional Theory,"
i b i d .
^ I b i d . , p . 245.
^Paul F. L az arsfe ld and Allen H. Barton, "Q ualita
tiv e Measurement in the S o cial Sciences: C l a s s i f ic a t io n ,
Typology and In d ic e s," The P o licy S cie n ces, ed. Daniel
Lerner and Harold L. Lasswell (Palo Alto: S tanford Univer
s it y P re ss, 1951), p. 156.
^•^Sheldon p u ts i t t h i s way: " I t must be remembered
th a t the order i s in the p r in c ip l e s , and th a t d if f e r e n t
kinds of order w ill emerge w ith the use of d if f e r e n t kinds
of p r in c ip le s . . . . A se t of data does not impose t h e o r e t i
c a l c a te g o rie s which w i l l have to be used." (Sheldon, o p .
c i t . , p . 38.)
15
The logic of p a r t i a l
i n t e r p r e t a t i o n : a modified
view of o p e ra tio n a lism
Since Bridgem an's c l a s s i c o p e r a t i o n a l i s t statem ent
in 19^8,^ p h ilo so p h e rs and s c i e n t i s t s have attem pted to ex
p l i c a t e the lo g ic of the connection between a concept and
i t s r e f e r e n t s in somewhat d i f f e r e n t ways than b efo re . Much
of t h e i r d isc u ssio n a p p lie s to the most a b s tr a c t concepts in
t h e o r e t i c a l p h y sic s only, and i s of no immediate i n t e r e s t in
12
form ation of concepts in so cio lo g y . This much i s agreed
upon, however: any s c i e n t i f i c th eo ry must be b u i l t upon un
d e fin e d term s ("o b se rv a tio n te rm s " ), the re fe re n c e of which
i s assumed to be unambiguous. The e m p iric a l co n ten t of a l l
other concepts of the th e o ry th e n r e s t s upon t h e i r lo g ic a l
co nnection to the undefined term s. This lo g ic a l connection
between t h e o r e t i c a l concepts and o b se rv atio n term s may be
through d e f i n i t i o n or s e ts of red u c tio n se n ten c es. So-
c a lle d o p e ra tio n a l d e f i n i t i o n of a concept w i l l h e re in be
considered as unambiguous d e s c r ip tio n of i t s measurement
o p e ra tio n in o b se rv a tio n term s.
Whether the o p e r a tio n a l d e f i n i t i o n of a concept
U p . W. Bridgeman, The Logic of Modern P h y sics (New
York: The Macmillan Company” 1928).
^■^Consult H erbert F e ig l and May Broadbeck, oq, c i t . .
pp. 235-381.
^3For a d is c u s s io n of th e lo g ic of re d u c tio n sen
te n c e s , see C a rl G. Hempel, "Fundamentals of Concept Forma
t i o n in E m pirical S c ie n c e ," pp. 23-29.
16
exhausts the meaning of the concept is s t i l l a moot p o i n t . ^
The s o lu tio n which seems most sen sib le to t h i s w r ite r a t
p rese n t in sociology w i l l be o u tlin ed h e r e . 1^ Let us begin
w ith the p h y s ic a l concept of tem p eratu re. Consider temper
a tu re to be o p e ra tio n a lly defined as the read in g of a mer
cury thermom eter. Thus d efin e d , the concept has no meaning
below the fre e z in g p o in t of mercury or above i t s b o ilin g
p o in t. Can we then not speak of tem peratures beyond t h i s
ra^ge? Not u n t i l f u r th e r d e f i n i t i o n i s provided. I t hap
pens th a t the gas laws r e l a t e the tem perature of a gas to
i t s p ressu re by a c e r t a i n form ula. Now c e r t a in gases remain
gases over a v a s tly wider range than the liq u id range of
mercury, and a ls o include t h i s range. This rev ea led the
p o s s i b i l i t y of a concept of tem perature based on the p res
sure of a gas a t a co n stan t volume. The gas laws provide a
conversion formula fo r the two d if f e r e n t o p e ra tio n s of
measurement of tem perature in the range of liq u id mercury.
R igid o p e ratio n alism would re q u ire th a t th ese two d i f f e r e n t
o p e ra tio n s f o r measurement of the concept be regarded as two
d i f f e r e n t concepts of te m p e ra tu re : tem p eratu rem and temper-
a tu re g , which are r e la te d by an e m p iric a l law.
More re c e n t t h e o r e t i c a l work has made i t seem more
*4 I b i d . , and Hempel, "The T h e o r e tic ia n 's Dilemma,"
o p . c i t .
^ T h i s is a m o d ificatio n of the o r i g i n a l argument
of Bridgeman, oq* c i t . . in the li g h t of Hempel's a n a ly s is .
.1 7
f e a s i b l e , in order to f a c i l i t a t e the c o n s tru c tio n of ab
s t r a c t th e o ry , to co n sid er te m p era tu rem and tem p eratu reg as
p a r t i a l i n t e r p r e t a t i o n s of a t h e o r e t i c a l concept of tempera
tu r e , which may a lso re c e iv e a d d itio n a l i n t e r p r e t a t i o n s .
Two o p e ra tio n s cannot be co n sid ered as p a r t i a l i n t e r p r e t a
t i o n s of the same concept u n le s s th e re i s a law r e l a t i n g the
two such t h a t i t allow s the s u b s t i t u t i o n of the measurements
of one f o r th e other in laws covering the domain where t h e i r
measurements o v erla p . Furtherm ore, i t would be im possible
t o e s t a b l i s h the r e l a t i o n s h i p between the two o p e ra tio n s un
le s s t h e i r domains overlapped.
Applied to s o c io lo g ic a l co n cep ts, th e meaning of the
example above i s c l e a r . I f two s o c io lo g is ts c o n s tru c t in
strum ents to measure a concept, s o c ia l c l a s s , l e t us c a l l
them c la s s y and c l a s s 2 » both r e f e r r i n g to the same domain
(iapplicable to the same range of s u b je c ts ) , th e se may be cor>-
sid e red as two p a r t i a l i n t e r p r e t a t i o n s of th e same concept
i f , and only i f , th e re i s a conversion formula by which
c la s s ^ measurements may be s u b s titu te d in a l l the laws in to
which the concept c la s s 2 measurements form tru e em p irical
laws. I f t h i s co n d itio n does not h o ld , th e re i s l i t t l e
p o in t in co n sid erin g both the s c a le s t o be i n t e r p r e t a t i o n s
of a sin g le t h e o r e t i c a l concept, since th ey obviously do not
measure the same phenomenon.
In sociology as w ell as o th e r s o c ia l sc ien c es the
question as to whether two d e f i n i t i o n s are p a r t i a l in te r p r e -
i
18
ta tio n s of the same concept is answerable only as a matter
of degree. In the example of c la s s , suppose th a t measure
ments of c l a s s i and c l a s s 2 a?e law fully connected to the
extent th a t t h e i r r^ (as discussed by R obinson^) = .95.
Would t h i s be accepted as a s a tis f a c to ry le v e l of agreement
between the two concepts? Would a lower r^ be acceptable?
Probably at t h i s point in the development of sociology, con
siderably lower c o e f f ic ie n ts would be accepted. At a l a t e r
date, one might wish to set higher standards of correspond
ence between p a r t i a l in te rp r e ta tio n s . This d ecisio n would
have to be made in terms of the p re c is io n of the theory into
which the terms were to e n te r, as w ell as the accuracy of
measurement of the other concepts of the theory.
I t should be noted here th a t i t is not reference
back to a single nominal d e f in itio n which determines whether
two operations are p a r t i a l in te rp r e ta tio n s of the same con
cept. Two operations may in a general way seem to in te r p r e t
the same nominal d e f in itio n , and yet not f u l f i l l the condi
tio n s set fo rth fo r being p a r t i a l in te rp r e ta tio n s of the
same concept. Furthermore, two operations developed from
two d if f e r e n t nominal d e f in itio n s may n evertheless be par
t i a l in te r p r e ta tio n s of the same concept i f they f u l f i l l the
two co n d itio n s: (1) th ere is an em pirical law connecting the
S. Robinson, "The Geometric In te rp re ta tio n of
Agreement," American Sociological Review. XXIV (June, 1959),
338-45.
1 9
two measurements in such a way th a t (2) the values of one
can be s u b s titu te d in the "tru e" laws f o r the o th e r, r e t a i n
ing the same c o n s ta n ts in the laws, and s t i l l r e s u l t in
"tru e" l a w s . ^
In t h i s study, a l l concepts to be used in the theory
w ill e v e n tu a lly be o p e ra tio n a lly d efin e d . These o p e ra tio n a l
d e f i n i t i o n s are to be considered p a r t i a l i n t e r p r e t a t i o n s of
the concepts of the th eo ry f o r t h e i r p a r t i c u l a r measurement
domain, since th e re w i l l always remain a p o s s i b i l i t y of ad
d i t i o n a l i n t e r p r e t a t i o n s . T his leaves open the o p p o rtu n ity
fo r a d d itio n a l i n t e r p r e t a t i o n s of each concept fo r th e same
and o th e r domains by means of conversion form ulae, should
th ese be e m p iric a lly determ ined.
I t may be noted from the d is c u s s io n above t h a t in
e s ta b lis h in g meaning fo r t h e o r e t i c a l concepts, the t r a d i
t io n a l dichotomy between d e f i n i t i o n a l and e m p irical s t a t e
ments becomes b lu rre d . Formerly d e f i n i t i o n was thought of
(except in th e a rc h a ic sense of " r e a l d e f in it io n " ) as being
s t r i c t l y a r b i t r a r y and co n v en tio n al. Under the io g ic of
p a r t i a l i n t e r p r e t a t i o n s d escrib ed above, two d i f f e r e n t oper
a tio n a l d e f i n i t i o n s may be sa id to be d e f i n i t i o n s of the
same c o n c e p t, not because of a lo g ic a l connection between
the two d e fin itio n s * but because of a s p e c ific type of
17
For the development of the logic supporting th e se
c r i t e r i a , see May Broadbeck, "Models, Meaning, and Theo
r i e s , " in G ross, 0£ . c i t . . pp. 348-72.
20
e m p iric a l r e l a t i o n s h i p between the measurements obtained by
the two d i f f e r e n t s e ts of o p e ra tio n s . In t h i s sense, lo g i
c a l and e m p iric a l statem en ts are combined in the d e f i n i t i o n
a l p ro c e s s, in order to make a b s tr a c t th e o ry p o s s ib le .
Summary of the lo g ic of
concept form ation
From a study of the lo g ic of concept form ation, the
follow ing p o in t of view has been adopted f o r use in t h i s
d i s s e r t a t i o n :
1. Concepts should be c o n s tru c te d f o r s o c io lo g ic a l
th e o ry on the b a s is of t h e i r t h e o r e t i c a l impor
ta n c e , and not on the b a s is of t h e i r s o c ia l im
p o rta n c e .
2. Concepts f o r t h i s th e o ry w ill be developed from
s o c ia lly im portant concepts by reducing the so
c i a l concepts to an elem entary le v e l of a b s tra c
t i o n in an e f f o r t to achieve s p e c i f i c i t y and
t h e o r e t i c a l im portance.
3. W ithin the requirem ent of t h e o r e t i c a l impor
ta n c e , the concepts of t h i s th e o ry should be
considered as " fre e c r e a tio n s of the i n t e l l e c t , "
t h a t i s , in no way imposed upon the s c i e n t i s t by
the n atu re of the ev en ts of th e world, but con
s tru c te d to s u it h i s purposes.
4. The o p e ra tio n a l i n t e r p r e t a t i o n of each concept
t o be p re se n te d in t h i s th e o ry should be consid-
ered a single p a r t i a l in te r p r e t a ti o n of the con
c e p t. Other p a r t i a l in te r p r e t a ti o n s may be con
sidered to be in te r p r e t a ti o n s of the same con
cept providing they meet the c o n d itio n s of sub
s t i t u t i o n which have been s ta te d .
The Logic of P ro p o sitio n s
The development of the logic of p ro p o sitio n s w ill be
concerned with two major to p ic s . The f i r s t to p ic i s the
form to be adopted fo r the p ro p o sitio n s of a formal theory.
Three p a tte r n s are d iscussed: the language of speech, lo g i
cal symbolic n o ta tio n , and mathematical equation. The sec
ond to p ic of t h i s se ctio n is the logic of deduction. By
using a p reviously developed deductive theory presented by
Z etterb erg , the usual lo g ic used in deductive theory in so
c i a l science is explained. The problems ra is e d by t h i s
deductive method are d isc u sse d . I t i s shown th a t i t is
necessary to make c e r t a in assumptions in order to make de
ductions from such a th e o ry , and c e r ta in a d d itio n a l assump
tio n s in order to ev alu ate the congruence between any empir
ic a l data and the th e o ry .
P a tte rn s fo r a formal theory
In the c o n s tru c tio n of a formal theory of in te ra c tio n
or of any theory in sociology, th ere are se v eral a v a ila b le
p a tte rn s from which to choose the one b est su ite d to the
level' of development of rese arch in a p a r tic u la r su b stan tiv e
22
a re a . Each p a t t e r n , however, sh ares th e se c h a r a c t e r i s t i c s :
1. The th e o ry i s p resen ted in the sm a lle st p o ssib le
number of p ro p o s itio n s .
2. This minimum number of p ro p o s itio n s i s so con
s t i t u t e d t h a t the r e l a t i o n s h i p s among them can
be s ta te d w ith some degree of p r e c is io n .
3. These p ro p o s itio n s are so r e l a t e d th a t o th er
p ro p o s itio n s can be derived from them by deduc
t i o n ( lo g ic a l or m athem atical).
Any s e t of p ro p o s itio n s having th e se th re e charac
t e r i s t i c s as w ell as a s e t of d e f i n i t i o n s connecting the
concepts to a se t of undefined term s, w ill h e re in be r e
f e r re d to as a form al th e o r y . The minimum s e t of p ro p o si
tio n s w ill be c a lle d axioms. The p ro p o s itio n s which can be
deduced from the axioms w i l l be c a lle d th e o re m s.10
The most common form of form al th e o ry in sociology
p re s e n ts th e axioms and theorems in words and sentences of
the u su a l language of speech. Recent p r e s e n ta tio n s in t h i s
form have appeared by Arnold R o se,19 L lew ellyn G ro s s,20 and
18
W e follow here the term inology used by H erbert
Hochberg, "Axiomatic Systems, F o rm a liz atio n , and S c ie n tif ic
T h e o rie s ," in Gross, i b i d . . pp. 407-37.
^ A rn o ld Rose, "A Deductive Id e a l Type Method,"
American J o u rn a l of S o cio lo g y . LVI (1950), 35-42.
^^Llewellyn G ross, "Theory C o n stru ctio n in S o cio l
ogy: A M ethodological In q u iry ," Symposium on S o c io lo g ic a l
Theory, ed. L lew ellyn G ross, oq. c i t .
23
Hans Z e tte r b e r g .2^ Here is a sample p ro p o sitio n in t h i s
form, from Gross: "In so ciety X upper- and lo w e r-statu s
groups se p arately a sso cia te to enlarge t h e i r means of ac-
quiring scarce goods." I t is conversion to t h i s kind of
e x p l i c i t p ro p o sitio n which c o n s titu te s the f i r s t step in
c o n stru c tio n of a formal theory. But t h i s remains a f i r s t
step . I t i s d i f f i c u l t to avoid ambiguity in such statem ents
even when a l l the terms are c a re fu lly defin ed . In drawing
deductions, the lo g ic a l im p licatio n s of a set of such propo
s itio n s are d i f f i c u l t to see. Furthermore, some kinds of
re la tio n s h ip s are very d i f f i c u l t to s ta te in common language
form. The next step in fo rm alizatio n is the red u ctio n of
these statem ents to symbolic form in which lo g ic a l manipula
tio n i s f a c i l i t a t e d . Here th e re are two a l te r n a ti v e s : (1)
lo g ic a l symbols and (2) mathematical symbols. Here is a
p ro p o sitio n from Woodger, set fo rth in h is own lo g ic a l sym
bols:
(xCy) equiv (xT & T“ ^ y ) .2^
Woodger ex p lain s th a t in words t h i s means approximately the
same as "A th in g x co in cid es in time with a th in g y i f , and
only i f , x stands in T to y and y stands in T to x." (The
r e l a t i o n T i s prev io u sly d e fin e d .) The meaning of t h i s
21
A Hans L. Z e tte rb erg , On Theory and V e r ific a tio n in
Socioloqy (New York: T re s s le r P re ss, 1954).
22Gross, o p . c i t . , p. 549.
2^Woodger, _ 0£. c i t . , p. 36.
24
p r o p o s itio n i s not the s l i g h t e s t i n t e r e s t in the argument
being developed h e re . The p r o p o s itio n i s included only as
an example of p ro p o s itio n s s ta te d in lo g ic a l symbols. Such
lo g ic a l symbolism i s considered by some lo g ic ia n s to be of
wider a p p lic a tio n th an m athem atical symbols, since i t is
a p p lic a b le t o c l a s s i f i c a t o r y and com parative conceptual
schemes as w ell as m e tric a l schemes, and since i t can even
fu n c tio n as a method fo r the in v e n tio n of new m athem atics.
P ro p o sitio n s se t f o r th in th e form of m athem atical
e q u a tio n s are a p p ro p ria te where f u l l y m etric concepts are
d e a lt w ith . In view of the f a c t t h a t many of the concepts
in prev io u s re se a rc h in sm all-group i n t e r a c t i o n have been or
can be measured in a t l e a s t i n t e r v a l s c a le s , i t was decided
to attem pt th e c o n s tru c tio n of the p rese n t th e o ry using
m athem atical p ro p o s itio n s , a t l e a s t in p a r t .
N ev erth eless the s t a r t i n g p o in t w i l l be a se t of
p ro p o s itio n s in the language of speech. Z e t t e r b e r g 's f o r
m a liz a tio n of Durkheim's th e o ry of d iv is io n of la b o r has been
s e le c te d as an example by means of which to explore the
problems of deduction from a th eo ry in t h i s form .^4 By
c a r e f u l exam ination of the problems involved in making ex
p l i c i t the lo g ic of Z e tt e r b e r g 's th e o ry , i t i s p o ssib le t o
develop the necessary r u l e s f o r c o n s tru c tio n of th e i n t e r
a c tio n th e o ry which i s the main concern of the d i s s e r t a t i o n .
^4Z e tte rb e rg , op. c i t .
25
The logic of deduction
In order to e x p la in the logic of deduction used in
the most deductive th eo ry in sociology, the Z etterb erg f o r
m alizatio n of a th eo ry based on six concepts of Durkheim's
D ivision of Labor was se le c te d as an example. I t should be
borne in mind th a t t h i s theory i s being used p u rely to ex
p la in the logic of deduction, and i t s su b stan tiv e content is
of no i n t e r e s t whatsoever in t h i s reg a rd .
Z e tte r b e r g 's fo rm a liz a tio n of Durkheim. Hans Zet
te rb e rg presented th e follow ing form alized th e o ry , embodying
six of the concepts Durkheim used in h is D iv isio n of Labor
(d iv is io n of la b o r, s o l i d a r i t y , u n ifo rm ity , number of mem
b e rs, d e v ia tio n , and r e j e c t i o n of d e v ia te s) in fo u rte e n
p ro p o sitio n s :
Axioms:
1. The g r e a te r the d iv is io n of la b o r, the g r e a te r the
s o l i d a r i t y .
2. The g r e a te r the s o l i d a r i t y , the g re a te r th e uniform
i t y .
3. The g re a te r the number of members, the g r e a te r the
d iv is io n of la b o r.
4. The g re a te r the s o l i d a r i t y , the le s s the r e j e c t i o n
of d e v ia te s .
Theorems:
5. The g r e a te r the d iv is io n of la b o r, the more the
u n ifo rm ity .
6. The g r e a te r the s o l i d a r i t y , the g re a te r the number
of members.
26
7.
The g re a te r the number of members, the le s s the
d e v ia tio n .
8. The more the uniform ity, the
d e v ia te s.
le s s the r e je c tio n of
9.
The g re a te r the number of members, the le s s the
je c tio n of d e v ia te s.
re -
10. The g re a te r the d iv is io n of
v ia tio n .
lab o r, the le s s the de-
11. The le s s the d e v ia tio n , the
d e v ia te s.
le s s the r e je c tio n of
12. The g re a te r the s o li d a r ity ,
of d e v ia te s.
the le s s the r e je c tio n
13. The g re a te r the number of members, the more the
uniform ity.
14. The g re a te r the d iv is io n of
r e j e c t i o n of d e v ia te s .^ 5
la b o r, the le s s the
He introduces the following concepts as undefined:
behavior, member, group, norm, s o li d a r ity , d iv is io n of labor,
r e je c tio n . The follow ing d e f in itio n s are s tip u la te d :
U niform ity: the proportion of members whose behavior
i s the norm of the group.
D eviation: the proportion of members whose behavior
i s not the norm of the group.
D eviate: member whose behavior is not the norm of
the group.
He then s t a t e s th a t " i t is e a s ily recognized th a t
from these four {'axioms'} a l l fo u rteen fin d in g s can be de-
2^ Z e tte rb e rg 1s lo g ic a l terminology is modified to
f i t the usage adopted fo r the p resen t d i s s e r t a t i o n . I b id . .
pp. 18-19.
2 7
riv e d as th ey are combined with each o th er and with the
nominal d e f i n i t i o n s . " ^
The d e riv a tio n of each of the theorems from the ax i
oms need not be dem onstrated h e r e . Z e tte rb e rg did not f o r
mally dem onstrate the d e r iv a tio n s . But examination of the
complete se t of p ro p o s itio n s shows th a t th e re are two d is
t i n c t ty p es of deductions involved: one type is deduction
through the d e f i n i t i o n s , and the o th er is through t r a n s i t i v e
r e l a t i o n s h i p s .
Deduction through the
'd e fin itio n s
Deductions through the d e f in it io n s can be demon
s tr a te d as follow s:
F i r s t , from the d e f i n i t i o n s , a p ro p o sitio n om itted
by Z e tte rb e rg must be deduced, which i s designated as A.
A. U niform ity + d e v ia tio n = 1. (by d e f in it io n of
p ro p o rtio n , u n i
fo rm ity , and devia
tio n )
By deduction from A,
B. The g r e a te r the u n ifo rm ity , the le s s d e v ia tio n .
W e can now proceed.
7. The g re a te r the number of members, the le s s the
d e v ia tio n .
B. The g r e a te r the u n ifo rm ity , the le s s d e v ia tio n .
The g r e a te r the number of members, the more the
un ifo rm ity (by A, B, and 7).
^ I b i d . . p. 19.
which is Theorem 13. Thus Theorems 7 and 13 are ta u to lo g i
c a l by the d e f in itio n s and t h e i r deductive consequences.
Deduction through t r a n s i tiv e
r e la tio n s h ip s
Deduction through tr a n s i tiv e re la tio n s h ip , the sec
ond type of deduction required in t h i s th e o r e tic a l system,
Z etterberg demonstrates as follows:
5. The g re ate r the d iv is io n of labor, the more the
uniform ity.
13. The g rea te r the number of members, the more the
uniform ity.
The g re a te r the number of members, the g re a te r
the d iv is io n of labor.
Only through the a p p lica tio n of numerous such deductions as
t h i s l a s t can the ten theorems be derived from the four axi
oms. The f i r s t type of deduction— through the d e f in itio n s —
is u n iv e rsa lly v alid . The second type, although widely used
in theory c o n stru c tio n among so c ia l s c i e n t i s t s such as Wal
t e r M a rtin ,27 Robin W illiam s,2® Arnold Rose,2^ and Llewel
lyn G r o s s , ^ r e s t s on c e r ta in assumptions which have appar
en tly not been c a re fu lly examined, or at le a s t not f u lly
27Jack P. Gibbs and Walter T. M artin, "A Theory of
S tatus In te g ratio n and I t s R elation to Suicide," American
S ocio lo g ical Review. XXIII (A pril, 1958), 140-4?.
2®Robin M. Williams, J r . , "C ontinuity and Change in
S ociological Study," American Sociological Review. XXIII
(December, 1958), 619-33.
^ R o s e , .op. c i t .
^ G r o s s , .op. c i t .
29
s ta te d .
Let us examine the t r a n s i tiv e deductive p a tte rn in
d e t a i l . Reducing the l a s t example above to p ro p o sitio n al
fu n ctio n s with an if -th e n s tru c tu re , i t becomes:
If A in creases, then B in c re a se s.
If C in c re a se s, then B in c rease s.
If A in creases, then C in c re ase s.
The r e la tio n s h ip between A and C through t h e i r common r e l a
tio n s h ip to B i s known as a tr a n s i tiv e r e la tio n s h ip . But
there is nothing which allows deduction from t h i s stru c tu re
without a d d itio n a l assumptions as to the nature of the r e la
tio n s h ip .
Assumptions necessary for
the tr a n s i t i v e deduction
I f the nature of the co v a riatio n expressed in the
premises is sp e cified more p re c is e ly , lim ited deductions of
a so rt can be made. Let i t be stip u la te d th a t Pearsonian
product-moment c o rr e la tio n s are being described. Then given
two p ro p o sitio n a l fu n ctio n s
r AB = f
rCB s 9
where r is a Pearsonian c o r r e la tio n , A, B, and C are v a ri
ab les, and f and g are the values of the c o rre la tio n coef
f i c i e n t s , is th ere anything which can be deduced about the
r e la tio n s h ip between A and C? There is .
I t can be demonstrated th a t when r 1 2 and r j ^ are
30
known ( s u b s c r ip t numerals r e f e r r i n g to any v a r i a b l e s ) , the
lim its of p o s s ib le v a r ia tio n of r 2 3 are given by the formu-r
la;31 I --------------------------------------------------------------------------------------2 —
r 23 = r 12r 13t y 1 ' r 12 ' r 13 * r 12r 13
Now in th e case where the c o r r e l a t i o n s r ^ a ^d r-^g are both
1.00, th e c o r r e l a t i o n of r 2 3 i s a ls o 1.00, which g iv es the
i n i t i a l p l a u s i b i l i t y to the o r i g i n a l deduction in Z e tte r
b e r g 's th e o ry . But s o c io lo g is t s are only too w ell aware of
the absence of p e r f e c t c o r r e l a t i o n s between t h e i r v a r ia b le s .
More u su a l in sociology are c o r r e la tio n s of .50 and below.
Some sample v alu es of r 2g fo r d i f f e r e n t v alu es of r-^2 and
r ^ 3 are given below.
When rj_2 and each = .90, l i m i t s of r 23 are ♦ .62 and
+ 1. 00.
When r ^ 2 a^d r^g each = .50, li m i t s of r 2 3 are - .5 0 and
+ 1. 00.
When r^2 and r^g each = .25, li m i t s of r 2g are - .8 8 and
♦ 1. 00.
T his g iv es some idea of the broad li m i t s of the de
d u c tio n s which can be made from low c o r r e l a t i o n s . Now the
Z e tte rb e rg fo rm u latio n r e q u ir e s a whole s e r ie s of ch a in de
d u c tio n s , which can only be made in th e case where lin e a r
c o r r e l a t i o n s of p r a c t i c a l l y 1.00 hold among the v a r ia b le s .
In the p a r t i c u l a r example under c o n s id e r a tio n , i f the c o rre -
^■Quentin McNemar, P sy c h o lo g ic a l S t a t i s t i c s (New
York: John Wiley and Sons, 1953), p. 2l£.
31
la t i o n between degree of d iv is io n of labor and degree of
conform ity were .50, and the c o r r e la tio n between the number
of members of a group and the degree of u niform ity were .50,
logic (mathematics) could deduce only t h a t the c o r r e la tio n
between degree of d iv is io n of labor and number of group mem
b ers must l i e between - .5 0 and +1.00— h ard ly grounds fo r
deducing the p ro p o s itio n Z e tte rb e rg deduces.
Assumptions necessary to f i t
e m p irical «iata to the th eo ry
An a l te r n a ti v e to drawing no deductions at a l l from
moderate c o r r e la tio n s of t h i s s o rt is to p o s tu la te p e r f e c t
c o r r e la tio n s between the v alu es of the concepts of the th e
ory. The deductions can then be made, but the em p irical
s o c io lo g ic a l data c o lle c te d w ill not f i t the th e o ry . This
then r e q u ir e s the a d d itio n a l assumption t h a t e r r o r s in meas
urement have been made, and th a t v a r ia b le s have been l e f t
unaccounted f o r which are a f f e c tin g the r e l a t i o n s h i p . As a
m atter of f a c t , t h i s i s e x a c tly what G u ilfo rd suggests we
assume:
Whenever a r e la tio n s h i p between two v a ria b le s is
e s ta b lis h e d beyond reasonable doubt, the f a c t t h a t the
c o r r e la tio n c o e f f ic ie n t i s small may merely mean th a t
the measurement s it u a ti o n i s contam inated by many th in g s
u n co n tro lle d or not held c o n s ta n t. One can e a s i l y con
ceive of an experim ental s it u a ti o n in which i f a l l
re le v a n t f a c t o r s had been held c o n s ta n t, the c o r r e la tio n
might have been 1.00 r a th e r th an .20. T his lin e of re a
soning in d ic a te s t h a t where any c o r r e la tio n between two
th in g s i s e s ta b lis h e d a t a l l . . . the fundamental law
im plies a p e rfe c t r e l a t i o n s h i p . Thus in nature c o r r e la
t i o n s of zero or 1.00 are the ru le between v a ria b le s
32
32
when is o la te d .
A ctually, c o r r e la tio n s of 1.00 as a p o stu la te of a
theory means th a t when A is found to be r e la te d to B, then
A can be said to be p e r f e c tly r e la te d to B, and hence by
t r a n s i t i v e deduction to everything to which B i s r e l a t e d .
This p u ts us in the absurd p o s itio n of having to m aintain,
when using t r a n s i t i v e deduction, th a t anything which is r e
la te d to one th in g i s also p e r f e c tly r e la te d to anything
e ls e which is r e la te d to anything e l s e . Furthermore, i f A
is not re la te d to B, and B i s r e la te d to something e l s e , we
must conclude th a t A is th e re fo re not r e la te d to anything
e l s e , otherw ise i t would also be r e la te d to B. The so c io l
o g ist who b u ild s t h i s assumption in to h is theory i s not in
an enviable p o s itio n . The only reason fo r such an assump
tio n would be th a t i t c o n s is te n tly leads us to make p red ic
tio n s which are confirmed e m p iric a lly , in sp ite of the f a c t
th a t i t leads u ltim a te ly to a red u c tio ad absurdum.
A ctually, th ere is an escape from t h i s ab su rd ity in
the form of p o s tu la tio n of n o n -lin ea r r e la tio n s h ip s r a th e r
than lin e a r . I t i s the opinion of t h i s w rite r along with
33
many o th e rs th a t e v e n tu ally in te ra c tio n theory w ill be
^2J . P. G uilford, Fundamental S t a t i s t i c s in Psy
chology and Education (New* York":" McGraw-Hill Book Company,
19567", p. 146'.
^^Herbert A. Simon, "A Formal Theory of In te ra c tio n
in S o cia l Groups," Small Groups, ed. A. Paui Hare, Edgar
F. B orgatta, and Robert F. Bales (New York: Alfred A. Knopf,
1955), pp. 132-48.
33
b u ilt around nonlinear r e la tio n s h ip s . At t h i s p oint in the
understanding of so c ia l in te ra c tio n , however, i t i s assumed
th a t we are not yet ready to elab o rate a nonlinear th eo ry .
Therefore a lin e a r theory w ill be tem porarily adopted, ac
cepting the absurd assumption above to t e s t i t s u t i l i t y in
building a theory with p re d ic tiv e value.
Review of the logic of
p ro p o sitio n s
From a study of the logic of p ro p o sitio n s , the f o l
lowing plan fo r a theory of in te ra c tio n has been adopted.
1. The p ro p o sitio n s fo r the theory of in te ra c tio n
w ill f i r s t be sta te d in the language of speech,
then given more p recise statem ent in mathemati
c a l form.
2. The theory w ill be presented in the form of a
set of axioms, and a set of theorems which are
deducible from the axioms.
3. Two general types of deduction w ill be used:
A. Deduction through d e f in it io n s .
B. Deduction through t r a n s i t i v e r e la tio n s h ip s .
4. In order to make the deductions through t r a n s i
tiv e r e la tio n s h ip s i t is necessary to specify
the nature of the c o v a ria tio n . I t w ill be as
sumed th a t the c o v a ria tio n i s lin e a r , and may be
a p p ro p ria te ly described by a lin e a r equation.
5. In order to make u sefu l deductions, i t is neces
sary to assume th a t the theory d escrib es p e rfe c t
lin e a r c o v a ria tio n .
6. In order to evaluate the congruence of em pirical
data to such a theory, i t is necessary to assume
th a t when any two v a ria b le s are em p iric a lly de
termined to have a s ig n if ic a n t c o r r e la tio n , t h i s
be accepted as evidence of p e rf e c t c o rr e la tio n
in th eo ry . D iscrepancies are then accounted fo r
in terms of e r r o r s of measurement and uncon
t r o l l e d sources of v a r ia tio n from extraneous
v aria b le s.
7. These assumptions lead to absurd consequences.
The absurd consequences can be avoided by adopt
ing a n onlinear model, but i t is assumed th a t
p resen t knowledge of the phenomena of in te ra c
tio n at t h i s time is in s u f f ic ie n t to j u s t i f y
e la b o ra tio n of a nonlinear theory. I t is de
cided to continue with the lin e a r model and the
questionable assumptions w ith the proviso th a t
i f they lead to knowledge they are j u s t i f i a b l e
even though they co n tain absurd elem ents.
Summary
This chapter presented the logic of theory construc
tio n to be used fo r the c o n s tru c tio n of a theory of in te ra c
t io n to be presented in the remainder of the paper. Theory
35
c o n s tru c tio n was d isc u sse d under th r e e main d iv is io n s : (1)
a g e n e ra l co n cep tio n of s c i e n t i f i c th e o ry , (2) the lo g ic of
concept fo rm a tio n , and (3) the lo g ic of p r o p o s itio n s ,
1, A g e n e ra l co n c ep tio n of s c i e n t i f i c th e o r y . The
co n cep tio n of s c i e n t i f i c th e o ry developed was as
a s e t of p r o p o s itio n s c o n ta in in g some d e fin e d
term s, to g e th e r w ith a s e t of d e f i n i t i o n s con
n e c tin g the term s of th e p r o p o s itio n s w ith unde
fin e d o b se rv a tio n te rm s . The th e o ry to be p r e
sen ted in t h i s paper w i l l be on a low le v e l of
a b s t r a c t i o n , and designed to e x p la in known
f a c t s . The paper i s based on the assum ption
t h a t the lo g ic a l p r i n c i p l e s of a l l sc ie n c e s are
the same, and t h a t o b s e rv a tio n and lo g ic are the
only o p e ra tio n s p e r m is s ib le in co n firm a tio n of
p r o p o s itio n s .
2. The lo g ic of concept form ation.. I t was con
cluded t h a t p r o p o s itio n s were to be s e le c te d on
th e b a s i s of t h e o r e t i c a l im portance, and t h a t
th ey would be on a low le v e l of a b s t r a c t i o n w ith
narrow r e f e r e n c e . I t was s ta te d t h a t th e y
should be co n sid ered " fre e c r e a t i o n s of th e in
t e l l e c t , " i . e . , n o t imposed by the n atu re of the
d a ta , and t h a t th e y should be c o n sid ere d p a r t i a l
i n t e r p r e t a t i o n s of t h e o r e t i c a l co n c ep ts f o r
which o th e r p a r t i a l i n t e r p r e t a t i o n s might be
36
added under c e r t a in c o n d itio n s.
3. The logic of p ro p o s itio n s . I t was decided to
c o n stru c t the theory f i r s t in the language of
speech, then in mathematical eq u a tio n s. Two
types of deductions were to be used: (l) through
d e f in itio n and (2) through t r a n s i t i v e r e l a t i o n
ship. The t r a n s i t i v e deduction re q u ired assump
tio n s of p e r fe c t lin e a r r e la tio n s h ip s f o r the
p ro p o s itio n s . This assumption req u ired an addi
tio n a l assumption to f i t em p irical data to the
theory: th a t any s ig n ific a n t em p irical c o r re la
tio n was to be in te rp r e te d as p e rfe c t c o r r e la
tio n in theory.
With the logic of theory c o n stru c tio n developed in
d e t a i l , i t i s fe a s ib le to proceed to c o n s tru c t a theory of
fiv e in te r a c tio n v a ria b le s fo r small groups.
CHAPTER I I I
A THEORY OF INTERACTION IN SMALL GROUPS
In tro d u c tio n
Having e s ta b li s h e d the lo g ic and s t r a t e g y to be used
i n concept form ation and p r o p o s itio n a l r e l a t i o n s , we a re p re
pared to c o n s tr u c t a th eo ry of i n t e r a c t i o n in sm all groups.
Beginning w ith common sense co ncepts and p r o p o s itio n s , se- .
le c te d e m p iric a l s tu d ie s by s o c i a l s c i e n t i s t s are review ed
in t h i s ch a p te r to a s s i s t in forming hypotheses about the
r e l a t i o n s h i p s between the v a r i a b l e s . The next ste p i n the
pro cess i s the development of a g e n e ra l th eo ry in the la n
guage of speech, w ith o u t too g r e a t concern fo r p r e c i s io n ,
and a p p lic a b le to groups g e n e ra lly . The co n cep ts of t h i s
r a t h e r in c lu s iv e th e o ry are then given s p e c i f i c e m p iric a l
re f e re n c e to a r e s t r i c t e d domain, and the in c lu s iv e p ro p o si
tio n s are i n t e r p r e t e d i n m athem atical form fo r t h i s domain.
This r e s t r i c t e d th eo ry i s then ready fo r a t e s t of i t s i s o
morphism to e m p iric a l d a ta .
An I n c lu s iv e Theory of I n t e r a c t i o n
S e le c tio n o f concepts
Five concepts were s e le c te d fo r i n c lu s io n i n the
37
38
theory of i n t e r a c t i o n to be p rese n ted i n t h i s study. They
were i n i t i a l l y s e le c te d because they a re of s o c ia l impor
tan ce, as w ell as because they have been stu d ied fre q u e n tly
by s o c io lo g is t s . Since they have been p re v io u sly stu d ie d ,
th e re i s a b a s is fo r c o n s tru c tio n of a th eo ry s ta t i n g t h e i r
i n t e r r e l a t i o n s h i p s , w ithout c o lle c tin g a d d itio n a l e m p iric a l
data as a s t a r t i n g p o in t. The fiv e concepts chosen a re ;
1. Consensus.
2. U nderstanding.
3. S ize .
4. Democracy.
5. Frequency of i n t e r a c t i o n .
In the c o n s tru c tio n of t h i s in c lu s iv e theory, th e re
w ill not be too g r e a t a concern with the r i g o r of d e f i n i
t i o n a l congruence among previous s tu d ie s using a sin g le word
as a name fo r a concept. I t would be absurd to m aintain
t h a t a l l s tu d ie s using the concept democracy use i t to r e f e r
to the same phenomenon, and th a t each measurement o p e ra tio n
of democracy should th e re fo re be assumed to be a p a r t i a l in
t e r p r e t a t i o n of a sin g le t h e o r e t i c a l concept. This i s obvi
ously im possible to determ ine i n some cases and f a ls e in
o th e r s . I t w i l l merely be assumed a t t h i s p o in t t h a t th e re
i s some agreement on the g en e ra l area of re fe re n c e of each
concept.
Common sense p ro p o sitio n s
The c o n s tru c tio n of the theory w ill begin with some
common sense notions about i n te r a c ti o n . The follow ing s ta te
ments seem to express reasonable and accep tab le id e as about
human r e la tio n s h i p s .
1. Frequency of in t e r a c t i o n and understanding:
The more time people spend to g e th er, the more
they w ill understand one another.
2. Understanding and consensus:
People can b e tte r understand o th e rs who agree
with them than o th e rs with whom they do not
agree.
3. Frequency of i n te r a c ti o n and consensus:
People u su a lly w ill spend more time to g e th e r i f
they agree with one another than i f they d is
agree w ith one an o th er.
4. Size of group and consensus:
The la rg e r number of people th ere are in a
group, the more disagreem ents th ere w ill be
among the id e as of the members.
5. Size of group and democracy:
The la rg e r a group i s , the more one person w ill
tend to dominate i t .
These fiv e statem ents embody the fiv e concepts to be used
in the theory, but only in common sense form.
40
E m pirical o b serv atio n of the
v a ria b le s bv s o c io lo g is ts
The importance of these common-sense p ro p o sitio n s in
the s tr u c tu re of i n te r a c ti o n has not escaped the a t t e n t i o n
of s o c io lo g is ts who, as was said e a r l i e r , u su a lly begin with
common-sense o b se rv a tio n s. Summarized below are the fin d
ings of se le c te d s tu d ie s of these v a r ia b le s .
Democracy and siz e of group. Bossard and B oll, in
a r e c e n t d e s c r ip tiv e study of large fa m ilie s , re p o rte d th a t
larg e fa m ilie s are co n sid erab ly le s s dem ocratic in s tr u c tu re
than small fa m ilie s .^ - They rep o rted th a t a u t h o r it a r ia n fa
th e rs tend to predominate in fa m ilie s with many c h ild re n .
They even suggest th a t i t i s the nature of la rg e fa m ilie s
th a t to be e f f e c tiv e they must be a u t h o r it a r ia n , to organize
and d i s t r i b u t e ta sk s and give s t a b i l i t y and order to the
fam ily u n it.
On an e n t i r e l y d i f f e r e n t level. Bales rep o rted from
o b serv atio n of jad hoc la b o ra to ry groups of c o lleg e stu d e n ts:
. . . as groups get la rg e r , more and more communica
tio n i s addressed to one member of the group. . . . At
the same time the r e c i p i e n t begins to d i r e c t more and
more of h is remarks to the group as a w hole.3
^■James H. Bossard, The Large Family System (P h ila
d e lp h ia : U n iv e rsity of Pennsylvania P re ss, 1956).
2I b i d . . p. 312.
^Robert F. Bales, "The E quilibrium Problem in Small
G roups," Small Groups. ed.,A . Paul Hare, Edgar F. B orgatta,
and R obert F. Bales (New York: A lfred A. Knopf, 1955), p.
439.
41
This would seem lik e a reasonable approxim ation of the au
t h o r i t a r i a n s tr u c tu r e , or a t l e a s t a s i t u a t i o n favorable to
the development of an a u t h o r it a r ia n s tr u c t u r e . I t i s fu r-
A ^
th e r enforced by the fin d in g s of Bavelas^ and L e a v itt, who
rep o rte d th a t in communication networks of varying p a t te r n s ,
persons upon whom communication converges tend to be the
most i n f l u e n t i a l in problem so lv in g . Hamblin, using labora
to ry groups, r e p lic a te d the Bales study, and re p o rte d the.
same fin d in g s as B ales.^
Consensus and Frequency of i n t e r a c t i o n . One of the
most p e r s i s t e n t ideas in the l i t e r a t u r e on in t e r a c t i o n is
t h a t the more two people i n t e r a c t , the more a l ik e t h e i r
id eas become. In The Human Group. Homans, summarizing in a
common framework the Hawthorne s tu d ie s , S t r e e t Corner Soci
e t y . an a n th ro p o lo g ic a l monograph on a n o n lit e r a te group,
and a community study in New Hampshire, re p o rte d th a t these
fo u r stu d ie s show t h a t "the more fre q u e n tly persons i n t e r a c t
4Alex Bavelas, "Communication P a tte r n s in Task-
O riented Groups," Group Dynamics; Research and Theory, ed.
D. C artw rig h t and A. F. Zander (Evanston: Row, P eterson and
Company, 1953).
^Harold J . L e a v itt, "Some E ff e c ts of C e rta in Commu
n ic a tio n Problems on Group Perform ance," Readings in S o c ia l
Psychology, ed. Guy E. Swanson, T. M. Newcomb, and E. L.
H artley (New York: Henry H olt and Company, 1952).
^R. Hamblin, "An Experim ental Study of the Rela
tio n s h ip of Communication, Power R e la tio n s h ip s , S p e c ia liz a
tio n , and S o c ia l Atmosphere to Group S ize" (unpublished Ph.
D. d i s s e r t a t i o n , U n iv e rsity of Michigan, 1955).
I
42
with one another, the more a lik e . . . both th e ir a c t i v i t i e s
and th e ir sentim ents become,' and th a t
A decrease in the frequency of in te ra c tio n between
the members of a group and in the number of a c t i v i t i e s
they p a r tic ip a te in together e n t a il s a decline in the
e x te n t to which norms are common and c l e a r . 8
Katz and Lazarsfeld observed the same development of norm
consensus as a product of frequent in te ra c tio n :
. . . small group norms are manufactured in the
process of in te ra c tio n in r e la tio n to sp ecific s itu a
tio n s . In dividuals in te ra c tin g together sim ultaneously
c re a te a shared way of looking a t things or of doing
them.9
Two in v e s tig a to rs have conducted experim ental re
search to t e s t th i s re la tio n s h ip . Newcomb concluded from
studying college boys liv in g in a house together th a t
"Communicators tend to become more sim ilar to each other
than they were before the communication . . . " p a r tic u la r ly
with referen ce to o b jects important and re le v a n t to th e m .^
Hare reported th a t in laboratory groups he studied, the
lower the frequency of p a r tic ip a tio n in his groups, the
lower the consensus among i t s members.
^George C. Homans, The Human Group (New York: Harper
and B rothers, 1950), p. 120.
8I b id . . p. 362.
^Elihu Katz and Paul F. L azarsfeld, Personal In flu
ence (Glencoe: Free P re ss, 1955), p. 57.
^Theodore M. Newcomb, "The P re d ic tio n of In te rp e r
sonal A ttra c tio n ," American P sy ch o lo g ist. XI (November,
1956), p. 578.
^-A. Paul Hare, "A Study of In te ra c tio n and Consen
sus in D iffe re n t Sized GrouDS." American S o cio lo aical Review.
XVII (1952), p. 264. K 3-----
43
Consensus and size of group. In the study by Hare
rep o rted above, the author found th a t the la rg e r the group
was, the le s s consensus was achieved during i t s m eetings.
Edwin Thomas, studying la rg e r o rg a n iz a tio n s, concluded from
a study of w elfare o rg an iz atio n s th a t the la rg e r the organi
za tio n , the le s s consensus th ere was among members with re -
1 ^
gard to o rg a n iz a tio n a l r o le s .
On a more sweeping d e s c rip tiv e le v e l, the r e l a t i o n
of group size to consensus i s im p lic it in most of the d i-
chotomous c l a s s i f i c a t i o n s of s o c ie ty . Durkheim is q u ite
e x p l i c i t in s ta tin g th a t the organic s o li d a r ity of p re -in
d u s t r i a l s o c ie tie s is based on consensus, in c o n tra s t to the
lack of consensus with i t s concommitant complex d iv is io n
of labor in large s o c ie tie s . The same idea is contained
a lso in the G em einschaft-G esellschaft dichotomy, in Red-
f i e l d ’s folk-urban continuum, and with le s s e x p lic itn e s s in
B ecker's sa cred -sec u lar c l a s s i f i c a t i o n .
Size and frequency of i n t e r a c t i o n . In a study in
volving the system atic observation of a large number of
fa c e -to -fa c e groups, John James rep o rted the conclusion th a t
12 Ib id .
■^Edwin J . Thomas, "Role Conception and Organiza
t io n a l S iz e ," American S o cio lo g ical Review. XXIV (1959),
pp. 30-37.
14Emile Durkheim, The D ivision of Labor in S o c ie ty ,
tr a n s . George Simpson (Glencoe: Free P ress," 1947), pp.
133-38.
44
. . . Groups c h a r a c te r iz e d by f a c e - t o - f a c e and spon
taneous i n t e r a c t i o n tend to g r a v i t a t e to the s m a lle s t
s i z e , i . e . , two— to the p o in t of th e l e a s t number of
i n t e r a c t i o n s (one.) re q u ire d fo r i n t e r a c t i o n . . . . I t is
p o s s ib le t h a t groups over s iz e two m a n ife st in c re a s in g
i n s t a b i l i t y per u n i t in c re a se in s i z e . ^ 5
The c o n c lu sio n here i s t h a t the sm aller the group, the more
fre q u e n tly i t forms as a group, i . e . , has o p p o rtu n ity fo r
i n t e r a c t i o n . In the study by Hare quoted above, the i n v e s t i
g a to r r e p o rte d a lower i n t e r a c t i o n r a t e fo r p erso n s in la rg e
groups, a r i s i n g from the obvious s i t u a t i o n t h a t th e re a re
more i n t e r a c t o r s among whom to d iv id e a lim ite d amount of
time in the la r g e r than in the sm a lle r g r o u p s . ^ F is h e r , in
a study s im ila r to t h a t of James, but based on a more in te n
sive a n a ly s is of a sm aller number of groups, re p o r te d sub
s t a n t i a l l y the same fin d in g as James: "As the siz e of a
group in c r e a s e s , the frequency of i t s occurrence d e c re a se s
1 7
in a t y p i c a l decay c u rv e. nx' This i s a p r e c is e sta tem en t of
the in fe re n c e drawn from the James r e p o r t .
Democracy and frequency of i n t e r a c t i o n . D efining
the asym m etrical in flu e n c e r e l a t i o n between two people as
" o r ig in a tio n of i n t e r a c t i o n , " Homans concluded from the four
s tu d ie s which he reviewed t h a t
^ J o h n James, "A P re lim in ary Study of the Size De
te rm in a n ts in Small Group I n t e r a c t i o n , " American S o c io lo g i
c a l Review. XVI (August, 1951), 476-77,
^ H a r e , l o c . c i t .
^ P a u l N. F is h e r , "An A n aly sis of the Prim ary
G ro u p ," S ociom etrv. XVI (1953), 272-76.
45
When two persons i n t e r a c t w ith one a n o th e r, the more
fre q u e n tly one of the two o r i g i n a t e s in t e r a c t i o n fo r the
o th e r, the more n ea rly w ill the frequency of in t e r a c t i o n
be kept to the amount c h a r a c t e r i s t i c of the e x te rn a l
sy stem .18
I t may be concluded th a t i f the req u irem en ts of the
e x te rn a l system ( f a c t o r s not contained w ith in the group) are
held c o n s ta n t, the more asym m etrical a r e l a t i o n s h i p w ith
re s p e c t to in flu e n c e , the l e s s fre q u e n tly in t e r a c t i o n w ill
occur.
U nderstanding and frequency of i n t e r a c t i o n . The
concept understanding or empathy as i t is sometimes named,
i s a very w e ll-s tu d ie d concept in s o c ia l psychology. Stud
ie s by F i e d l e r , ^ C ronbach,20 G age,2-* H a s t o r f ,22 B i e r i , 2^
■^Homans, .op. c i t . . p. 247.
^ F r e d e . F ie d le r , " I n te rp e rs o n a l P erc e p tio n and
Group E f f e c tiv e n e s s ," Person P e rc e p tio n and In te rp e rs o n a l
B ehavior, ed. R e n a to .T a g u iri and L u ig i P e t r u l l o (Palo A lto:
S tan fo rd U n iv e rs ity P re s s , 1958).
20
Lee J . Cronbach, "Processes A ffe c tin g Scores on
'U nderstanding of O th e rs ' and 'Assumed S i m i l a r i t y ', " Psy
c h o lo g ic a l B u l l e t i n , L II (1955), 177-93.
2^N. L. Gage and Lee J . Cronbach, "Conceptual and
M ethodological Problems in In te rp e r s o n a l P e rc e p tio n ," Psy
c h o lo g ic a l Review. LXII (1955), 411-23.
22A. H. H a sto rf, I . E. Bender, and D. J . W eintraub,
"The In flu en ce of Response P a tte r n s on the 'R efined Empathy
§core, '" Jo u rn a l of Abnormal and S o c ia l P sychology. LI
(1955), 341-43.
2°J. B ie r i, "Changes in In te r p e r s o n a l P erc e p tio n
Following S o c ia l I n t e r a c t i o n ." J o u rn a l of Abnormal and So-
c i a l P sychology. XLVII (1953), 61-66.
46
Newcomb,and o th ers have used the concept in very sim ila r
ways. Newcomb and B ie ri re p o rte d s tu d ie s of the r e l a t i o n
ship between t h i s kind of " in te rp e r s o n a l p ercep tio n " and
the frequency of in te r a c ti o n of p erso n s. B ie ri worked w ith
stran g e r same-sex d ia d s, p re d ic tin g one a n o th e r 's r e a c tio n s
to the Rosensweig P ic tu r e - F r u s t r a tio n Test adapted fo r the
experim ent. With t h i r t e e n p a i r s in the experim ental group,
B ie ri re p o rte d th a t the su b je c ts were able to p r e d ic t one
a n o th e r's responses to the t e s t s i g n i f i c a n t l y b e t t e r a f t e r
only twenty minutes of inform al i n te r a c ti o n in p a ir s than
they could before i n te r a c ti o n . The d if fe re n c e in p r e d ic tiv e
power was sm all, but then the in te r a c ti o n time was very
sh o rt.
Newcomb, in h is o f t-r e p o r te d study of a 17-man hous
ing u n it, re p o rte d a f t e r a y e a r 's o b se rv a tio n of in te r a c ti o n
th a t "the accuracy of e s tim a te s of s i m i l a r i t y made by sub
j e c t s increased w ith time, p a r t i c u l a r l y a t the higher le v e ls
of i n t e r a c t i o n ."
There are o ther p a tte r n s of r e l a t i o n s h i p s between
the fiv e v a ria b le s which we have l i s t e d , and a number of
other s tu d ie s which impinge on the concepts to a g re a te r or
le s s e r degree, but those reviewed above give enough in d ic a
24Theodore M. Newcomb, "The C ognition of Persons as
C ognizers," T a g u iri and P e tr u llp , 0£. c i t .
25- .
J . B ie r i, c i t .
2^Newcomb, jop. c i t .. p. 186.
47
tio n of the e m p iric a l work which has been done in an attem p t
to v e r if y the common-sense r e l a t i o n s h i p s w ith which we be
gan. I t i s concluded t h a t the supposed p a t te r n of i n t e r r e
la t i o n s of the fiv e v a r ia b le s i s reaso n ab ly c l e a r , and th a t
many of them have a lre a d y been confirmed by more or le s s
p re c is e e m p iric a l s tu d ie s , as w ell as appearing in the theo
r e t i c a l fo rm u latio n s of many s o c io lo g is ts .
Nominal d e f i n i t i o n s of the
concepts
Although many r i g i d o p e r a t i o n a l i s t s scorn nominal
d e f i n i t i o n s as empty words, in the l i g h t of the a n a ly s is of
concept form ation above, i t i s the p o s itio n of t h i s d i s s e r
t a t i o n t h a t nominal d e f i n i t i o n s have a u se fu l fu n c tio n in
27
theory c o n s tru c tio n . I f the nominal d e f i n i t i o n of a con
c e p t i s p re se n te d in o b se rv a tio n term s, then t h i s d e f i n i
tio n w ill unambiguously d e sc rib e what i s to be observed, and
w ill be an o p e ra tio n a l d e f i n i t i o n . I f the nominal d e f i n i
t io n of a concept i s not p re se n ted in o b se rv a tio n term s,
then (1) e i t h e r i t i s a h ig h e r - le v e l concept c o n s tru c te d of
d efin e d term s, or (2) i t s d e f i n i t i o n s can be co n sid ered a
g e n e ra l statem ent of a program fo r p a r t i a l i n t e r p r e t a t i o n s .
The nominal d e f i n i t i o n s p re se n te d below are of the
27For a more g en e ra l trea tm en t of the r o le of nom
i n a l d e f i n i t i o n in th e o ry , see C arl G. Hempel, "Fundamentals
of Concept Formation in E m pirical S cie n ce ," I n te r n a ti o n a l
Encyclopedia of U n ified S cie n ce, Vol. I I : Foundations oF~the
U nity of S c ie n c e , No. 7 (Chicago: U n iv e rs ity of Chicago
P re s s , 1952).
48
l a t t e r type (2), and should be considered general statem ents
suggesting the area fo r more rig o ro u s p a r t i a l in t e r p r e t a
tio n . In a l l fu tu re d e f i n i t i o n s the conventional n o ta tio n
fo r nominal d e f in it io n w ill be adopted, where " . . . =df
. . " means " l e t . . . be synonymous in meaning w ith
I I
• • •
The nominal d e f i n i t i o n s of the in c lu s iv e theory
follow :
Group =df p l u r a l i t y of i n t e r a c t o r s .
(Let a group be defined as a p l u r a l i t y of i n t e r a c t o r s . )
Small =df having no more than 12 i n t e r a c t o r s .
(Small i s defined fo r t h i s study as meaning having no more
than 12 i n t e r a c t o r s .)
Consensus =df degree of agreement.
(Consensus as used in t h i s re se a rc h w ill mean the same as
degree of agreem ent.)
All fu tu re d e f in it io n s w ill appear w ith the conven
tio n a l " =df " n o ta tio n w ithout accompanying sentence def
i n i t i o n . )
The common-sense meaning of the statem ent "A under
stands B" i s something lik e "A knows how B th in k s , f e e l s ,
and can a n tic ip a te how he w ill r e a c t to s it u a ti o n s , e t c ."
For our purposes t h i s is r e s t r i c t e d to ;
Understanding =df a b i l i t y to p r e d ic t response to
s p e c ific s tim u li.
Size =df number of i n t e r a c t o r s .
49
Democratic =df c o n sistin g of in te ra c to rs with equal
influence on one another.
A and B are in te ra c to rs =df A and B are in one
a n o th e r's presence and are aware of one an o th er's
a c tio n s.
Obviously the d e f in itio n s presented above do not
e x p l ic i tly specify unambiguous r e fe r e n ts in a single case.
Even the concept of size does not allow the measurement of
the values of t h i s concept. Therefore the theory which is
b u ilt on the concepts w ill not be capable of em pirical t e s t
without fu rth e r d e f in itio n s . The concepts in th e ir present
form of nominal d e f in itio n s w ill be re fe r re d to as "in clu
sive concepts" and the theory b u ilt on them as the "inclu
sive th eo ry ." The more precise concepts tie d to measurement
operations to be presented l a t e r w ill be re fe rre d to as
" r e s tr ic te d concepts" and th e ir resp e ctiv e measurement oper
a tio n s as "operational d e f in itio n s ." The statem ent of r e l a
tio n s h ip s between the " r e s tr ic te d concepts" w ill be re fe rre d
)
to as the " r e s t r ic te d th eo ry ." The inclusive concepts only
suggest a general area of observation for th e ir p a r t i a l in
te r p r e ta tio n in the form of r e s t r i c t e d concepts.
P ro p o sitio n s of the inclusive
theory
Following the Z etterberg model developed in Chapter
I I , the p ro p o sitio n s derived from the research described
above are l i s t e d as follow s:
50
1. The g re a te r the consensus, the g re a te r the f r e
quency of in te r a c tio n .
2. The g re a te r the siz e , the le s s democratic the
group.
3. The g re a te r the siz e , the le s s the frequency of
in te ra c tio n .
4. The g re a te r the frequency of in te ra c tio n , the
g re a te r the understanding.
5. The g re a te r the siz e , the le s s the consensus.
Following the argument presented in Chapter I I above
which s t a t e s the assumptions necessary for deduction from
t h i s th eo ry , we include in our stru c tu re the assumption of
p e rfe c t lin e a r re la tio n s h ip between each p a ir of v a ria b le s .
From p ro p o sitio n 3, the g re a te r the s iz e , the le s s the f r e
quency of in te ra c tio n , the reverse can be s ta te d : the
sm aller the siz e , the g re a te r the frequency of in te ra c tio n ,
which is designated as 3a. I t i s then c le a r , following the
deductive model p resented, th a t statem ent 5 i s deducible
from 3a and 1.
3a The sm aller the siz e , the g re a te r the frequency
of in te ra c tio n (from 3).
1 The g re a te r the consensus, the g re a te r the f re
quency of in te ra c tio n .
5 The sm aller the s iz e , the g re a te r the consensus
(from 3a and 1).
This deduction is according to the t r a n s i t i v e r e l a t i o n a l
deduction explained on page 28. Accordingly, we designate
51
the f i r s t four statem en ts as axioms of the in c lu siv e th e
ory, and r e le g a te 5 to the theorems. The follow ing theorems
follow the axioms. Numbers a f t e r each theorem in d ic a te the
two p ro p o s itio n s from which i t was deriv ed .
1. The le s s dem ocratic a group, the le s s i t s f r e
quency of in te r a c ti o n (ax. 2, 3).
2. The g r e a te r the s iz e , the le s s the consensus
(ax. 1, 3).
3. The g r e a te r the consensus, the g re a te r the un
d e rsta n d in g (ax. 1, 4).
4. The sm aller the s iz e , the g r e a te r the under
standing (ax. 3, 4).
5. The g r e a te r the consensus, the more dem ocratic
the group (ax. 1, th . 1).
6. The g r e a te r the democracy, the g re a te r the un
d e rsta n d in g (ax, 2, th . 4).
C onfirm ation of the th e o ry . The theory as i t now
stands i s not confirm able as such. The nominal d e f in it io n s
of the concepts do not p o in t unambiguously to r e f e r e n t s , but
suggest general a re a s of o b se rv a tio n . To make the theory
t e s t a b l e , unambiguous r e f e r e n ts must be s p e c ifie d fo r the
concepts, i . e . , they must be o p e ra tio n a liz e d . But t h i s
o p e r a tio n a liz a tio n w ill be fo r a p a r t i c u l a r domain of phenom
ena, not as broad as the general concept, and alm ost cer
ta i n l y w ith l e s s g en eral ''meaning1 ' fo r the domain to which
52
i t gives re fere n c e . In a sense then, one cannot confirm
the in clu siv e theory above; one can only confirm or fa il to
confirm p ro p o sitio n s which are developed in observation terms
from the theory.
Summary of the inclusive
theory of in te ra c tio n
This p a rt of the chapter has been devoted to the
development of a theory of in te ra c tio n in small groups in
which the p ro p o sitio n s are presented in the language of
speech, and the concepts are nominally defined. Five con
cepts of so c ia l importance and freq u en tly studied by so c ia l
s c i e n t i s t s were se lec ted , and a set of common-sense proposi
tio n s was developed from them. A review was then presented
of stu d ies involving these v a ria b le s, in d ic a tin g the nature
of th e ir in te rr e la tio n s h ip s . On the b asis of these stu d ies,
a set of ten p ro p o sitio n s was developed, s ta tin g the in te r
re la tio n s h ip s among the v a ria b le s. Four of these were des
ignated axioms, and the other six, designated theorems, were
deduced from the four axioms. I t was pointed out th a t in
t h i s form the theory is s t i l l in a p recise sense not t e s t
ab le, since the nominal d e fin itio n s of the concepts do not
in d icate unequivocal r e f e r e n ts .
A R e stric te d Theory of In te ra c tio n
Introduction
In the previous section an in clu siv e theory of in-
53
te ra c tio n s was developed: inclusive because i t is applicable
to a wide range of groups, or a t le a s t does not specify pre
c is e ly i t s domain; inclusive because the meaning of the con
cepts of the theory is purposely l e f t vague. As was in d i
cated, the theory in t h i s form is not subject to em pirical
t e s t , because i t s th e o r e tic a l concepts are not defined in
observation terms the meanings of which are unambiguous.
This section is devoted to r e s t r i c t i n g the meaning
of the in clu siv e theory. The f i r s t p a rt of t h i s section is
devoted to giving precise reference to the r e s t r i c t e d con
cepts for the domain of same-sex ‘'n atu ral" adolescent
groups. The second p a rt of the section is devoted to a pre
sen tatio n of the axioms and theorems of the r e s t r i c t e d the
ory in the form of mathematical equations of lin e a r form.
In t h i s form, given the assumptions and logic developed in
Chapter I I , the r e s t r i c t e d theory is capable of being sub
je c te d to an em pirical t e s t .
Domain of the r e s t r i c t e d theory
The decision was made to in te r p r e t the inclusive
in te ra c tio n theory for "natural" same-sex adolescent groups.
Adolescent frien d sh ip groups were chosen for th e o re tic a l
reasons connected with assumptions as to the nature of the
v a ria b le s . I t was considered d e sira b le to work with fre ely -
forming groups whose members might come and go as they
pleased, asso cia te with one another as much or as l i t t l e as
they chose, and s tr u c tu r e t h e i r r e l a t i o n s h i p s w ith a minimum
of c u l t u r a l p r e s c r ip t io n e x te r n a l to the group. Same-sex
a d o le sc e n t frie n d s h ip groups seemed to meet th ese c h a r a c te r
i s t i c s b e t t e r than any o th e r type of a v a ila b le group. The
concepts of the th eo ry are defined below by measurement op
e r a t i o n s fo r t h i s domain.
The domain of the r e s t r i c t e d theory i s " n a tu r a l”
same-sex ad o lesc en t groups.
Group = df p l u r a l i t y of i n t e r a c t o r s .
A and B are i n t e r a c t o r s =df A and B are in one
a n o t h e r 's presence and aware of one a n o t h e r 's
a c tio n s .
A dolescent =df of age 13 through 19 y e a rs.
N atural = df having in te r a c te d before e n te rin g the
study.
Measurement of the v a r ia b le s
The measurement o p e ra tio n s were designed to be ad
m in iste re d to groups one a t a time w ith in te r-a c to rs assembled
in a room.
Measurement of c o n sen su s. Consensus has been meas
ured by a number of d i f f e r e n t te c h n iq u e s. F ie d le r used the
degree of congruence between resp o n se s on the Osgood Seman-
28
t i c D i f f e r e n t i a l . ° Hare had h is s u b je c ts rank o rd er the
^® Fiedler, .op. c i t .
55
importance of equipment for a c e r ta in ta sk , and used the
degree of s im ila rity of the rankings as a measure of consen
s u s . ^ Newcomb used s im ila rity of responses on a m u ltip le-
choice q u e s t i o n n a i r e .^ The technique adopted fo r the p res
en t study i s sim ila r to th a t of Newcomb. Responses on a
m ultiple choice questionnaire allow one to measure two other
v a ria b le s, understanding and democracy. These w ill be de
scribed below.
A m ultiple-choice q u estio n n aire co n tain in g 36 items
was co n stru cted on behavioral m atters of i n t e r e s t and impor
tance to ad o lesc en ts. Relevance was judged by the rese a rc h
er in accordance with several y e a rs ' teaching of adoles
ce n ts, to g eth er with a c a re fu l p eru sa l of p e r io d ic a l mater
i a l d ire c te d a t t h i s age group. A prelim inary form of the
questionnaire was adm inistered to two so c ia l science c la sse s,
including about 65 p u p ils , in a lo c a l high school for p re
te s tin g . Wording and design of the q u estio n n aire were then
rev ised to improve c l a r i t y and make marking e a s ie r . The
m ultiple choices were d e lib e r a te ly a lte r e d where the pre
t e s t in d ic ated very high consensus in the p r e - t e s t group.
This assured a broader spread on consensus scores in the
study, as w ell as enough disagreem ents even in high consen-
29
Hare, ojg. c i t .
3^Theodore M. Newcomb, "The Study of Consensus,"
Sociology Today: Problems and P ro s p e c ts , ed. Robert K. Mer
ton, Leonard Broom, and Leonard S. C o t t r e l l , J r . (New York;
Basic Books, 1959).
56
sus groups to allow measurement of r e l a t i v e in flu e n c e , as
d e sc rib ed below. (A copy of the q u e s tio n n a ire used is in
cluded in Appendix B .)
Assembled i n t e r a c t o r s were asked to answer the ques
tio n s in d iv id u a lly , and noncommunication between p a r t i c i
p a n ts was assu red by the r e s e a rc h e r during the response p er
iod. Consensus was then measured by counting the number of
q u e s tio n s on which the group agreed. H enceforth, consensus
w ill be r e f e r r e d to as C.
C =df number of q u e s tio n s on which a l l i n t e r a c t o r s
in a given group marked i d e n t i c a l resp o n ses to
the q u e stio n .
Measurement of u n d e rsta n d in g . A number of d i f f e r e n t
in stru m e n ts have been used in the measurement of understand
ing or empathy, but e s s e n t i a l l y the same o p era tio n i s p er
formed on the responses in each c a se . In d iv id u a ls A and B
are asked to respond to a l i s t of q u e stio n s and then each i s
asked to " p re d ic t" how the o th e r would respond to the given
q u e s tio n s. The number of c o r r e c t p r e d ic tio n s i s then used
as the measure of u n d erstan d in g . F ie d le r used the Osgood
Semantic D i f f e r e n t i a l as an i n s t r u m e n t i ^ while B ie r i, in
the study re p o rte d above, used p i c t u r e s from a p r o je c tiv e
t e s t . ^ Dymond used the Minnesota M ultiphasic P e r s o n a lity
— —
F ie d le r, c i t .
^ B i e r i , ,£[£. c i t .
57
In v e n to ry ,33 and C o rs in i used a sta n d a rd iz e d a d je c tiv e s o r t
l i s t . 34, Locke, Thornes, and Sabagh used s i t u a t i o n a l ques
tio n s w ith scale resp o n ses on the lik e lih o o d of the sin g le
35
response provided to the s i t u a t i o n .
Cronbach has c r i t i c i z e d empathy measures based on
the o p e ra tio n s l i s t e d above on s e v e ra l grounds, but most
im portant, t h a t (1) they are " d i s t r e s s i n g l y u n r e l i a b l e ," and
(2) the items to be p re d ic te d are seldom homogeneous as to
c o n te n t. He r e p o r ts th a t c e n tro id f a c to r a n a ly s is of r e s
ponses to such measurements r e v e a l t h a t the empathy score
can be broken down in to a t l e a s t four f a c t o r s . He suggests
t h a t a n a ly s is of such d ata should be made to determ ine
whether th e se four f a c t o r s should be combined as a sin g le
36
concept, or tr e a te d s e p a ra te ly as p r e d ic tiv e f a c t o r s . No
doubt some of C ronbach's c r i t i c i s m s apply in p a r t to the
instrum ent used in t h i s study. F u rth e r a n a ly s is of empathy
scores was not undertaken a t t h i s p o in t because the prim ary
33R osalind Dymond, " In te r p e rs o n a l P erc ep tio n and
M a rita l H appiness," Canadian J o u rn a l o f Psychology. V III
(1954), 164-71.
34Ravmond J . C o r s in i, "Understanding and S im il a r ity
in M arriage," Jo u rn a l of Abnormal and S o c ia l Psychology.
L II (1956), 327-32.
35
Harvey J . Locke, Georges Sabagh, and Mary M argaret
Thornes, " C o rre la te s of Prim ary Communication and Empathy,"
Research S tu d ie s of the S ta te C ollege of Washington. XXIV,
No.' 2 (1956), 116-24.
3^Lee J . Cronbach, £ £ . c i t . : and "P roposals Leading
to A nalytic Treatment of S o c ia l P e rc e p tio n S co re s," T a g u iri
and P e t r u l l o , oj£. c i t . . pp. 357-58.
58
concern is not p u r if i c a t io n of t h i s p a r tic u la r concept.
For t h i s study, understanding was measured on the
consensus qu estio n n aire in the follow ing manner. A fter com
p le tio n of the measurement of C, the q u estio n n aire s were col
le c te d , and id e n tic a l new ones passed o u t. S ubjects were
in s tru c te d as follow s:
(For two-person groups) "Think over each of these
questions c a re fu lly , and answer them t h i s time the way you
think the other person would answer them."
(For groups la rg e r than two) "Now. w ithout saying
anything or looking around, s e le c t one of the other persons
here and w rite his id e n t i f i c a t i o n number (to be explained
la t e r ) in the upper l e f t corner of the q u e stio n n a ire . Think
over each of these questions c a re fu lly , and answer them t h i s
time the way you think the person whose number you have
w ritte n would answer them."
The o p e ra tio n a liz a tio n of understanding, h e re a fte r
re f e r re d to as E, was scored as follow s:
e =df the number of questions for which A 's p red ic
tio n for B was id e n tic a l with B 's answer on the
consensus q u estio n n a ire .
E =df mean e fo r the group.
R e s tr ic tio n of the meaning of dem ocratic. The term
democratic has been used in so many d i f f e r e n t ways, many of
them vague, t h a t i t is d i f f i c u l t to decide on what p a r t of
59
the meaning to use in theory c o n s tru c tio n . The famous stud
ie s by Lewin and L ip p e tt of twenty y ea rs ago used a subjec
tiv e e v a lu a tio n of the group atmosphere based on the behav
io r of the le a d er of the g ro u p .3^ Most subsequent s tu d ie s
have focused on the behavior of the le a d e r. A few have
chosen the dimension of in flu en ce d i s t r i b u t i o n . S tu d ies by
M ills 38 and S tro d tb e c k 3^ are c h a r a c t e r i s t i c of re s e a rc h on
t h i s dimension.
The in flu e n c e dimension of democracy was s e le c te d ,
since i t seemed to be the more basic p a r t of the meaning of
democracy, and since i t promised to lend i t s e l f to o b je c tiv e
i n t e r p r e t a t i o n . The follow ing r e s t r i c t e d nominal d e f i n i t i o n
is s ti p u la t e d :
Group X i s dem ocratic =df a l l the members of group
X are eq u a lly i n f l u e n t i a l in making group de
c i s i o n s .
Group X i s not dem ocratic =df members of group X
vary w idely in t h e i r in flu en c e in making group
d e c is io n s .
The measurement problem then becomes how to measure the
37Kurt Lewin and Ronald L i p p i t t , "An Experim ental
Approach to the Study of Autocracy and Democracy: A P re lim
inary Note," Sociom etrv. I (1938), 293-300.
38Theodore M. M ills , "Power R e la tio n s in Three-Person
G roups," American S o c io lo g ic a l Review. XVIII (1953), 351-57.
3^Fred L. S tro d tb eck , "The Family as a Three-Person
Group," American S o c io lo g ic a l Review. XIX (1954), 23-29.
60
r e l a t i v e influence of members of the group, and then to con
s tr u c t a measure of d isp e rs io n or co n cen tratio n of t h i s in
fluence in such a way th a t d if f e r e n t groups w ill be com
parable .
Measurement of in flu e n c e . The influence dimension
of in te r a c tio n was measured by a m odification of a technique
o r ig in a lly described by Fred S tro d tb e c k ,4^ which he c a lle d
the "revealed d iffe re n ce technique." His technique is as
follow s. Each in te ra c to r i s asked in d iv id u a lly to respond
to a stim ulus instrum ent p r iv a te ly , a l i s t of s itu a tio n a l
questions w ith dichotomous responses. The group is then
asked to compare t h e i r responses and reach agreement as to
the b est answer for each question. Strodtbeck measured in
d iv id u a l influence in the group by a weighting process
p a r t i c u l a r l y adapted to the th ree -p erso n groups with which
he worked.
The p rese n t measurement technique, although based on
sim ila r o p eratio n s, was d if f e r e n t in several re s p e c ts .
F i r s t , the responses were m ultiple choice, which produced a
g re a te r spread of responses than would have been po ssib le
w ith dichotomous responses. Second, the technique needed to
be g e n e ra liz a b le for groups of varying s iz e . Third, the
th e o r e tic a l use of the concept suggested the a d v is a b ility of
a m etric w ith a t le a s t an in te r v a l sc a le . The basic d -sca le
4 0 Ib id
d escrib ed below has a tru e zero, where S tro d tb e c k 's zero
p o in t was a r b i t r a r y . The d -s c a le has equal in te r v a ls based
on a simple, counting o p e ra tio n , while S tro d tb e c k 's involved
a r b i t r a r y score w eights. I t should be pointed out th a t
S tro d tb e c k 's measure i s probably somewhat more s e n s itiv e
than the d -s c a le in one r e s p e c t: since there were only two
answers from which to choose on h is , h is groups were not
given the opportunity to compromise on a n e u tra l answer
which none had o r ig in a l ly chosen.
The measurement of influence fo r t h i s study was
achieved in the follow ing way. A fter the G -questionnaire
and E -q u estio n n aire were completed, the C -q u e stio n n a ire s
were re tu rn e d to the s u b je c ts , with one a d d itio n a l unused
q u e s tio n n a ire . The s u b je c ts were then in s tru c te d as f o l
lows: "Look over your answer to g e th e r. On any questions on
which you a l l ag ree, mark the answer on which you a l l agree
on the e x tra form. On the qu estio n s on which you do not
agree, ta lk over the q u estio n s u n t i l you can agree on which
is the b e st answer. Mark t h a t answer on the e x tra form. If
a f t e r fiv e minutes or so of d is c u s s io n on a single question
you should fin d t h a t you simply cannot agree, mark th a t
q uestion 'c a n 't a g re e ' and go on. Try to agree on a l l of
them, but do not have more than a couple on which you c a n 't
a g re e ." The re s e a rc h e r then l e f t the room, but stayed w ith
in e a rs h o t. The r e s u l t s were scored in the follow ing way:
62
d^ =df .the number of questions on which in d iv id u al
i ' s answer on the consensus qu estio n n aire was
d if f e r e n t from the answer on the group-answer
form.
Persons who were most in f l u e n t i a l had the lowest d^ scores,
since the group 's answers would correspond most c lo s e ly to
th e i r own. The s t a t i s t i c dr , which can be considered a
" re la tiv e influence score of in d iv id u al compared to h is
group" was computed by the follow ing formula:
where is the mean of d^ fo r the group. Persons with the
lowest d^ scores had the h ig h e st dr sco res, and hence the
highest r e la tiv e influence compared to t h e i r group.
c lo se ly corresponds to a scale for the measurement of the
le v e l of the in c lu siv e concept democracy fo r the group, was
computed by the formula:
N
where N is the size of the group. D is a measure of the
average d ev ia tio n of dj[ from the mean d^ fo r the group. The
lower the D score fo r a group, the more eq u ally i t s members
influence the group's d e c is io n s . Use of the mean d^ as
p o in t of base makes groups with low consensus scores and
dr =df 1 +
The s t a t i s t i c D, which can be considered a " re la tiv e
influence d isp e rsio n score fo r the group" and which most
63
those with high consensus scores comparable on the D s c a le .
Measurement of size of the group. Measurement of
the size of the group p resen ted some p a r t i c u l a r l y p erp lex in g
t h e o r e tic a l problems in the case of the same-sex a d o lesc en t
groups which were the su b je c t of the em p irical study. This
is what has custom arily been c a lle d the "boundary problem"
in s o c io lo g ic a l th e o ry . In formal groups such as c lu b s,
membership is u su a lly defined in the o r g a n iz a tio n s ' r u l e s .
In fa m ilie s , boundaries of the fam ily membership can be
e s ta b lis h e d by e a sy -to -a p p ly r u l e s . In the ad o lescen t
frie n d sh ip groups, the only c r i t e r i o n of membership in a
group is the f a c t of i n te r a c ti o n .
I f these groups had been a t a l l s im ila r to the groups
which H ollingshead d escrib ed in Elmtown, the problem would
have been r e l a t i v e l y sim ple. From h is d e s c r ip tio n s , the
read e r i s led to believe th a t nearly every a d o lesc en t "be
longed" to a c e r t a i n "bunch" with which he r e g u la r ly a s s o c i
a te d . Each was i d e n tif ie d as a member of a c e r t a i n group by
nonmembers and a d u lts . Each, i f asked what "bunch" he ran
around w ith, would a u to m a tic a lly name the members of h is
crowd and no o th e rs . Each f e l t a sense of membership in a
recognizable group.
For reasons which a t t h i s p o in t can only be sur
mised, H o llin g sh ead 's d e s c r ip tio n did not f i t the groups in
^ A u g u s t B. H ollingshead, Elmtown's Youth (New York:
Wiley and Sons, 1949).
the p o p u la tio n from which t h i s sample was drawn. From four
years of c a s u a l o b se rv a tio n of the stu d e n t body of the
school which was stu d ie d , the f l u i d i t y of the ad o lesc en t
group s tr u c tu r e was ap p a re n t. To be su re, th e re were recog
n izable c liq u e s which f i t t e d H ollingshead*s d e s c r ip tio n , but
they were an e x c e p tio n a l p a t te r n of a s s o c ia tio n . In the
process of in terv iew in g the sample chosen fo r t h i s study, i t
immediately became c l e a r th a t group boundaries were extreme
ly vague in the minds of the a d o le sc e n ts them selves. One
th in g which com plicated the p ic tu re was m u ltip le membership
in groups. For example, some of the time a boy would i n t e r
a c t w ith two or th re e o th e r boys, while a t o th e r tim es he
in te r a c te d w ith another boy, and a t s t i l l o th e r tim es he in
te r a c te d w ith s t i l l o th e rs . When the s u b je c ts were asked to
name those in the group they in te r a c te d w ith r e g u la r ly and
"considered t h e i r b e s t f r ie n d s ," most were in a quandary as
to whom to l i s t . I f they had a b e st f rie n d , t h i s would be
l i s t e d . The number of o th e rs l i s t e d depended upon memory of
names and o th e r in s c ru ta b le f a c t o r s .
One of the most obvious reaso n s fo r the lack of
d e f i n i t i v e boundaries in the groups is the f a c t th a t more
than h a l f of the stu d e n ts in the sample had moved to t h e i r
p re se n t homes in the p a s t two y e a rs , and more than th r e e -
fo u rth s in the p a s t four y e a rs . Few long-term frie n d s h ip
p a t te r n s had been e s ta b li s h e d .
I t was noted above t h a t the term group was defined
65
as a p l u r a l i t y of in te ra c to r s with no o th er q u alify in g
c h a r a c t e r i s t i c s . Consider the follow ing h y p o th e tic a l exam
p le , based on many concrete o b serv atio n s. John and B ill are
best frie n d s and spend f i f t e e n hours a week to g eth er on the
average. Bob, B i l l , and John spend another ten hours a week
to g eth er in ad d itio n to t h a t . Bob's b est frie n d is Jack,
with whom he spends about twenty hours a week. Once in a
while, say th ree hours a week, Bob, John, B i l l , and Jack a l l
do something to g e th e r. Bob and Jack a lso have other frie n d s
with whom they spend considerable time, and so on. Now pre
sumably each of these many combinations has behavior p a t
te rn s which are d i s t i n c t in some way from those of any other
p a tte rn of other com binations. In order to analyze the d if
ferences and hence is o la te the v a ria b le s which d i f f e r e n t i a t e
the groups, each grouping needs to be considered a group.
Any other c r i t e r i o n th re a te n s to elim in ate ranges of the
phenomena we wish to study, with no very c le a r reason for
the e lim in a tio n . For t h i s study, any p l u r a l i t y of in te ra c
to r s i s a group when i t is in te r a c tin g . Even i f two strang
e rs c o llid e on a corner, excuse them selves, and never meet
again, they c o n s titu te d a group during the time in which
they were in te r a c tin g .
As a consequence of these c o n s id e ra tio n s, any per
sons who came to g e th er w ith the person on the sample who was
interview ed were accepted as a group. This included cases
where one of the persons in the group had met the o th ers
only sh o rtly before the conduct of the research, as well as
cases where the subjects were well acquainted. This re
sulted in a wide range of v a ria tio n on frequency of in te r
a c tio n (to be described below), as w ell as considerable
range on the other v a ria b le s . Size of group, designated
h e re a fte r as N, was th e re fo re measured as follows:
N =df number of in te ra c to r s , excluding the research
e r , present in the measurement s itu a tio n .
Measurement of frequency of in te r a c ti o n . The idea
frequency of in te ra c tio n was o p eratio n alized in three d if
fe re n t ways. On an inform ation sheet, (See Appendix A),
su b jects were asked to indicate for each other in te ra c to r
(1) the number of hours spent together during the preceding
week, (2) the average number of hours spent together during
a school week, and (3) the number of years each had known
the o th e r. The group was then asked how much time the whole
group spent together during an average week (fo r groups of
more than two only). Each person made the computations in
co n su lta tio n with the other in every case, usually on a day-
by-day b a s is, beginning with the previous day and working
backwards for the f i r s t computation.
I = d f the number of hours individual i said he
spent with j the preceding week.
I a =df
67
J i j xik * • • ■
N
1 a + + • • •
I =df --------------------------
N
I was chosen as the measure of frequency of i n te r a c ti o n in a
given group from in c lu sio n in the r e s t r i c t e d th e o ry . The
follow ing measures were a lso included as an ex p lo ra to ry
ste p .
F^j =df average number of hours i said he spent w ith
j during a school week.
p . = d f F v * ' • . •
N
F j + F j + • • .
F =df ----------- *-------------
N
F corresponds c lo s e ly to I, except I p e r ta in s to the sp e c if
ic week before p a r t i c i p a t i o n in the study, while F is fo r an
average school week.
=df number of years i said he had known j .
L • • + Lji. + . . .
L. =df ------- —--------------
- _ N
Li L• + . . «
L =df — ------- J-------------
N
T =df group agreement of the average number of hours
the whole group spent to g e th e r in a week.
Nature of the measurement s c a l e s . A ll the basic
measurements are based on simply counting u n its : hours,
68
agreements, c o rre c t p re d ic tio n s , e t c . The measurements were
in te n tio n a lly co n stru cted on the counting of eq u iv alen t
u n its , each with a tru e zero p o in t, fo r two reaso n s. F i r s t ,
t h i s allow s the use of the Pearsonian c o r r e la tio n c o e f f i
c i e n t. This s t a t i s t i c has av a ila b le formulae fo r p a r t i a l
and m ultiple c o r r e la tio n which can a s s i s t in making c le a r
the stru c tu re of the in te r r e l a ti o n s h ip s of the theory.
Second, such sc a le s allow the c o n stru c tio n of other con
cep ts from the basic sc a le s by m u ltip lic a tio n and d iv is io n ,
giving considerably more f l e x i b i l i t y to the form ulation.
Ordered or p a r t i a l l y ordered scale s could have been con
stru c te d for most of the concepts, but such sc ale s lack the
property of having a v a ila b le m ultiple and p a r t i a l c o r r e la
tio n s t a t i s t i c s , and lack the property of being m u ltip lic -
able and d i v i s i b l e .
Correspondence of the r e s t r i c t e d to in c lu siv e con
c e p t. The question w ill c e r ta in ly be r a is e d as to the "in
te r n a l v a lid ity " of the o p e ra tio n a l concepts: to what ex ten t
do they correspond to the in clu siv e concepts from which they
were derived? F i r s t , i t i s c le a r th a t in every case the
o p e ra tio n a l concept is considerably more r e s t r i c t e d than the
in clu siv e one. For example, the C -scale p u rp o rts to measure
agreement on only 36 s p e c ific item s, while consensus i s nom
in a lly defined as "degree of agreement." To what e x te n t do
they correspond? I t i s c le a r th a t the C -scale i s more
r e s t r i c t e d in meaning, but the degree of t h e i r correspond
ence i s a su b je c tiv e judgment, since we cannot measure the
in c lu siv e concept. I f the re a d e r concludes th a t the D -scale
i s not what he means by democracy, or t h a t the E -scale is
not what he means by und erstan d in g , l e t him not connect the
r e s t r i c t e d th eo ry to the in c lu siv e th e o ry . No one can say
to what e x te n t any o b se rv a tio n s confirm or d isc o n firm the
in c lu siv e th e o ry , except by re fe re n c e to the r e s t r i c t e d
th e o ry .
P ro p o s itio n s of the r e s t r i c t e d
theory
The s c a le s of th e v a r ia b le s of t h i s theory have
been c o n s tru c te d such t h a t every value of every concept can
be expressed as a p o s itiv e r a t i o n a l number. Furtherm ore,
the t h e o r e t i c a l s tr u c tu re in c lu d e s an assumption of "per
f e c t" l i n e a r c o r r e l a t i o n between each of the v a r ia b le s . The
axioms of the r e s t r i c t e d th eo ry can now be s ta te d in term s
of l i n e a r e q u a tio n s .
Axioms.
1. C =
ar +
b i l
2. N =
a2 "
b2D
3. N =
a3 "
b3 I
4. I = a4 + b4E
Theorems deducible from the axiom s. From the axioms
can be developed by the r u l e s of elem entary alg eb ra the pro
p o s itio n s which correspond to the theorems of the in c lu siv e
70
theory. Note th a t the six theorems are d ir e c tly deducible
including the two c o e ff ic ie n ts in each equation. The coef
f i c i e n t s of the theorems have been reduced to su b sc rip ts
because i t was not proposed to have data of s u f fic ie n t pre
cisio n to j u s t i f y a t e s t of the c o e f f ic ie n ts . The theorem
d e riv a tio n from the axioms follows the p a tte rn below.
Axiom: Y = a + bX
Axiom: Z = c + dX
dX = Z - c
y — Z — c
bX = Y - a
Z - c _ Y - a
d “ b
Theorem: Y = a - ~ + -^(Z)
d d
A ctually, eaGh of the above equations might be w ritte n as a
leg itim ate theorem of the r e s t r i c t e d theory, but there is no
need fo r an exhaustive l i s t of the theorems a t the p resen t
tim e.
Theorems.
1. D = a& - b^I 4. N = ag + bgE
2. N = a6 - b6C 5. C = a 9 - b9D
3. C = 8 7 + byE 6 . D = a^Q -
I t i s possible to t e s t these axioms and theorems in
71
t h e i r p re s e n t form. That i s , from the d ata gathered in the
measurement s i t u a t i o n s p e c if ie d , one can s ta te the degree of
co n firm atio n of the p r o p o s itio n s in term s of p r o b a b ili ty
sta te m e n ts. Given the assumption of " p e rfe c t" l i n e a r i t y
whenever a c o e f f i c i e n t of c o r r e l a t i o n , however sm all, i s sig
n i f i c a n t l y d i f f e r e n t from zero, i t i s re q u ire d fo r confirm a
tio n of each of the p ro p o s itio n s only t h a t the c o r r e l a t i o n
c o e f f i c i e n t fo r each eq u a tio n be s i g n i f i c a n t l y d i f f e r e n t
from zero in the d i r e c tio n p r e d ic te d . The co n v en tio n al .05
le v e l of s ig n ific a n c e i s used in t e s t i n g the n u ll hypotheses.
Null h y p o th e se s. The n u ll hypotheses to be t e s t e d
are as fo llo w s: fo r each p ro p o s itio n of the th e o ry , the r i s
zero. D ire c tio n of the c o r r e la tio n i s p re d ic te d , th e re fo re
o n e - ta ile d t e s t s are in o rd e r.
Im p lic a tio n s of the r e s t r i c t e d
theory
When the axioms and theorems are in te r p r e te d in
term s of the r e s t r i c t e d concepts, they r e p re s e n t a s t a t i c
system, and d escrib e the s ta te of the system a t any one
tim e .4^ When the s t a t i c s are thoroughly understood, time
may then be introduced as an e x p l i c i t v a ria b le in the th e
ory. Such a dynamic system would probably re q u ire a theory
4^Von Neumann's advice has been follow ed h ere: "The
h is to r y of o th e r sc ien c es in d ic a te s t h a t th e re i s l i t t l e
hope of working out a dynamic system u n t i l the s t a t i c s are
thoroughly u n d ersto o d ." (John von Neumann and Oskar Morgen-
s te rn , "The B asis of Game Theory," Readings in Game Theory
and P o l i t i c a l Behavior 1 ed. M artin Shubik (Garden C ity :
72
in which the axioms were simultaneous d i f f e r e n t i a l equa
t io n s . At t h i s time in in te r a c tio n theory the se rio u s en
te rta in m e n t of a dynamic th e o r e tic a l scheme as a guide fo r
e x p lo ra tio n seems prem ature.
Meanwhile o th er problems seem more p re ssin g :
1. What v a ria b le s are to be included in a basic
system?
2. W ill a deductive scheme based on d e te rm in is tic
assumptions fu rth e r knowledge of the phenomena
involved, or is i t also premature?
3. Can a workable s t a t i c system be c o n s tru c te d
which w i l l lead to new d isc o v e rie s ?
4. What assumptions is i t necessary to make in or
der to make accurate p re d ic tio n s from such th e -
o rie s?
Summary of the r e s t r i c t e d theory
of in te r a c tio n
The f i r s t se ctio n of t h i s ch ap ter was devoted to
the development of an in c lu siv e theory of in te r a c ti o n in
small groups in which ten p ro p o s itio n s were p resen ted in the
language of speech, w ith nominally d efin ed concepts. Four
of the ten p ro p o s itio n s were designated axioms, and the
other s ix , d esig n ated theorems, were deduced from the ax i
oms, using the logic developed in the preceding se c tio n on
Doubleday, 1954), p. 60.)
73
the lo g ic of p ro p o s itio n s . I t was pointed out t h a t in t h i s
form the theory was s t i l l not te s ta b l e in any p rec ise sense,
since the nominal d e f i n i t i o n s of the concepts did not in d i
ca te unequivocal r e f e r e n t s .
The second se c tio n of the ch a p ter was devoted to
developing a more r e s t r i c t e d theory from the in c lu siv e th e
ory of the f i r s t s e c tio n . The concepts of the r e s t r i c t e d
theory were defined so as to r e f e r to the same g en e ra l area
of phenomena as the concepts of the in c lu siv e th e o ry , but
w ith unequivocal em p iric a l re fe re n c e , and consequently more
r e s t r i c t e d but more p re c is e meaning. The r e la t i o n s h i p s be
tween concepts expressed in the in c lu siv e theory were made
more p re c is e and le s s ambiguous fo r the p ro p o s itio n s of the
r e s t r i c t e d th e o ry . This was done by s ta tin g the r e l a t i o n
sh ip s between concepts of the r e s t r i c t e d theory in mathe
m a tic al n o ta tio n , in t h i s case in the form of sim ultaneous
lin e a r e q u a tio n s.
Having s ta te d the four axioms of the r e s t r i c t e d
th eo ry in the form of sim ultaneous lin e a r eq u atio n s, from
them i t was p o ssib le to deduce th e six theorems of the r e
s t r i c t e d th e o ry . The r e s u l t i n g se t of te n eq u atio n s approx
imated the ten p ro p o s itio n s of the in c lu siv e th e o ry , but
could not be considered e q u iv a le n t to those of the in c lu siv e
th e o ry . From th e se te n p ro p o s itio n s , te n s t a t i s t i c a l n u ll
hypotheses were ready fo r e m p iric a l t e s t . The n u ll hypothe
ses were th a t in each eq u atio n , the corresponding lin e a r
c o r r e la tio n c o e f f i c i e n t would not be s i g n i f i c a n t l y d i f f e r e n t
from zero in the d i r e c t i o n p re d ic te d , a t the .05 le v e l of
s ig n i f i c a n c e .
I t was p o in te d out t h a t the e m p iric a l t e s t , to be
re p o rte d in the next c h a p te r, i s a t e s t only of the r e
s t r i c t e d th e o ry . To what e x te n t the r e s u l t s of the em pirical
study can be said to apply to the in c lu s iv e th eo ry i s a mat
t e r of su b je c tiv e judgment.
CHAPTER IV
AN EMPIRICAL TEST OF THE RESTRICTED
THEORY OF INTERACTION
In tro d u ctio n
Most previous em p irical rese a rc h on v a ria b le s in
small-group in te r a c tio n has e i th e r been ex p lo rato ry and
d e s c rip tiv e (b est example: S tre e t Corner S o c ie ty 1), or i t
has been experim ental and conducted w ith ad hoc experim ental
groups (b est example: the work of the Human R e latio n s Labor
atory a t H arvard). Although much valuable work has been
done with these experim ental groups, a number of sociolo
g i s t s have questioned the g e n e r a liz a b ility of the r e s u l t s to
n a tu ra l human groups.
A small number of stu d ie s have appeared over the
l a s t few years using e x is tin g n a tu ra l groups for experiment
a l s tu d ie s . S tro d tb e c k 's c r o s s - c u l tu r a l study of marriage
2 - “ 5
p a ir s , and h is re se arc h w ith th ree-p e rso n family groups,
^-William F. Whyte, S tre e t Corner S ociety; The S ocial
S tru c tu re of an I t a l i a n Slum (Chicago: U n iv ersity ' o f Chicago
P re s s , 1943).
2
Fred L. Strodtbeck, "Husband-Wife I n te ra c tio n Over
Revealed D if fe re n c e s ,” American S o cio lo g ica l Review. XVI
(1951), 468-73. -------------------
^Fred L. S trodtbeck, "The Family as a Three Person
75
76
demonstrate the kind of precise experim ental work which can
be done with n a tu ra l groups. Of course the problem of gen
e ra liz in g from experim entally manipulated in te ra c tio n to
spontaneous n atu ra l in te ra c tio n is s t i l l th e re , but at
le a s t one obstacle has been removed.
The present study was conducted with n a tu ra lly oc
curring groups in an experim ental s e ttin g . An attempt was
made to secure a random sampling of groups from a s p e c ifi
c a lly r e s t r i c t e d universe. So fa r as can be determined from
published statem ents of previous re se a rc h e rs, t h i s i s the
only experim ental-type study of small-group in te ra c tio n
known to the author which was conducted with a random sample
of n a tu ra l groups.
The universe
The universe of t h i s study was the n atu ral same-sex
groups in the student body of a suburban Southern C a lifo rn ia
high school of about 2500 p u p ils . I t was selected because
of the f a c i l i t i e s and o p p o rtu n itie s fo r freedom of actio n
offered the rese arch er at t h i s school. Conference rooms for
the conduct of the research were provided on a 24-hour basis
for the e n tire period of the rese arch . Complete access was
granted to school records for a d d itio n a l inform ation on sub
j e c t s . The research er was provided with an enumeration of
the universe fo r se le c tio n of the sample, and names and
Group." American S ocio lo g ical Review. XIX (1954), 23-29.
77
addresses and phone numbers of a l l stu d e n ts to f a c i l i t a t e
c o n tact w ith su b je c ts during the summer, when most of the
re se a rc h was conducted. A d m in istrato rs of the school gave
th e ir endorsement of the re se a rc h to se v e ral worried p a re n ts
of su b je c ts who otherw ise would have been l o s t from the
sample •
The sample
This study re q u ire d a random sample of n a tu ra l
groups from the school p o p u la tio n . Since i t was impossible
to know of the ex isten ce of a l l the groups (under the d e fin
i t i o n of group adopted above), the d e c isio n was to take a
random sample of stu d e n ts, and ask each person se le c te d to
designate a group in which he wa$ an in te r a c to r fo r p a r t i c i
p atio n in the study. A ccordingly, from a l i s t of the stu
dent body a simple random sample of s ix t y - s i x stu d en ts was
s e le c te d . S ix ty -fo u r of th e se were interview ed during
school hours, before the close of the school year in June,
1959. Students were c a lle d from c l a s s e s , and sev eral a t a
time were presented w ith an e x p lan atio n of the study. I t
was re p re se n te d to them as a study "to help a d u lts b e t te r
understand what makes te en ag e rs t i c k . " Each person was
asked to o b ta in the cooperation of the group of h is f rie n d s
who ran around to g e th e r, whom he considered h is c lo s e s t
group of f r ie n d s . Most agreed to w rite down the names of
those persons whom they intended to ask . No person refu sed
78
to cooperate during the i n i t i a l p re se n ta tio n of the req u e st.
A ll were asked to come back the next day to re p o rt th a t they
had th e i r f r ie n d s ' cooperation. About h a lf retu rn ed un
prompted to give affirm ativ e r e p o r ts . The remainder were
again c a lle d out of c la s s for re p o rts on obtaining th e ir
f rie n d s ' cooperation. Of th e se , one refused to cooperate,
two in d icated th a t th e i r frie n d s were unw illing to do i t ,
one rep o rted th a t h is only frie n d was moving out of town
th a t very day. The remaining six ty agreed to p a r tic ip a t e .
Case accounting. Of the s ix ty -s ix names on the
o rig in a l sample, th ir ty - f o u r were boys and th irty -tw o were
g i r l s . Table I below in d ic a te s f in a l d is p o s itio n of the
o r ig in a l cases.
Examination of the school records for the six re
fu sa ls in boys ind.icated th a t they were as a group average
in I.Q ., and below average in school achievement. Three had
records as chronic d is c ip lin e problems in school. The
breakdown of f a t h e r s ' occupations of su b jects a c tu a lly in
cluded matched very clo sely in percentage d is tr ib u tio n a
previous random sample of 150 p u p ils from t h i s universe for
another study, in which there was no case a t t r i t i o n . I t
th e re fo re seems j u s t i f i e d to conclude th a t the sample for
t h i s study i s re p re se n ta tiv e of the universe as to f a t h e r s '
occupations.
The problem of randomness. The question of the
79
TABLE 1
DISPOSITION OF CASES OF
THE ORIGINAL SAMPLE
Boys G ir ls
Number o r ig in a lly drawn in the sam-
pie 34 3*
Unable to locate a t any
time 1
R efusals ( i n i t i a l or sub
sequent contact) 6 1
I n i t i a l co n tact but moved
before completed 2
I n i t i a l co n tact but unable
to locate l a t e r 2
Chose other member of
sample in group 3
14 _1 '
Total groups studied 20 31
80
randomness of the groups s e le c te d i s d i f f i c u l t to answer.
The sample of p erso n s s e le c te d o r i g i n a l l y was random. But
each p erso n i s an i n t e r a c t o r in a m u ltitu d e of groups. I t
i s d i f f i c u l t to assume t h a t each person s e le c te d the p erso n s
to p a r t i c i p a t e w ith him so t h a t the r e s u l t i n g groups were a
random sample of a l l groups in the u n iv e rs e . Furtherm ore,
some p erso n s on the sample named c e r t a i n in d iv id u a ls as
those in the group to p a r t i c i p a t e , l a t e r were unable to g e t
those p erso n s to come w ith them, and chose o th e r s . I t i s
in e sca p ab le under th e se c irc u m sta n ces t h a t s e le c tiv e f a c t o r s
were a t work to give some groups g r e a t e r chance to e n t e r the
sample th a n o th e r s . The most which can be sa id i s t h a t the
sample was as n e a rly random as p o s s ib le under the p r a c t i c a l
c irc u m sta n c e s.
I t has been s ta te d by Z e tte r b e rg 4 and Chapin5 t h a t
i t i s not necessary to have random samples on which to t e s t
h yp o th eses. Z e t t e r b e r g 's argument i s p re s e n te d below:
C onsider ag ain the u n i v e r s a l i t y of a th e o ry . The
r e l a t i o n s h i p s in the hypotheses we have to v e r if y , in
o th e r words, cla im to be u n iv e r s a lly p r e s e n t. They are
a c c o rd in g ly p re s e n t both in r e p r e s e n ta ti v e and non-rep
r e s e n t a t i v e sam ples. To disprove or dem onstrate t h e i r
e x is te n c e i s , hence, p o s s ib le in any kind of a sample,
b ia se d or u n b iased . . . . In c o n c lu sio n , i t ap p ears t h a t
n o n -re p re s e n ta tiv e samples are not much i n f e r i o r to
r e p r e s e n ta ti v e samples when we want to d isp ro v e a
4Hans L. Z e tt e r b e r g , On Theory and V e r i f i c a t i o n in
Sociology (New York: T r e s s le r P re s s , 1954), pp. 56-57.
5F. S tu a r t Chapin, E xperim ental Design in S o c ia l
Re search (New York: H arpers and B ro th e rs, 1947), p. 103".
31
hypothesis claim ing u n iv e rsa l v a l id ity .^
I t is tru e th a t a nonrepresentative sample can serve to d is
prove a s t r i c t l y u n iv e rsa l p ro p o sitio n . But i t is also tru e
th a t a single negative case equally disproves such a propo
s itio n .
I t i s the p o s itio n of t h i s d i s s e r ta ti o n th a t in any
em pirical in v e s tig a tio n , p ro b a b ility or random samples are
e s s e n t i a l . I t w ill be remembered th a t a p o stu la te of per
f e c t r e la tio n s h ip s has been tem porarily adopted in the
th e o r e tic a l system. Thus the theory co n tain s p ro p o sitio n s
of u n iv e rsa l form. On the other hand the hypotheses which
have been derived from the theory for em pirical t e s t w ill
c e r ta in ly not be r e je c te d should th e re be found cases in
which they do not hold, for i t was decided to account for
these cases in terms of v a r ia tio n of other v a ria b le s not in
cluded or accounted f o r, as w ell as in terms of e r r o r s of
measurement. Under what co n d itio n s are the hypotheses to be
accepted as confirmed by the data? Following the e a r l i e r
7
lead from G uilford, a hypothesis w ill be accepted as con
firmed i f the data show a r e la tio n s h ip of the so rt p red icted
by the hypotheses and of s t a t i s t i c a l sig n ific a n c e . C learly
s t a t i s t i c a l hypotheses are being te s te d . Although the th e
ory co n tain s u n iv e rsa l p ro p o sitio n s , the hypotheses w ill be
^Z etterb erg , _op. c i t . . pp. 56-57.
7p age 32 above.
82
accepted according to the r e s u l t s of a p p lic a tio n of prob
a b i l i t y s t a t i s t i c s to the d a ta .
In sociology i t i s safe to say t h a t one i s alm ost
never d i r e c t l y working on the e m p iric a l v e r i f i c a t i o n of a
s t r i c t l y u n iv e rs a l p ro p o s itio n . Furtherm ore, as Reichen-
bach p o in ts o u t, no e m p iric a l science works w ith anything
but p r o b a b ili ty statements® but whenever d ata i s assembled
to t e s t p r o b a b ili ty sta te m en ts, p r o b a b ili ty s t a t i s t i c s are
used in the comparison of the data w ith the p ro p o s itio n to
be confirm ed. These p r o b a b ili ty s t a t i s t i c s are based on the
assumption of p r o b a b ili ty samples. I f the sample from which
the data were derived i s not a p r o b a b ili ty sample, the s ta
t i s t i c s are in a p p lic a b le , and th e re i s no b a s is fo r s ta tin g
the p r o b a b ili ty le v e l of co n firm atio n . Hence i t i s con
cluded t h a t p r o b a b ili ty samples are e s s e n t i a l in the con
firm a tio n of hypotheses as w ell as in d e s c r ip tiv e s tu d ie s .
This i s the reason fo r the e f f o r t devoted to the attem pt to
procure a random sample fo r t h i s d i s s e r t a t i o n .
S o c ia l c h a r a c t e r i s t i c s of the sample. A t o t a l of
f ifty -o n e groups p a r t i c i p a t e d in the study. Twenty of th ese
were boys' groups, and th ir ty - o n e were groups of g i r l s .
Groups v a rie d in size from two to seven, w ith a mean size of
®Hans L. Reichenbach, " P ro b a b ility Methods in S o cia l
S c ie n c e s ," I n te r n a ti o n a l Encyclopedia of U n ifie d S c ie n c e .
Vol. I I : Foundations of the U nity of Science-! No'. 1 (Chicago:
U n iv e rs ity of Chicago P re s s , 1944), pp. 25-26.
33
th re e . A ltogether, 151 su b je cts p a r tic ip a te d in the study,
including 54 boys and 97 g i r l s . Subjects ranged in age from
14 to 19, with 90 per cen t between 15 and 17.
The boys and g i r l s in t h i s study were nearly a l l
(96 per cent) native born, but th e i r b irth p la c e s were scat
te re d in nearly every s t a t e . Only 6 per cent were born in
the county in which the study was conducted, and only 40 per
cent were born in Southern C a lif o rn ia . South c e n tr a l, north
c e n tr a l, and n o rth ea stern s ta te s accounted fo r nearly h a lf
the b ir th p la c e s . One fo u rth had liv e d a t t h e i r p rese n t ad-'
dresses a year or le s s , and th re e -fo u rth s had moved to th e i r
p rese n t homes w ithin the p ast four years.
According to f a t h e r s ’ occupations as rep o rted by the
s u b je c ts , about 20 per cent were c l a s s i f i e d as p ro fe s s io n a l,
about 13 per cent nonprofessional white c o l la r , and about
46 per cent s k ille d workers, and about 17 per cent semi
s k ille d and u n sk ille d workers. Four per cent were deceased
or u n c l a s s i f i a b l e . About 36 per cent of the subject r e
ported employed mothers. As for t h e i r own occupational as
p ir a tio n s , about 44 per cent said they expected to en te r
p ro fe ssio n s, while 12 per cen t chose s k ille d tra d e s and none
mentioned sem i-sk ille d or u n sk ille d jo b s.
According to re p o rted r e lig io u s preferen ce, 68 per
cent of the su b je cts were P ro te s ta n t, 17 per cent C ath o lic,
about 3 per cen t Jewish, and 11 per cent o ther responses.
More than h a lf (56 per cent) of the su b je cts re-
84
p orted being employed p a rt-tim e or f u ll-tim e a t the time of
the study ( summer v a c a tio n ), and 39 per cent re p o rte d having
been employed during the school y ea r.
The mean I.Q . of the su b je c ts as measured on the
O tis Gamma t e s t was 107, about two p o in ts higher than the
school average. S ubjects re p o rte d th a t they had known the
o th er members of t h e i r group an average of about two and a
h a lf y e a rs, but t h i s ranged from a few hours to lb y ears.
P r a c t i c a l Mechanics
Nearly a l l groups were assembled during summer vaca
tio n in the follow ing manner. The sample member was con
ta c te d by phone, or in person i f he had no phone, and asked
to make an appointment fo r meeting w ith the experim enter.
He was asked to c o n ta c t b is frie n d s and arrange fo r a meet
ing a t a s p e c ifie d time and p la c e . About tw o -th ird s of the
groups were met and measured a t the conference rooms in the
high school, and about o n e -th ird were met in the homes of
members of the group. T ra n sp o rta tio n was provided when nec
e ssa ry by the experim enter. Broken appointments were a con
s ta n t hazard, and the re s e a rc h e r was "stood up" se v e ral
time s.
Once the group had assembled, a b r ie f exp lan atio n of
the purposes of the p r o je c t was given, and each person was
asked to s e le c t an i d e n t i f i c a t i o n number from 1 to 9, p re
sented by the re s e a r c h e r . These were re g u la r p l a s t i c house
85
numbers hung on s tr in g s . The su b je cts were asked to put the
numbers around t h e i r necks, and to use numbers in stea d of
names fo r themselves and th e i r frie n d s on a l l papers of the
re se a rc h . Members were to ld th a t since th ere were no names
on the papers, they could say what they chose in answer to
the q u estio n s. Everyone seemed to accept t h i s as a guaran
te e of anonymity. In a c tu a l f a c t , the name of each person
had been a sce rta in ed before the meeting. Since the informa
tio n sheet which each person f i l l e d out contained h is number
and h is date of b ir th (although no name), i t was simple to
use school reco rd s to match number w ith name by means of
date of b i r t h . This was done in order to secure a d d itio n a l
data on each su b ject fo r other v a ria b le s , from the school
reco rd s.
Each person then f i l l e d out the f i r s t page of the
inform ation sheet (see Appendix A). After t h i s was com
p le te d (from 5 to 25 minutes depending on the size of the
group), the q u e stio n n a ire s were handed to each person, who
was in s tru c te d to put h is own number in the upper r i g h t cor
ner of the page. (The op eratio n s for measurement of C were
then c a rrie d o u t.) A fter t h i s , C -q u estio n n aires were turned
over on the ta b le , and new q u e stio n n a ire s were handed to
each. The o p eratio n s fo r measurement of E were next c a rrie d
o u t. In an attem pt to insure candor of the E -questionnair
responses, su b je c ts were assured th a t they would have a
chance to fin d out how well they had guessed t h e i r f r ie n d s '
86
responses, because they would see h is C -q u estio n n aire, but
th a t t h e i r frie n d s would have no way of knowing what the
guesses had been, because the £ - q u estio n n aire s would be tak
en up and put away immediately.
As the su b je cts fin ish e d the E -q u e stio n n a ire, p r e t
z e ls and so ft d rin k s were served to everyone fo r the ensuing
d iscu ssio n p erio d . E -q u estio n n aires were taken up and put
away, and C -q u estio n n aires were turned face up. An addi
tio n a l clean questionnaire was put on the ta b le . In stru c
tio n s were then given and the o perations for the measurement
of D were c a rrie d o u t. The d isc u ssio n s la s te d from 20 min
u te s to an hour and a h a lf, a t the d is c re tio n of the group.
The completion of the e n tire sequence took from an hour and
f i f t e e n minutes fo r the f a s t e s t group to two hours and a
h a lf fo r the slow est, with an average of about two hours,
counting only time spent a t the ta b le .
CHAPTER V
RESULTS OF THE EMPIRICAL STUDY
Introduction
This chapter p rese n ts in d e t a i l the findings of the
em pirical study reported in the l a s t ch ap ter, as they apply
to the r e s t r i c t e d theory of in te ra c tio n presented in Chap
te r I I I , The f i r s t section of the chapter is an evaluation
of the in d iv id u al hypotheses of the theory. The second sec
tio n of the chapter is an evaluation of the concepts, lo g ic,
and assumptions of the theory.
Evaluation of the Hypotheses
I n te r p re ta tio n of the v aria b le s
for p a irs
In the evaluation of hypotheses, each hypothesis was
te s te d as in te rp re te d for whole groups. However, as an an
a l y t i c a l aid , some of the concepts of the theory were in
te rp re te d fo r p a ir s of in te ra c to r s . In p a rt the p a ir analy
s is a s s is te d in c la rify in g the whole group a n a ly sis , and in
p a rt i t o ffered an evaluation of which le v e l was the most
th e o r e tic a lly productive: the group le v e l, cx the more e le
mentary p a ir le v e l. Most of the in te rp r e ta tio n s for p a irs
87
88
are contained in the development of the r e s t r i c t e d theory
in Chapter I I I , but they are reviewed here.
Cij =df the number of questions on the consensus
questionnaire on which in d iv id u al i and in d iv id
ual j gave id e n tic a l answers.
e^j =df the number of times i c o r re c tly p re d ic te d
j ' s response on the consensus q u e stio n n a ire .
P ij ("Perceived agreement") =df the number of times
i p re d ic te d for j the response i him self had
made on the consensus q u estio n n aire .
I - - d f the number of hours i in d icated he spent with
" J
j during an average school week.
The concept dr is discussed a t length in Chapter
I I I , and i s not repeated here. For the p a ir in te r p r e ta tio n ,
dr values are only used from groups of two, since only in
groups of two is dr derived from p a ir in te ra c tio n .
The concept of size of group i s , of course, not ap
p lic a b le to the p a ir in te r p r e t a ti o n .
General summary fo r whole groups. Table 2 p re se n ts
a summary of the ev a lu atio n s of the hypotheses of the re
s t r i c t e d th eo ry , juxtaposed w ith the statem ents from the
in c lu siv e theory from which the r e s t r i c t e d p ro p o sitio n s
were derived. I t must not be supposed th a t t h i s in d ic a te s
th a t the e v a lu a tio n a p p lie s to the in clu siv e theory as well
as to the r e s t r i c t e d th e o ry . As was in d ic ated before, the
TABLE 2
DISPOSITION OF HYPOTHESES DERIVED FROM THE THEORY
P roposition from the Inclusive Theory
R e stricted . . .
Hypothesis Tested D isposition
1. The g reater the consensus, the g reater the
frequency of in te ra c tio n .
2. The g reater the siz e, the le ss democratic the
group.
3. The g reater the siz e, the le ss the frequency of
in te ra c tio n .
4. The g reater the frequency of in te ra c tio n , the
g reater the understanding.
5. The le ss democratic a group, the le ss i t s
frequency of in te ra c tio n .
6 . The g reater the size, the le ss the consensus.
7. The g reater the consensus, the g reater the
understanding.
8 . The smaller the siz e, the greater the under
standing.
9. The g reater the consensus, the more democratic
the group.
C
N
N
I
D
N
C
a l + k l 1
a 2 " b2^
a3 " b3I
a4 + b4E
a5 - h s 1
a 6 " b6C
N — 3g + bgE
C = a g — b9D
Confirmed
Rejected
Confirmed
Rejected
Contradicted
Confirmed
Confirmed
Rejected
Rejected
o o
N O
TABLE 2—Continued
Proposition from the Inclusive Theory
R e stricted
Hypothesis Tested
D isposition
10. The more democratic the group, the g reater
the understanding
D = a lQ - b1QE Rejected
vO
o
91
e m p iric a l study i s a t e s t only of the r e s t r i c t e d th e o ry .
Table 3 p re s e n ts a d e t a il e d r e p o r t of the measure
ments of the in d iv id u a l v a r ia b le s of the th e o ry . Table 4
p r e s e n ts v alu es f o r the v a r ia b le s of the i n t e r c o r r e l a t i o n
c o e f f i c i e n t s . Table 5 p r e s e n ts i n t e r c o r r e l a t i o n c o e f f i c i e n t s
fo r the v a r ia b le s in te r p r e te d f o r p a i r s of i n t e r a c t o r s .
Hypotheses from previous
re s e a rc h
Size of group and in flu en ce c o n c e n tra tio n : N on D.
The c o r r e l a t i o n d eriv e d between the size of group and the
c o n c e n tra tio n of in flu e n c e in a single person (N on D) was
close to zero . This re s e a rc h did not in d ic a te any tendency
toward one person dominating the decision-m aking p ro c e ss in
larg e groups. Scores on the D -scale ranged from .03 to .68
( t h e o r e t i c a l l y p o ssib le range of zero to 1.00) with a mean
of .23, the la r g e r scores re p re s e n tin g the most co n cen tra
tio n of in flu e n c e . The h ig h e st scores as w ell as the lowest
came from groups of two. A s c a t t e r diagram of o b tain ed v al
ues in d ic a te d th a t the g r e a t e s t v a r ia tio n on the D -scale was
fo r groups o f two, w ith le s s extreme v alu es fo r in c re a sin g
group s iz e . The sc a tte rg ra m was d i s t i n c t l y pyram idal. This
h y p o th esis was r e j e c t e d .
Consensus and frequency of i n te r a c ti o n : C on I . The
r e l a t i o n s h i p between the C -sc o res fo r consensus, and f r e
quency of in t e r a c t i o n as measured by C, which b est in d ic a te s
92
TABLE 3
SUM M ARY OF THE POSSIBLE AND OBTAINED
RANGES, MEANS, AND STANDARD
DEVIATIONS OF THE VARIABLES
Inclusive
Concept
Re
s t r i c t e d
Concept
P o ssib le
Range
Obtained
Range
Mean
Standard
D eviation
Consensus C 0-55 5-47 25 10
Understanding E 0-55 13-37 24 5
Democracy D 0-1 .0 0 .03-,68 .23 .16
Frequency of
I n te r a c tio n
I 0-120 0-88 13 17
Size of Group N 2-12 2-7 3 1 .1
93
TABLE 4
PEARSONIAN INTERCQRRELATION COEFFICIENTS
FOR THE FIVE VARIABLES INTERPRETED
FOR W HOLE GROUPS (D .f. = 49)
D I N
Consensus C
Understanding E + .32a
Influence
C oncentration
D +.10 -.0 5
^ In te ra c tio n 1 ‘
Size of Group N -.81*3 +.11 - .0 5 - .3 3 a
a S ig n ific a n t beyond .05 le v e l,
k S ig n ific a n t beyond .01 le v e l.
94
TABLE 5
PEARSONIAN INTERCORRELATION COEFFICIENTS
FOR THE VARIABLES INTERPRETED
FOR PAIRS OF INTERACTORS
C C E P I dr
Consensus
Cij
-
Understanding
e i j
+ .69b
-
Perceived
Consensus Pi j
X3
C M
•
+
+ . 50b
Frequency of
In te ra c tio n *ij
+ . 25b + .20a +.10 -
R elative
Influence
dr *
+ .14 +.07 +.07 - .0 2
a S ig n ific a n t beyond .05 le v e l.
b S ig n if ic a n t beyond .01 le v e l.
* For p a ir s of course N does not afjply. Values of
dr a g a in st the other v a ria b le s were used from
groups of two only, since these are the only de
values derived from p a ir in te ra c tio n ; d . f . = 25
for c o r r e la tio n c o e f f ic ie n ts of dr .
I
i
95
the g ro ss amount of in t e r a c t i o n among members of the group,
is in d ic a te d by r = .30, which i s s i g n i f i c a n t beyond the .05
le v e l . C -v alu es (number of q u e stio n s on which a l l persons
agreed) ranged from 5 to 47, out of a p o ssib le range of 0 to
55. The d i s t r i b u t i o n was p o s i t i v e l y skewed. However, sep
a ra te d i s t r i b u t i o n s f o r groups of two and fo r groups of
la r g e r siz e y ie ld e d approxim ately normal d i s t r i b u t i o n s .
I-v a lu e s ranged from l e s s than 1 to 88, w ith a mean of 13
and a stan d ard d e v ia tio n of 17, w ith the d i s t r i b u t i o n p o si
t i v e l y skewed by the c o n c e n tra tio n of I-v a lu e s fo r groups of
more than two in the lower ran g e. Here ag ain , sep arate d is
t r i b u t i o n s fo r groups of two and fo r la rg e r sized groups
y ie ld e d approxim ately normal d i s t r i b u t i o n s .
On the b a s is of the c o r r e l a t i o n between C and I , the
h y p o th e sis i s accepted t h a t groups w ith high fre q u e n c ie s of
i n t e r a c t i o n have h ig h er consensus than groups w ith low in
t e r a c t i o n fre q u e n c ie s . In sp e c tio n of s c a t t e r diagrams sug
g ested the p a r t i a l l i n g of the in flu en ce of group size on
t h i s c o r r e l a t i o n . The r e s u l t i n g c o r r e l a t i o n between C and I
w ith N held c o n s ta n t, i s near zero ( r = .1 3 ), dem onstrating
t h a t most of th e c o r r e l a t i o n between C and I can be ac
counted fo r by v a r i a t i o n of N. This in tu r n suggested the
i n t e r p r e t a t i o n of C and I on the le v e l of p a i r s . For the
101 d i s t i n c t p a i r s used in measuring E, the c o r r e la tio n be
tween C and I was .25, s i g n i f i c a n t a t the .01 le v e l . These
fin d in g s can be summarized in t h i s way: fo r groups of size
N, most of the c o rr e la tio n between C and I is accounted for
by v a ria tio n of N, but when in te rp re te d for p a irs of in te r
a c to rs, C and I are s ig n ific a n tly c o rre la te d .
Consensus and size of group: C on N. C o rre la tio n
between N and C proved to be highest of any in the matrix of
t h i s study, with r = - .8 1 . A ctually, a s c a tte r diagram of
the obtained values in d ic a te s a probable c u r v ilin e a r ity ,
with the la rg e s t differen ce in C between groups of size two
and th re e , and with s lig h t increase in C between consecutive
values of N of three and above. The Pearsonian c o e ff ic ie n t
quoted is th erefo re probably a s lig h t underestimate of the
degree of c o r r e la tio n . The hypothesis is confirmed th a t
smaller groups have higher consensus than la rg e r groups.
Size and frequency of in te ra c tio n : N on I . Fre
quency of in te ra c tio n when in te rp re te d as I proved to be
s ig n ific a n tly re la te d to N, size of group, with r = - .3 3 ,
which is s ig n ific a n t beyond the .05 le v e l. This confirms
the hypothesis th a t the smaller the group, the higher the
frequency of in te ra c tio n . If frequency of in te ra c tio n is
in te rp re te d as T, the c o rre la tio n between T and N is -.5 3 ,
s ig n ific a n t beyond the .01 le v e l. S c a tte r diagrams revealed
th a t the re la tio n s h ip is n o n -lin ea r, however, p a r tic u la r ly
for N on T. The shape of the s c a tte r diagram looks lik e
t h is :
97
5
N 4
3 ........................
2 ...............................................................................
0 10 20 30 40 50
T
I t is c e r ta in ly dubious whether one is j u s t i f i e d in using a
lin e a r equation as a mathematical model to describe t h i s
type of c o n fig u ra tio n .
Frequency of in te ra c tio n and influence concentra
tio n ; I on D. Homan's hypothesis th a t the more equal the
influence of members of a group* the le s s the frequency of
in te ra c tio n was co n tra d ic te d by the r e s u l t s of t h i s study.
The c o r r e la tio n between D and I was r = +.29, s ig n if ic a n t a t
beyond the .05 le v e l. I n te rp re te d , t h i s means th a t groups
with the le a s t equal influence have the h ig h est frequencies
of in te ra c tio n . Inspection of the s c a tte r diagram however
was in t h i s case very i n s t r u c tiv e . P lo ttin g obtained values
of D a g a in st I and in d ic a tin g the size of the group in which
each observation was taken showed th a t the e n tire p o sitiv e
c o r r e la tio n is accounted fo r by groups of two, with a nega
tiv e c o r r e la tio n in d ic ated in the p a tte r n fo r groups of more
than two. Although the d i s t r ib u t io n s of the two groups taken
se p a ra te ly are reasonably homoscedastic, the combined d is
tr ib u ti o n is not a t a l l homoscedastic. Most values are then
98
concentrated in the lower l e f t of the diagram, with other
values sp a rse ly s c a tte re d . The fin d in g s here again p o in t
to the probable a n a ly tic a l gain in separate treatm ent of
groups of two from groups of more than two.
Other than i t s r e la tio n s h ip to I, Do showed no sig
n if ic a n t c o r r e la tio n s to any of the other v a ria b le s .
Understanding and frequency of in te ra c tio n ; E on I .
With understanding in te rp re te d as E and frequency of in t e r
a ctio n as I, the hypothesis th a t the more fre q u e n tly a group
i n te r a c ts , the more they w ill understand one another i s not
confirmed. The r = -.1 2 , which is quite in s i g n if i c a n t, and
in the opposite d ire c tio n to the p re d ic tio n .
In view of the fin d in g s described above in connec
tio n w ith the r e la tio n s h ip of C and I, scores fo r the E and
I v a ria b le s were in te rp re te d and c o rre la te d fo r the same 101
p a irs used above. When in te rp re te d on the le v e l of p a i r s ,
the c o r r e la tio n between E and I yielded an r of .20, s ig n if
ic a n t a t beyond the .05 le v e l, and in the d ir e c tio n p red ic
te d . Although the hypothesis was r e je c te d fo r groups of
size N, i t was confirmed fo r p a ir s of in te r a c to r s .
Hypotheses deduced from the
theory
Consensus and understanding: C on E. In te rp re tin g
understanding as E and consensus as C, the obtained values
of these sc a le s c o r re la te s ig n if ic a n tly a t the .05 le v e l.
99
r = .32. Examination of the s c a t t e r diagram however r e v e a ls
two q u ite sep arate d i s t r i b u t i o n s . When v alu es of E a g a in s t
C are p l o t t e d in d ic a tin g the siz e of group in which each
value was observed, groups of siz e two form one c l u s t e r , and
groups of more than two form another c l u s t e r . Separate cor
r e l a t i o n fo r groups of two y ie ld s an r of .69 and fo r groups
of more than two an r of .66. The h y p o th esis i s accepted
th a t the g r e a t e r the consensus in a group, the more able are
i t s members to p r e d ic t one a n o t h e r 's resp o n se s.
Consensus and in flu en c e c o n c e n tra tio n : C on D.
Another of the hypotheses deduced from the th eo ry was t h a t
the more consensus a group e x h ib ite d , the more equal i t s
in flu en ce d i s t r i b u t i o n . This h y p o th esis was not supported
by the d ata of t h i s study. The r of .11 was i n s i g n i f i c a n t ,
and in the d i r e c t i o n opposite to the p r e d ic tio n .
Understanding and siz e of group: E on N. The hy
p o th e s is t h a t the sm aller the group, the more the members
would be able to p r e d ic t one a n o t h e r 's resp o n se s, which was
deduced from the th e o ry , was not supported. The c o r r e l a t i o n
obtained between E and N was .11, in the d i r e c t i o n opposite
from the p r e d ic tio n , and q u ite i n s i g n i f i c a n t . P a r t i a l co r
r e l a t i o n a n a ly s is re v e a le d t h a t when the in flu en c e of C on E
and N i s p a r t i a l l e d o u t, the c o r r e l a t i o n between E and N is
i s high: r = .67, w hile w ith C varying i t was only .11. I t
i s not immediately apparent why the average amount of under-
1 0 0
standing in the group should be higher as the size of the
group in c re a se s. This is e x a c tly the opposite of the p re
d ic te d r e la tio n s h i p . I t also produces a p e c u lia r se t of
p ro p o sitio n s:
1. The g re a te r the consensus, the g re a te r the un
d erstan d in g , size of group held constant
(r = .71).
2. The g re a te r the size the le s s the consensus,
understanding held co n stan t (r = - .9 0 ) .
3. The g re a te r the size the more the understanding,
consensus held co n stan t (r = .67).
Under the logic developed in Chapter I I , these would be
incompatible or even c o n tra d ic to ry , yet each i s v e r if ie d by
em p irical d a ta .
Understanding and influence co n cen tratio n : E on D.
The deduced hypothesis th a t the more understanding th e re is
in a group, the more equal W ill be the influence of members
on group d e c isio n s, was not confirmed. The obtained c o rre
la tio n c o e f f ic ie n t of E a g a in st D was r = - .0 5 , of no sig
n ifica n ce .
p o ssib le confounding f a c to rs
In te llig e n c e as a p o ssib le confounding f a c t o r . I t
was thought th a t th e re was a p o s s i b i l i t y th a t in te llig e n c e
of the members of the groups or the mean in te llig e n c e of
1 0 1
d if f e r e n t groups might be confounding v a ria b le s in the anal
y s is . In order to explore t h i s p o s s i b i l i t y , I.Q. 's were
obtained on a l l but ten su b je c ts on the O tis Gamma t e s t . On
the le v e l of group a n a ly s is , mean group I . Q . 's had very low
c o r r e la tio n s w ith a l l the v a ria b le s of the study. Neverthe
le s s , the p o s s i b i l i t y remained th a t in d iv id u a l v a r ia tio n in
I.Q ., obscured in group averages, was in flu en c in g our v a r i
a b le s. A ccordingly, values of the concept " r e la tiv e I.Q ."
measuring the in d iv i d u a l's I.Q. w ith the mean I.Q . of h is
group as 100 were computed as follow s:
IQr =df 100 - (15 - IQ)
where " I Q i s the mean I.Q. of members of the group, and IQ
is the in d iv i d u a l's I.Q. C o rre la tio n s of IQr w ith the v a r i
ab les of t h i s study were sm all, none reaching s ig n ific a n c e .
I t i s concluded th a t I.Q. i s not an im portant confounding
f a c to r in these r e s u l t s .
Age or grade le v e l as p o ssib le confounding f a c t o r s . '
The p o s s i b i l i t y was explored th a t the o ld e r a person was in
r e l a t i o n to other members of the group, the higher h is in
flu en c e, and th a t the higher a p e rs o n 's grade le v e l in r e l a
tio n to o th e rs in the group, the higher h is in flu e n c e .
These hypotheses proved w ithout support. Most groups were
very c lo se in age. In 28 groups, a l l members were in the
same grade. In the o th er groups, th e re was no tendency fo r
the upperclassman to have higher influence in h is group.
1 0 2
S o cia l c la s s and in flu e n c e . I t was thought th a t the
d is t r ib u t io n of influence scores (dr and perhaps D) might be
a ffe c te d by so c ia l c la s s p o s itio n . The groups in t h i s study
were considerably more heterogeneous in c la s s composition
(as measured by a four-way breakdown of f a t h e r s ' rep o rted
occupations) than Hollingshead rep o rted in Elmtown. The oc
cu p atio n al c l a s s i f i c a t i o n used was as follow s: p ro fe ssio n a l
and major managerial; other white c o l la r ; s k ille d workers;
se m i-sk ille d and u n s k ille d workers. Only ten groups were
homogeneous on t h i s v a ria b le . Sixteen groups had a t le a s t
a th r e e - c la s s span, and of these seven covered a f o u r- c la s s
span. Of the 41 nonhomogeneous groups, only 16 supported
the hypothesis th a t c la s s and influence were p o s itiv e ly r e
la te d . In 13 cases c la s s and influence were negatively re
la te d , and in 12 cases th ere was no re la tio n s h ip d isc e rn ab le .
C lass was th e re fo re dism issed as a p o ssib le confounding fac
to r in the influence r e l a t i o n .
E valuation of the Theory
A nalysis of the r e s u l t s of the em pirical study r e
ported in Chapter IV, which was conducted to t e s t the pre
d ic tiv e value of the theory presented in Chapter I I I , sug
g e s ts sev eral p o ssib le m odifications of the th e o ry . But in
m odification of the th eo ry , i t i s d i f f i c u l t to discover from
what p ro p o sitio n s or assumptions of the theory stem the e r
r o r s and in c o rre c t p r e d ic tio n s . Are they due to fa u lty
103
p r o p o s itio n s , in a c cu ra te measurement, in a p p ro p ria te lo g ic a l
assum ptions, com binations of th e s e , or to o th e r s t i l l un
named causes? I t i s im possible to say from one a n a ly s is
such as t h i s . O tto Neurath put i t t h i s way:
. . . p a r t i c u l a r l y in the s o c ia l scien c es we have
h a rd ly any use fo r the r e f u t a t i o n - v e r i f i c a t i o n asymmetry
of c e r t a i n c a lc u la to r y schemes. W e look a t a network of
hypotheses only, and we cannot say from which hypotheses
c e r t a i n d i f f i c u l t i e s a r i s e . W e have to s t a r t p l u r a l i s
t i c a l l y a l l over again . . . S c i e n t i s t s in a l l sc ien ces
t r y to f i t hypotheses in to a c l u s t e r of o th er hypothe
ses, o b se rv a tio n sta tem en ts and o th e r accepted s t a t e
ments. But c e r t a i n d e f e c ts , e . g . , w ell d e scrib ed con
t r a d i c t i o n s , do not always induce s c i e n t i s t s to d is c a rd
a h y p o th e sis. They may m aintain t h a t t h i s h y p o th esis is
o fte n u s e fu l and th a t th e re i s no o th er more a t t r a c t i v e
h y p o th e sis. Newton's g r a v ita tio n h y p o th esis has been
used in s p ite of the f a c t th a t fo r about 150 years
sc h o la rs have f e l t again and again t h a t th e re were con
t r a d i c t o r y and ambiguous elem ents in Newton's hypothe
s i s , but i t appeared so su c c e s sfu l in analyzing the
movements of bodies t h a t only a few sc h o lars r e a l l y
c r i t i c i z e d the d e fe c ts of t h i s h y p o th e s is .1
Three ty p es of m o d ific atio n of the th eo ry are d is
cussed in t h i s se c tio n : (1) m o d ific atio n and e lim in a tio n of
v a r ia b le s , (2) m o d ific atio n of the le v e l of a n a ly s is , and
(3) m o d ific a tio n of deductive assum ptions.
M o d ificatio n and e lim in a tio n
of v ariab le's
M ultiple lin e a r r e g r e s s io n a n a ly s is was computed fo r
each of the fiv e v a r ia b le s as dependent v a r ia b le s . (See
Table 6 .) This a n a ly s is to g e th e r w ith the c o r r e la tio n anal-
^Otto N eurath, "Foundations o f the S o c ia l S c ie n c es,"
I n te r n a ti o n a l Encyclopedia of U n ified S c ie n c e . Vol. I I :
iFoundations of the U nitv of S c ie n c e . No. 1 (Chicago: U niver
s i t y of Chicago P re s s , 1944), pp. 25-26.
104
TABLE 6
MULTIPLE CORRELATION ANALYSIS
FOR THE FIVE VARIABLES
INTERPRETED FOR W HOLE GROUPS
(D .f. = 46)
Dependent
V ariable
C o e ff ic ie n t
of
D eterm ination
M ultiple
C o rre la tio n
C o e f f ic ie n t
Size of Group N
.82
. 90
Consensus .84 .92*
U nderstanding .54 .73*
Influence
C o n c e n tra tio n
D .10 .32
Frequency of
I n te r a c tio n
.20 .45*
a S ig n if ic a n t beyond .05 le v e l
k S ig n if ic a n t beyond .01 le v e l
105
y s is used to re p o rt r e s u l t s in Chapter IV form the b a sis of
the ev a lu atio n of each v aria b le in the theory.
Consensus: C. While the c o r r e la tio n between C and
N was r = - .8 1 , the m ultiple c o r r e la tio n c o e ff ic ie n t for C
as dependent v ariab le w ith the other four v a ria b le s as in
dependent p re d ic to rs was R = .91, in d ic a tin g th a t the other
v a ria b le s co n trib u te s u b s ta n tia lly and independently from N
to the p re d ic tio n of the v a ria tio n of C. I t is th e re fo re
considered worth-while to r e t a i n C in the system.
U nderstanding: E . Although none of the c o r r e la tio n s
with E were very high in the group in te r p r e ta tio n , the high
e s t being w ith C (r = .32), the m ultiple c o rr e la tio n c o e f f i
c ie n t for E in the system was R = .74, in d ic a tin g th a t more
than h a lf the variance of E is accounted for by the other
v a ria b le s of the system. I t i s th e re fo re considered worth
while to r e t a i n E in the system.
Influence c o n cen tratio n : D. Influence con cen tratio n
showed l i t t l e r e la tio n s h ip to the other v a ria b le s of the
system. I t s h ig h e st in d iv id u al c o r r e la tio n was w ith I
(r = .29). I t s c o e f f ic ie n t of m ultiple c o rr e la tio n w ith the
other v a ria b le s of the system is R = .31, in d ic a tin g no
worth-while independent r e la tio n s h ip to any of the other
v a ria b le s . In t h i s case R does not even reach sig n ific an c e
a t the .05 le v e l. I t i s th e re fo re not considered worth-
while to r e t a i n D in the system.
106
Frequency of in t e r a c t i o n : I . Frequence of in te r a c
tio n as in te r p r e t e d as I showed s i g n i f i c a n t c o r r e l a t i o n w ith
a l l the v a r ia b le s in the system except E. I t s c o r r e l a t i o n
w ith C was .30, w ith D .29, and w ith N - .3 3 . P re d ic te d from
the o th e r four v a r ia b le s , the m u ltip le c o r r e la tio n c o e f f i
c i e n t fo r I as dependent v a ria b le was R = .45. Below i s a
comparison of the c o r r e l a t i o n s of I and T w ith the o th e r
v a r ia b le s of the system.
I T
c .30
O '
if)
•
E
-.1 2 .15
D .29
.21
N - .33
i
•
cn
CO
F i r s t , i t i s ev id en t from the l i s t of c o r r e l a t i o n s t h a t p re
d ic tio n of C and N would be much b e t t e r from T than from I.
Second, the c o r r e l a t i o n between T and E i s in the d i r e c t i o n
p r e d ic te d by the th e o ry (o p p o site from the r e l a t i o n s h i p of I
and E ), and only la ck s ach iev in g s ig n ific a n c e a t the .05
le v e l by a f r a c t i o n in the t h i r d decim al p la c e . The r e l a
tio n s h ip of T to D i s not much lower than t h a t of I to D.
I t i s th e re fo re co n sid ered d e s ira b le th a t fo r a th e o ry of
small groups, frequency of in t e r a c t i o n be in te r p r e t e d as T.
T of course has i d e n t i c a l meaning fo r groups of size N and
107
for p a i r s of i n t e r a c t o r s . F u rth e r, i t i s suggested th a t
measurement of T be re f in e d to include only time when a c tu a l
communication i s tak in g place between i n t e r a c t o r s , in ste a d
of simply time spent to g e th e r. This c r e a te s measurement
problems not p resen ted by the p re se n t i n t e r p r e t a t i o n , but
may r e s u l t in increased t h e o r e t i c a l relevance of the v a r i
able .
Size of group: N. Size of group showed s ig n i f ic a n t
c o r r e la tio n s to C (r = -.8 1 ) and I ( r = - . 3 3 ) . I t s m u ltip le
c o r r e la tio n c o e f f ic ie n t from the four v a ria b le s of the sys
tem was R = .91, but i t showed l i t t l e r e la tio n s h ip to E
(r = .1 1 ). The a n a ly s is of the r e la tio n s h i p s of N p resen ted
in Chapter IV suggests th a t i f a theory i s going to co n tain
a siz e-o f-g ro u p v a r ia b le , some m o d ific atio n s w ill be needed
in i t s in c o rp o ra tio n . The d i s t r i b u t i o n shapes obtained in
t h i s study showing size of group p lo tte d a g a in st o th e r v a r i
ab les seem lik e ly to be ty p ic a l of fre ely -fo rm in g groups.
Most d i s t r i b u t i o n s p lo tte d on size were of the shape in d i
cated on page 98 above, w ith a c u r v ilin e a r boundary, and the
lower l e f t co rn er of the graph f i l l e d . Groups of two show
much g r e a te r v a rie ty on a i l the v a r ia b le s stu d ied here than
any o th er group s iz e . Each la rg e r size shows a g r e a te r r e
s t r i c t i o n of v a r ia tio n , w ith the g r e a te s t d iffe re n c e between
size two and size th r e e , and p ro g re ssiv e ly sm aller d i f f e r
ences in each succeeding in te r v a l.
108
The d iffe re n c e between groups of two and la rg e r
sized groups i s so marked th a t in some cases r e s u l t s from
groups of two formed a separate d i s t r i b u t i o n on graphs from
the r e s t of the groups. This suggests th a t separate p ro v i
sions may have to be made in the theory fo r groups of two.
A review of the in te ra c tio n frequency data above
confirm s o b se rv a tio n s of previous r e s e a rc h e rs th a t most
groups of more than two persons are o ccasio n al groupings
composed of se v era l p a irin g s , each of which i n t e r a c t s a t
le a s t as much and almost always more than the in c lu siv e
group. These p a i r s w ith in la rg e r groups d isp la y a l l the
v a r ia tio n in frequency of in te ra c tio n of the groups of two,
ranging from a few minutes per week to more than a hundred
hours per week. James has even suggested th a t any s t a b i l i t y
of s tru c tu re th a t groups la rg e r than two d isp la y i s due to
2
t h e i r dyadic s u b s tru c tu re .
I t i s the proposal of t h i s r e s e a rc h e r, in view of
the above f a c t s , th a t two th e o r ie s are c a lle d fo r a t t h i s
stage in the e x p lo ra tio n of group in te r a c tio n : one d ealin g
w ith c h a r a c t e r i s t i c s and p a tte r n s of dyads, or ev en the two
r e la tio n s h i p s w ith in the dyad; the o th e r, w ith concepts
d e f i n i t i o n a l l y coordinated to the f i r s t , d ealin g w ith small
groups and including size as a variable,,
*\John James, "A P relim in ary Study of the Size D eter
minants in Small Group I n te r a c tio n ," American S o c io lo g ic a l
Review. XVI (August, 1951), 476-77.
109
M odification of the le v e l of
a n a ly sis'
Comparison of group and p a ir i n t e r p r e t a t i o n s . The
p o s s i b i l i t y th a t a n a ly tic a l advantage would be gained in
t h i s theory by in te r p r e ta tio n of the v a ria b le s fo r p a i r s of
in te r a c to r s r a th e r than fo r whole groups has been explored
at sev eral p o in ts above. M ultiple c o rr e la tio n a n a ly s is of
the v a ria b le s fo r groups and for p a ir s was computed to a f
ford the opportunity to assess the r e l a t i v e p re d ic tiv e e f f i
ciency of each in te r p r e ta tio n . In Table 7 we p rese n t the
m ultiple c o rr e la tio n c o e f f ic ie n ts for each v aria b le in te rp r e
ted both for groups and fo r p a ir s .
From the a n a ly sis of in d iv id u al v a ria b le s presented
e a r l i e r , i t was concluded th a t much could be learned about
the nature of the re la tio n s h ip between v a ria b le s by studying
p a ir r e la tio n s h ip s , but from the fig u re s presented here, i t
is p la in th a t th ere is g re a te r p re d ic tiv e e ffic ie n c y in the
group theory than in the theory in te rp r e te d for p a i r s .
S ta tic and dynamic p ro p o sitio n s d is tin g u is h e d . There
i s no reason to assume th a t s t a t i c p ro p o sitio n s r e f e r r in g
c r o s s - s e c tio n a lly to a population a t a given time w ill sta te
the same r e la tio n s h ip s between v a ria b le s as dynamic proposi
tio n s r e f e r r in g to the r e la tio n s h ip s between two v a ria b le s
over time as measured on the same ca ses. For example, i t
may be tru e th a t for given p a irs of in te r a c to r s , as one
1 1 0
TABLE 7
COMPARISON OF MULTIPLE CORRELATION
COEFFICIENTS OF THE VARIABLES
INTERPRETED FOR W HOLE GROUPS
AND FOR PAIRS OF INTERACTORS
Groups P a ir s
Consensus
Understanding
C
E
.91
.74
.71
(e) .76
Frequency of
I n te r a c tio n
.45 .27
Influence
C oncentration
.31 (dr ) .15
Size of Group N .91 not ap p licab le
P erceived Consensus P not ap p lica b le .51
I l l
person in a p a ir more and more dominates the influence r e l a
tio n , the in te ra c tio n frequency for th a t p a ir w ill decrease,
while a t the same time not be true th a t a t any given time
those p a irs with the le a s t symmetrical influence r e la tio n
w ill in te ra c t le s s fre q u en tly . These p ro p o sitio n s, i t is
suggested, should be a p a rt of two d if f e r e n t th e o rie s for
the time being, and need to be te s te d by d if fe re n t methods.
Some of the p ro p o sitio n s of the present theory (the symmet
r i c a l influence p ro p o sitio n used as an example in t h i s para
graph among them) were o rig in a lly proposed as dynamic propo
s itio n s , but were te s te d here as s t a t i c p ro p o sitio n s.
M odification of the deductive
assumptions
Of the six hypotheses which were deduced from the
four axioms of the theory, two were confirmed (theorems 2
and 3), two were co n trad icted by th e i r opposites (theorems 1
and 4 ), and two were not confirmed (theorems 5 and 6).
Three which were fa u lty were deduced from axiom 2, which
t h i s rese arch has re fu te d . I t is th e re fo re d i f f i c u l t to as
sess the p re d ic tiv e value of the deductive theory, since
h a lf the deductions were made from a fa u lty axiom. Of the
th ree which were deduced from axioms which are s t i l l accept
able, two were confirmed and the th ir d co n tra d icte d . This
theory i s a case lesson to the s o c io lo g is ts engaged in de
ductive theory b u ild in g . The p ro p o sitio n s on which the th e
ory was constructed rep resen ted some low -level c o rre la tio n s
112
and some u n r e lia b le c o r r e la tio n s . A u se fu l deductive theory
w ill e v id e n tly re q u ire r e l i a b l e , high c o r r e la tio n s as i t s
foundation.
In view of the r e s u l t s of the em p irical t e s t of the
theory p re se n te d , i t can be concluded th a t the p ro p o sitio n s
deduced were not r e l i a b l e . Should t h i s tu rn out to be ty p i
c a l of deductions made on the b a s is of the assumption which
was in co rp o rated in t h i s theory ( th a t any s ig n if ic a n t c o rre
la tio n should be in te r p r e te d as a p e r f e c t c o r r e la tio n in
th e o ry ), t h i s assumption should be abandoned on the mathe
m atical le v e l. I t is d o u b tfu l a t th a t p o in t whether th e re
would be any j u s t i f i c a t i o n fo r continuing i t s use on the
in c lu siv e le v e l of theory c o n s tru c tio n .
CHAPTER VI
SUM M ARY
The problem of t h i s d is s e r a tio n was th re e fo ld :
1. To develop the logic fo r co n stru c tio n of a fo r
mal theory of in te ra c tio n in small groups.
2. To use t h i s logic fo r the a c tu a l co n s tru c tio n of
a substantive fiv e -v a ria b le theory of in te ra c
tio n in small groups.
3. To conduct an em p irical t e s t of the im p licatio n s
of the theory to evaluate i t s u se fu ln e ss in pre
d ic tin g in te r a c tio n in a sample of same-sex nat
u r a l adolescent groups.
The f i r s t problem was to develop the logic of theory
c o n s tru c tio n . This req u ired the adoption of a general con
ception of s c i e n t i f i c theory, a logic of concept form ation,
and a logic of p ro p o sitio n s . The general conception adopted
for t h i s study was th a t a theory is a set of p ro p o sitio n s
some of the terms of which are defined, and a set of d e f in i
tio n s connecting the defined terms of the p ro p o sitio n s to a
set of undefined observation term s. I t was s p e c if ic a lly
proposed th a t t h i s theory was to be p resen ted on a veiy low
113
i
1 1 4
le v el of a b s tr a c tio n , and was to be developed to ex p la in
known f a c t s of a q u ite lim ite d range. This was based on the
assumption th a t i t i s p r a c t i c a l to develop th e o r ie s of high
e r a b s tr a c tio n only a f t e r the lo w er-lev e l th e o rie s which
they include have been developed. W e adopted as presupposi
tio n s th a t the logic of s c i e n t i f i c procedure is the same fo r
a ll scien c es, and th a t the only adm issible o p e ra tio n s in
confirm ation were logic and o b se rv atio n .
The follow ing p o in t of view was developed fo r con
cept form ation fo r t h i s study.
1. Concepts should be chosen fo r a th eo ry on the
b a s is of t h e i r close r e la tio n s h i p to o th e r con
c e p ts , not so le ly on the b a s is of t h e i r s o c ia l
importance.
2. Concepts fo r t h i s theory were to be developed
from s o c ia lly im portant concepts by reducing the
s o c ia lly im portant concepts to an elem entary
le v e l of a b s tra c tio n .
3. The concepts of the th eo ry were to be considered
" fre e c r e a tio n s of the i n t e l l e c t , " i . e . , not im
posed by the nature of the world of ev e n ts.
4. The o p e ra tio n a l i n t e r p r e t a t i o n of each concept
should be considered one of many p o ssib le par
t i a l in t e r p r e t a t i o n s which might be recognized
as i n t e r p r e t a t i o n s of the same concept under
s tip u la te d c o n d itio n s.
115
Under the logic of p ro p o sitio n s , i t was decided to
c o n stru c t the theory of in te ra c tio n f i r s t in the language
of speech, then in mathematical eq u atio n s. Two types of
deductions were described as necessary for a theory of the
type under d iscu ssio n : deductions using d e f in itio n s only,
and deductions through t r a n s i t i v e r e la tio n s h ip s . Deductions
made through d e f in itio n s are only those made on the p a tte r n
of the t r a d i t i o n a l syllogism using the logic of c la s s e s .
Deductions through t r a n s i t i v e r e l a t i o n s are those in which
the re la tio n s h ip between two concepts is deduced on the ba
s is of t h e i r common r e la tio n s h ip to a t h ir d .
I t was demonstrated th a t in order to use deductions
through t r a n s i t i v e r e la tio n s h ip s in our theory, i t was nec
essary to assume th a t the p ro p o sitio n s described p e rfe c t
lin e a r r e la tio n s h ip s . This assumption req u ired an a d d itio n
a l assumption to f i t em p irical so c io lo g ic a l data to the
theory: th a t any s ig n if ic a n t em pirical c o r r e la tio n was to be
in te rp re te d as p e rfe c t c o r r e la tio n in theory.
The logic for co n stru c tio n of the theory of in te r a c
tio n having been developed, the f i r s t problem was completed.
The co n stru c tio n of a fiv e -v a ria b le theory of in te ra c tio n in
small groups was the second problem. The theory was f i r s t
presented in the language of speech with nominally defined
concepts. Five concepts of so c ia l importance, freq u en tly
studied by s o c io lo g is ts , were se lec ted : consensus, under
standing, frequency of in te ra c tio n , democracy, and size of
116
group. From these a s e t of common-sense p ro p o s itio n s was
developed. A review was p re se n ted of s tu d ie s involving
these v a ria b les* in d ic a tin g the nature of t h e i r i n t e r r e l a
ti o n s h i p s . On the b a s is of th ese s tu d ie s , a s e t of te n
p r o p o s itio n s was developed. Four of th e se were d esig n ate d
axioms, and the o th e r s ix , d esig n ate d theorem s, were deduced
from the four axioms. I t was p o in te d out t h a t in t h i s form
the th e o ry was s t i l l not t e s t a b l e , since the d e f i n i t i o n s
s t i l l d id not ambiguously e s t a b l i s h connection between the
t h e o r e t i c a l term s and a se t of o b se rv a tio n term s.
Although the te n p ro p o s itio n s were not te s te d per
s e . they formed the b a s is fo r the p ro p o s itio n s which were
t e s t e d . The te n p r o p o s itio n s of the in c lu s iv e theory were
as fo llo w s:
Axioms:
1. The g r e a te r the consensus, the g r e a te r the f r e
quency of i n t e r a c t i o n .
The g r e a te r the s iz e , the le s s dem ocratic the
group.
3. The g r e a te r the s iz e , the le s s the frequency of
i n t e r a c t i o n .
4 . The g r e a te r the frequency of in t e r a c t i o n , the
g r e a te r the u n d erstan d in g .
Theorems:
1. The le s s dem ocratic a group, the le s s i t s f re -
117
quency of in te ra c tio n .
2. The g re a te r the siz e , the le s s the consensus.
3. The g re a te r the consensus, the g re ater the un
derstanding.
4. The sm aller the siz e , the g re a te r the under
standing.
5. The g re a te r the consensus, the more democratic
the group.
6 . The more democratic a group, the g reater the
understanding.
Next, a r e s t r i c t e d theory was developed from the in
clusive theory in which each concept of the theory was de
fined o p e ra tio n a lly for the domain of same-sex n a tu ra l ado
lescen t groups. These d e f in itio n s were b u ilt around re
sponses to q u estio n n aires designed s p e c if ic a lly for t h i s
study, but based upon operations commonly used in previous
stu d ies of the v a ria b le s, so as to maximize the r e p lic a tiv e
value of t h i s re se a rc h . The p ro p o sitio n s of the theory were
then sta te d in terms of the r e s t r i c t e d concepts, in the form
of ten simultaneous lin e a r equations. With the theory in
t h i s form, and the logic developed fo r theory c o n stru c tio n ,
so lu tio n to the th ird problem was undertaken: the conduct of
an em pirical t e s t to evaluate the usefulness of the theory
in p re d ic tin g in te ra c tio n in a sample of same-sex n atu ra l
adolescent groups.
The universe selec ted was the student body of a
i
118
Southern C a lif o r n ia high school. Because p ro b a b ility s ta
t i s t i c s were to be used in a ssessin g the f i t of the data to
the theory* i t was considered e s s e n t i a l to have a random
sample of groups, other w r i te r s to the c o n tra ry notw ith
sta n d in g .. The f i n a l sample included 51 groups of adoles
c e n ts . The experim ental work was conducted w ith these
groups during the summer school v a c a tio n , 1959.
Using data gathered from the in te r a c ti o n of th ese
adolescent groups, t e s t s of the te n hypotheses of the r e
s t r i c t e d theory were made, w ith the follow ing r e s u l t s :
1. The g re a te r the consensus, the g re a te r the f r e
quency of in te r a c tio n : confirm ed.
2. The g re a te r the s iz e , the le s s democratic the
group: r e je c te d .
3. The g re a te r the s iz e , the le s s the frequency of
in te r a c tio n : confirm ed.
4. The g re a te r the frequency of in te r a c tio n , the
g re a te r the understanding: r e je c te d .
5. The le s s democratic a group, the le s s i t s f r e
quency of in te r a c tio n : c o n tra d ic te d . The data
supported the conclusion t h a t the le s s democrat
ic the group, the g r e a te r i t s frequency of in
te r a c tio n .
6 . The g re a te r the s iz e , the le s s the consensus:
confirmed.
7. The g re a te r the consensus, the g r e a te r the un-
119
d e rsta n d in g i confirm ed.
8 . The sm aller the s iz e , the g r e a te r the under
sta n d in g : r e j e c t e d .
9. The g r e a te r the consensus, the more dem ocratic
th e group: r e j e c t e d .
10. The more dem ocratic the group, the g r e a te r the
u n d erstan d in g : r e j e c t e d .
The r e s u l t s showed t h a t the th e o ry was of only lim
ite d u s e fu ln e s s in p r e d ic tin g the i n t e r a c t i o n measurements
ob tain ed from the groups. The r e s t r i c t e d i n t e r p r e t a t i o n s of
consensus, u n d ersta n d in g , frequency of i n t e r a c t i o n , and size
of group showed moderate to high i n t e r c o r r e l a t i o n s in some
c a s e s , but the in flu en ce v a ria b le developed from th e concept
democracy showed l i t t l e r e l a t i o n s h i p to the o th e r v a r ia b le s
of the system. Many of the r e l a t i o n s h i p s proved to be non
l i n e a r as measured by our in stru m e n ts, and se v e ra l s c a t t e r
diagrams d id not meet the req u irem e n ts of h o m o sced asticity
because of i r r e g u l a r d i s t r ib u t io n s * A n aly sis of p a r t i a l
c o r r e l a t i o n d ata and in s p e c tio n of the s c a t t e r diagrams of
o btained measurements in d ic a te d t h a t as in te r p r e t e d in t h i s
th e o ry , i n t e r a c t i o n p a t te r n s in dyads d if f e r e d so d i s t i n c t l y
from p a t t e r n s in la r g e r groups th a t f o r th e p re se n t i t
seemed t h a t more t h e o r e t i c a l p ro g re ss could be made by sep
a ra te a n a l y t i c a l trea tm e n t of groups of two from groups of
la r g e r s iz e . At the same tim e i t was shown t h a t on the
b a s is of d ata c o l le c te d , i n t e r p r e t a t i o n of most v a r ia b le s
120
for whole groups r e s u lte d in g re a te r p re d ic tiv e power than
when the v a ria b le s were in te rp re te d fo r p a ir s of in te ra c
t o r s .
Although from th e confines of the p resen t work i t
was d i f f i c u l t to determine from what p a rt of the theory d is
crepancies arose, i t was c le a r th a t deductions based on the
assumption of p e rfe c t lin e a r c o r r e la tio n proved to be unre
l i a b l e . Most of the deduced theorems were not confirmed by
the d a ta . I t was suggested th a t should sim ila r r e s u l t s
prove r e g u la rly to be the case when deductive assumptions
such as were found necessary are used, such assumptions
should be abandoned, and only deductions j u s t i f i e d by the
approximate degree of obtained r e la tio n s h ip s should be used.
This was recommended in sp ite of the d i f f i c u l t i e s which t h i s
w ill cre a te fo r deduction in sociology, and in sp ite of the
p o p u la rity among s o c io lo g is ts of deductions such as have
been used.
I t was fu rth e r suggested th a t th e o r e tic a l progress
would be f a c i l i t a t e d by always m aintaining a c le a r d i s t i n c
tio n between p ro p o sitio n s which s ta te changes over time and
those which s ta te e x is tin g co n d itio n s in a c ro s s -s e c tio n of
population a t a given tim e.
B I B L I O G R A P H Y
I
BIBLIOGRAPHY
Books
B ossard, James H. The Large Family System. P h ila d e lp h ia :
U n iv e r s ity of P ennsylvania P r e s s , 1956.
Bridgeman, Percy W . The Logic of Modern P h y s ic s * New York:
The Macmillan Company, 1928.
C a rtw rig h t, Dorwin, and Zander, A lvin F. ( e d s . ) . Group
Dynamics: R esearch and Theory. Evanston: Row,
P e te rs o n and Company, 1955.
C hapin, F. S t u a r t . E xperim ental Design in S o c ia l R e se a rc h .
New York: Harper and B ro th e rs , 1947.
Durkheim, Em ile. The D iv isio n of Labor in S o c ie ty . Trans
l a te d by George Simpson. Glencoe: The Free P re s s ,
1947.
F e ig l, H e rb e rt, and Broadbeck, May. ( e d s . ) . Readings in the
P hilosophy of S c ie n c e . New York: Apple ton-Ce n tu ry -
C r o f t s , I n c ., 1953.
F e ig l, H e rb e rt, S c riv e n , M ichael, and Maxwell, G rover.
M innesota S tu d ie s in the Philosophy of S c ie n c e .
'Vol. I I of C oncepts. T h eo ries, and the Mind-Body
Problem . M inneapolis: U n iv e rs ity ol M innesota
P r e s s , 1958.
G ross, L lew ellyn ( e d . ) . Symposium on S o c io lo g ic a l Theory.
Evanston: Row, P e te rso n and Company, 1959.
G u ilfo rd , J . P . Fundamental S t a t i s t i c s in Psychology and
Education" New York: McGraw-Hill Book Company,
19^6.
Hare, A. P a u l, B ales, Robert F ., and B o rg a tta , Edgar F.
( e d s . ) . Small Groups. New York: A lfre d A. Knopf,
1955.
H o llin g sh e a d , August B. Elmtown' s Youth. New York: John
Wiley and Sons, 1949.
122
123
Homans, George C. The Human Group, New York: Harper and
B rothers, 1950.
Katz, E lihu, and L azarsfeld, Paul F. Personal In flu en ce.
Glencoe: The Free P ress, 1955.
Lerner, D aniel, and Lasswell, Harold T. ( e d s .) . The Policy
Sciences. Palo A lto; Stanford U niversity Press,
- m r .
McNemar, Quentin. Psychological S t a t i s t i c s . New York: John
Wiley and Sons, 1953.
Merton, Robert K. S ocial Theory and Social S tru c tu re .
Glencoe: The Free Press, 1957. — — — ——— — — —
Neurath, Otto (e d .). In te rn a tio n a l Encyclopedia of Unified
Science. Vol. II: Foundations of' the' Unity of
S cie n ce . Chicago: U n iv ersity of Chicago P ress,
1944.
Parsons, T a lc o tt. The S ocial System. Glencoe: The Free
P re ss, 1951.
________, and S h ils , Edward. Toward a General Theory of
A ction. Cambridge; Harvard U niversity P re ss, 1951.
Shubik, Martin (e d .). Readings in Game Theory and P o l i t i c a l
Behavior. Garden City: Doubleday and Company," 1954.
Swanson, Guy E ., Newcomb, Theodore M., and H artley, E. L.
( e d s .) . Readings in Social Psychology. New York:
Henry Holt and Company, 1952.
T a g u iri, Renato, and P e tru llo , Luigi (e d s .). Person Percep
t i on and In terp erso n al Behavior. Palo A lto: is'ta'n-
Tord U niversity Press', 196B.
Z etterb erg , Hans L. On Theory and V e rific a tio n in Sociol
ogy. New York: T re ssle r P ress, 1954.
A rtic le s in P e rio d ic a ls
B ie ri, J . "Changes in In terp erso n al Perception Following
S ocial In te ra c tio n ." Journal of Abnormal and S ocial
Psychology. XLVIII (January, 19b3), 61-66.
C o rsin i, Raymond J . "Understanding and S im ila rity in Mar
ria g e ," J o u r a a l o f j t o n ^ ^
LII (1956), 327-3^.
124
Cronbach, Lee J . "P rocesses A ffe c tin g Scores on 'Under
standing of O th e r s ' and 'Assumed S i m i l a r i t y , " '
P sy c h o lo g ic a l B u l l e t i n . L II (May, 1955), 177-93.
Dymond, R o salin d . " I n te r p e r s o n a l P e rc e p tio n and M a rita l
H appiness." Canadian J o u rn a l of Psychology. V III
(September, 1954), 164-711
F is h e r , Paul N. "An A nalysis of the Primary Group," Soci-
om etrv. XVI (August, 1953), 272-76.
G ibbs, Jack P ., and M artin, W alter T. "A Theory of S ta tu s
In te g r a tio n and I t s R e la tio n to .S u ic i d e ," American
S o c io lo g ic a l Review. XXIII ( A p ril, 1958), l4b-47.
Hare, A. P a u l. "A Study of I n te r a c tio n and Consensus in
D if fe re n t Sized G roups." American S o c io lo g ic a l Re-
view . XVII (June, 19S2J, -----
H a sto rf, A lb e rt H., Bender, E. I . , and W eintraub, D. J .
"The Influence of Response P a tte r n s on the 'Refined
Empathy S c o re , 1" J o u rn a l of Abnormal and S o c ia l
Psychology, LI (September, 1955), 341-43.
James, John. "A P re lim in a ry Study of the Size D eterm inants
in Small Group I n t e r a c t i o n ," American S o c io lo g ic a l
Review. XVI (August, 1 951),.476-77.
Newcomb, Theodore M. "The P re d ic tio n of I n te r p e r s o n a l At
t r a c t i o n , " American P s y c h o lo g is t. XI (November,
1956), 575-66"!
Rose, Arnold. "A Deductive Ideal-Type Method." American
J o u rn a l of S o cio lo g y . LVI (J u ly , 1950), 35-42.
Robinson, W . S. "The Geometric I n t e r p r e t a t i o n of Agree
ment." American S o c io lo g ic a l Review. XXIV (June,
1959), 336-45.
S tro d tb e c k , Fred L. "The Family as a Three-Person Group,"
American S o c io lo g ic a l Review. XIX (February, 1954),
23- 29.
Thomas, Edwin J . "Role Conception and O rg a n iz a tio n a l S iz e ."
American S o c io lo g ic a l Review. XXIV (February, 1959),
30-57.
W illiam s, Robbin M., J r . "C o n tin u ity and Change in Socio
l o g i c a l S tu d y ," American S o c io lo g ic a l Review. XXIV
(June, 1959), 336-45'.'
125
A rtic le s in C o llec tio n s
Bavelas, Alex. "Communication P a tte rn s in Task-Oriented
Groups," Group Dynamics: Research and Theory. Ed-
ite d by .Dorwin Cartwright and Alvin r . Zander.
Evanston: Row, Peterson and Company, 1953.
Bendix, R einhardt, and Berger, Bennett. "Images of Society
and Problems of Concept Form ation.in Sociology,"
Symposium on S o cio lo g ical Theory. Edited by
Llewellyn Gross. Evanston: Row, Peterson and Com
pany, 1959.
Bales, Robert F. "The Equilibrium Problem in Small Groups,"
Small Groups. Edited by A. Paul Hare, Robert F.
Bales, and Edgar F. B orgatta. New York: A lfred A.
Knopf, 1955.
Broadbeck, May. "Models, Meaning, and T heories," Symposium
on S ociological Theory. Edited by Llewellyn Gross.
Evanston: Row, Peterson and Company, 1959.
Campbell, Norman R. "The S tru ctu re of T heories," Readings
in the Philosophy of S cience. Edited by Herbert
iFeigl and May Broadbeck. New York: Appleton-Cen-
tu ry -C ro fts, In c ., 1953.
Cronbach, Lee J . "Proposals Leading to Analytic Treatment
of S o cial Perception Scores," Person Perception and
In terp erso n al Behavior. Edited by Renato T aguiri
and Luig'i P e'trullo. Palo A lto: Stanford U niversity
P re ss, 1958.
F ied le r, Fred E. "In terp erso n al P erception and Group Ef
fe c tiv e n e s s ," Person P erception and In terp erso n al
Behavior. Edited by Renato T aguiri and Luigi Pe-
t r u l l o . Palo A lto: Stanford U niversity P ress, 1958.
Gouldner, Alvin W . "R eciprocity and Autonomy in Functional
Theory." Symposium on S ocio lo g ical Theory. Edited
by Lie we l'ly n Gro s s . Evanston: Row, P eterson and
Company, 1959.
Gross, Llewellyn. "Theory C onstruction in Sociology: A
Methodological In q u iry ." Symposium on S ociological
Theory. Edited by Llewellyn Gross. Evanston: Row,
Peterson and Company, 1959.
Hempel, C a rl G. "Fundamentals of Concept Formation in
Empirical Science," In te rn a tio n a l Encyclopedia of
U nified Science. Vol. II: Foundations of the' Unity
126
of S cie n c e . No. 7. Chicago: U n iv e rsity of Chicago
P re s s , 1952.
_________ . "The T h e o r e tic ia n 's Dilemma," Minnesota S tu d ies
in the Philosophy of S cie n ce. Vol. I I : Concept s .
T heories, and the Mind-Body Problem. Edited Joy
H erbert F e ig l, Michale S criven, and Grover Maxwell.
M inneapolis: U n iv e rsity of Minnesota P re s s , 1958.
Hochberg, H erb ert. "Axiomatic Systems, F orm alizatio n , and
S c i e n t i f i c T h e o rie s . 1 1 Symposium on S o c io lo g ic a l The
o ry . E dited by Llewellyn G ross. Evanston: Row,
P e te rso n and Company, 1959.
L az arsfe Id , Paul F ., and B arton, A llen H. " Q u a lita tiv e
Measurement in the S o c ia l S ciences: C l a s s i f i c a t i o n ,
Typology, and In d ic e s ," The P o licy S cie n ces. Edited
by D aniel Lerner and Harold L. Lassw ell. Palo A lto:
Stanford U n iv e rsity P re ss, 1951.
L e a v itt, Harold J . "Some E ff e c ts of C e rta in Communication
Problems on Group Perform ance," Readings in S o cia l
Psychology, E dited by Guy E. Swanson, Theodore" M.
Newcomb, and E. L. H artle y . New York: Henry Holt
and Company, 1952.
Neurath, O tto. "Foundations of the S o c ia l S cie n ces," I n te r
n a tio n a l Encyclopedia of U n ified S c ie n c e . Vol. I I :
Foundations of the Unity of Science. No. 1.
Chicago: U n iv e rsity of Chicago P re ss, 1944.
Newcomb, Theodore M. "The C ognition of P ersons as Cogniz-
e r s , " Person P e rc e p tio n and In te rp e rs o n a l B ehavior.
E dited” by Renato Taguiri' and L uigi P e t r u l l o . [Palo
A lto : S tanford U n iv e rsity P re s s , 1958.
Reichenbach, Hans L. " P ro b a b ility Methods in S o c ia l Sci
en ce," The P o licy S c ie n c e s . E d ited by Daniel Lerner
and Harold L. L assw ell• Palo A lto: Stanford Univer
s i t y P re s s , 1951.
Sheldon, Richard C. "Some O bservations on Theory in S ocial
S c ie n c e ." Toward a G eneral,Theory of A ction. Edited
by T a lc o tt Parsons and Edward S h iis . Cambridge:
Harvard U n iv e rsity P re s s , 1951.
Simon, H erbert A. "A Forman Theory of I n te r a c tio n in S o cia l
Groups," Small Groups. E dited by A. Paul Hare,
Robert F. B ales, and Edgar F. B o rg atta. New York:
A lfred A. Knopf, 1955.
127
Von Neumann, John, and M orgenstern, Oskar. "The B asis of
Game Theory," Readings in Game Theory a n d _ P o litic a l
B ehavior. Edited by M artin S h u b i k .G a r d e n C ity :
Doubleday, 1954.
Woodger, J . H. "The Technique of Theory C o n stru ctio n ,"
I n te r n a tio n a l Encyclopedia of U n ified S c ie n c e. Vol.
I I : Foundations of the U nitv of S c ie n c e. No. 5.
Chicago: U n iv e rsity of Chicago P re s s , 1939.
Unpublished M a teria l
Hamblin, Richard L. "An Experim ental Study of the R e latio n
ship of Communication, Power R e la tio n sh ip s , S p e cial
iz a ti o n , and S o cia l Atmosphere to Group S iz e ."
Unpublished Ph.D. d i s s e r t a t i o n , U n iv e rsity of Michi
gan, 1955.
A P P E N D I X E S
APPENDIX A
INFORMATION SHEET
130
1 . Name
INFORM ATION SHEET
2 . D ate o f B ir th
3* A ddress P la ce o f B ir th
5 . How lo n g have you liv e d a t th e p r e se n t address?
6 . How lo n g have you liv e d in t h i s sc h o o l d is tr ic t? ^
7 . F a th e r 's o ccu p a tio n _____________________ ^______
8 . M other’s o ccu p a tio n
9 . Your fu tu r e occupation_
1 0 . R e lig io u s p referen ce
1 1 . Your grade in sc h o o l in Septem ber, 1959
1 2 . Were you employed during th e sc h o o l year?________
1 3 . Are you employed now?__________ How many hours p er week?
How many hours per week?
1 4 . Do you own youi own car? 1 5 . Do you d r iv e th e fa m ily car?
1 6 . P le a se l i s t below th e e x tr a -c u r r ic u la r a c t i v i t i e s you have b een in a t sc h o o l t h i s
y e a r (c lu b s , com m ittees, team s, e t c . )
1 7 . In column ( l ) b elo w , l i s t th e names o f each o f th e p erso n s in t h i s group NOT in c lu d ii
y o u r s e lf . In column ( 2 ) s t a t e how many y e a r s you have known ea ch p erso n . ( £n column (3 )
e stim a te how many hours you sp e n t w ith each person d u rin g th e p a s t w eek, in column (4 )
e s tim a te th e number o f hours in a t y p ic a l sc h o o l week you sp en t w ith each p erso n . For
ea ch person cou n t th e tim e you are w ith him in th e p resen ce o f o th e r s a s w e ll a s tim e yoi
a r e w ith him and no one e l s e .
(1) (2) (3 ) (4)
Name How lo n g known
(Y rs. & M onths)
How many h r e . w ith
him l a s t week?
How many hr s . w ith
him a v g . sc h o o l wee' .
18 . Check below the a c t i v it i e s you u su a lly engage in w ith the members o f t h is group,
double d ates _ _ _ _ _ C lasses a t sch ool _ Go home from sch ool to g eth er
homework V is it a t home "Hang around" to g eth er____________
Lunch Go to beach Other _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
shows ______ _ _ D rive around
19* L is t below the names o f persons o f your sex whom you have run around w ith in the
p ast year, but whom you no lon ger run around w ith . A fter each name in d ica te why you do
not run around w ith t h is person any more.
20. L is t below the names o f persons you have dated in the p ast s ix months.
21. Use the space below to w rite anything you think w ill help a d u lts to understand
b e tte r the teenagers you run around w ith .
APPENDIX B
QUESTIONNAIRE
I
132
SECTION I .
On each o f th e statem en ts below , c ir c le th e l e t t e r in fr o n t o f th e answer you
most c lo s e ly agree w ith . Do n ot mark more than one answer fo r any q u estio n .
1 . Which statem ent ex p resses your op in ion about smoking fo r teenagers?
a) I t i s p e r fe c tly O K fo r teen a g ers to smoke i f th ey want t o .
b) Anyone who d o e sn 't smoke i s a square.
c ) I t i s O K fo r boys to smoke, b u t n ot fo r g i r l s .
d) No teen ager should smoke.
e) I t i s O K fo r teen a g ers to smoke a f t e r th ey graduate from h igh s c h o o l.
2 . Which ex p resses your id ea on d rin k in g fo r teenagers?
a) Teenagers should not drink a t a l l .
b} I t i s O K fo r boys to d rin k , b u t not fo r g i r l s .
c ; I t i s O K fo r teen a g ers to d rin k , b u t n ot t o e x c e s s .
d) D rinking i s up to each p erso n s, and each teen a g er should d ecid e fo r h im seli
e ) One drink once in a w h ile i s n o t wrong, but no more.
3 . Which ex p resses your id ea on necking fo r teenagers?
a) A good -n igh t k is s i s OK, but necking i s a bad id ea fo r anyone.
b) A good -nigh t k is s i s O K w ith someone you r e a lly l i k e , b u t t h a t 's a l l .
c ) Necking i s O K on ly fo r cou p les who are going stea d y .
d) Necking i s O K w ith anybody, ju s t so th ey l i k e each o th er.
e ) Necking i s O K w ith anybody.
1 + . Which most a c c u r a tely ex p resses your op in ion on sch o o l woi’k?
a) S o c ia l a c t i v i t i e s are a n . th a t count fo r te e n a g e r s. School work d o e sn 't
r e a lly m atter much.
b) S o c ia l a c t i v i t i e s are th e most im portant th in g fo r tee n a g e rs, b u t sch o o l
work should be kep t up i f i t d o e s n 't in te r fe r e w ith having fu n.
c ) You should tr y to s tr ik e a b alan ce betw een s o c ia l a c t i v i t i e s and reason
ab ly good sch o o l work.
d) School work should come f i r s t , and you should do w e ll in sch o o l even i f
i t means c u ttin g down on s o c ia l a c t i v i t i e s and fu n .
e ) You should do your b e s t in s c h o o l, even i f i t means p r a c t ic a lly no s o c ia l
l i f e and r e c r e a tio n .
5 . Which statem ent m ost c lo s e ly e x p r esses your a ttitu d e toward r e lig io n ?
a ) R e lig io n should be th e most im portant th in g in a te e n a g e r 's l i f e .
b) R e lig io n should be one o f th e most im portant th in g s in a tee n a g e r ’s l i f e .
c ) R e lig io n and church are fin e fo r te e n a g e r s. Every teen ager should go
to church or Sunday sc h o o l.
d) Church i s O K fo r th o se who lik e i t , but i t d o e sn 't make much d iffe r e n c e
whether you go or n o t.
6 . Which statem ent on allow ances i s c lo s e s t to your opinion?
a) Every teenager i s e n title d to a regu lar allowance from h is p aren ts.
b) Teenagers are e n t it le d to an allowance only i f they perform certa in
d u ties a t home.
c) Teenagers are e n title d to money from th e ir parents only fo r sp e c ia l th in gs
th ey need or fo r s p e c ia l occa sio n s.
d) Teenagers should be required to earn any spending money th ey need through
part-tim e work, so th a t they d o n 't depend on th e ir parents fo r any.
7. A l6 -y e a r old boy wants a car o f h is own. His parents are w e ll enough o f f to
buy him one. What should they do?
a) They should buy him a ca r.
b) They should make him earn h a lf the c o st o f the car, and they should
pay h a lf.
c ) They should a llo w him to buy the ca r, but make him earn the money to pay
fo r i t h im self.
d) They should not l e t him buy a car a t a l l .
8 . When i s i t a l l r ig h t for a g i r l to ask a boy for a date?
a) Whenever a g i r l wants t o .
b) Only when the date i s fo r a party bein g given by the g i r l s .
c) Only when they are going stead y.
d) I t i s never a l l rig h t fo r a g i r l to ask a boy for a d a te.
9 . When i s i s a l l rig h t fo r a g i r l to pay date expenses?
a) Only when she in v ite d the boy.
b) Only when they are going stead y.
c) Only when they are going steady and she has more money than he does.
d) Any time she f e e ls lik e i t .
e ) I t i s never a l l r ig h t for a g i r l to pay date exp en ses.
10. Which statem ent most c lo s e ly exp resses your opinion on d iscu ssio n w ith parents?
a) Parents d o n 't understand teen agers, so there i s no sense in t e llin g them
anything you d on 't have t o .
b) You should keep parents informed on your a c t i v i t i e s , but there i s no use in
d isc u ssin g your problems w ith them, because they wouldn't understand.
c) You should keep parents up on what you are doing, and co n su lt them for
advice on some kinds o f problems.
d) You should d iscu ss a ll your problems w ith your p aren ts, because they are
your b e s t source o f good ad vice.
11. A seventeen-year old g i r l should never date a man any older than
a)
27
years old
to) 25 years old
c) 23
years old
a)
21 years old
e) 19
years old
2
12. What do you think is the youngest age at which teenagers should go steady?
a) 17
b) 1 6
c) 15
d) 1U
e) 13
f) They should not go steady at a ll.
13. A boy and a date have gone to a party where everyone i s drinking in to x ic a tin g
b everages. What should th ey do?
a) Join in the d rinking.
b) Take one drink ju s t to be p o lit e .
c ) Turn down th e drinks and ask fo r something e l s e .
d) Leave a t th e e a r lie s t opportunity w ithout b ein g n o tic e d .
e ) T e ll the h o st you are le a v in g because you d o n 't drink.
14. S ev era l h ig h -sch o o l g i r l s are planning to ren t a beach house during v a ca tio n .
Do you th in k th ey should have a chaperone?
a) No, th ey do n ot need a chaperone.
b) Y es, but someone who i s not too s t r i c t .
c) Only i f they are going to have boys v i s i t them.
d) Y es, p refera b ly one o f t h e ir m others.
15. One o f th e g i r l s in a group o f very c lo se frien d s has begun d atin g a boy
w ith a bad rep u ta tio n . What do you th in k her frien d s should do?
a) N othing. I t sh ou ld n 't make any d iffe r e n c e to them.
b) Try to persuade her to drop him.
c) Stop running around w ith her u n t il she drops him.
d) Encourage her to b rin g him on a double-date w ith them.
1 6 . You are in a car w ith your fr ie n d s , and th ey decide to race w ith another
ca r. What would you do?
a) Ask to g e t out im m ediately.
b) Keep q u iet and see what happens.
c) Encourage the d riv er to go to i t .
d) J u st enjoy y o u r s e lf.
17. How do you f e e l about teen agers u sin g profane language?
a) No teenager should ever use p r o fa n ity .
b) P rofan ity i s O K when you are not in mixed company, and when you are not
around a d u lts .
c) P rofan ity i s O K under some circum stances fo r b oys, but not fo r g i r l s .
d) P rofan ity never hurt anyone, and i t d o e sn 't make any d iffe r e n c e whether
you use profane language or n o t.
3
SECTION I I .
On each o f the follow in g q u estion s, c ir c le the X in fron t o f the one choice
you agree with most, and c ir c le the 0 in fro n t o f the one ch oice you agree with
l e a s t . -
18. I f you were the parent o f a teenaged daughter, what behavior on her part would
embarrass you the most?
X 0 a) Dope ad d iction .
X 0 b) H abitual drunkenness.
X 0 c) Unmarried pregnancy.
X 0 d) S te a lin g reg u la rly .
19. I f you were the parent o f a teenaged son, what behavior on h is part would
embarrass you the most?
X 0 a) Dope ad d iction .
X 0 b) Robbery.
X 0 c) H abitual drunkenness.
X 0 d) Sex o ffe n se s.
20. A g ir l w ith a date i s a t a party a t the home o f a frien d . Her date is g ettin g
drunk. What should she do?
X 0 a) Try to understand h is problem, and l e t him g et drunk i f he must.
X 0 b) Try to g et her date to stop drinking and sober up.
X 0 c) Forget about her d a te, lea v e with other frien d s, and never date him again.
X 0 d) Ju st not worry about i t one way or the other, but go on and have fun.
21. A boy i s picked up by the p o lic e on susp icion o f having s e t f ir e to the sch ool.
His b e st frien d s are c a lle d in and questioned. They know he is g u ilty . What
should they do?
X 0 a) Say they know nothing, or refu se to ta lk .
X 0 b) Make up an a lib i for him.
X 0 c) Answer questions they are asked and nothing more.
X 0 d) Try to help the p o lic e by t e llin g everything they know.
22. Which o f th ese does most to make a boy popular with the g ir ls?
X 0 a) A sharp car.
X 0 b) A th le tic a b ilit y .
X 0 c) Good lo o k s.
X 0 d) Good manners.
23
• Which one
X 0 a) Nice
X 0 b) Good
X 0 c)
Good
X 0 d) Good
o f th ese does most to make a g i r l popular w ith the boys?
c lo th e s .
k
X 0
X 0
X 0
X 0
25
•
2 * 4 -. A boy and g i r l want to g e t married im m ediately. They have juBt graduated
from high sch o o l. T heir parents t e l l them th a t once they are m arried, th ey
w i l l have to support th em selves. They boy and the g i r l both want to go to
c o lle g e . What should they do?
a) Get married now, and have the w ife work to support them w hile the husband
goes to c o lle g e .
b) Get married now, go to work, and take c o lle g e p art-tim e a t n ig h t.
c) Forget ghaut c o lle g e . Get married and go to work.
d) Go to c o lle g e and g et married s e c r e tly .
You are the parent o f a teenaged boy who has begun to run around w ith a group
o f boys who have spent tim e in Ju ven ile H all fo r minor o ffe n s e s . What should
you do?
X 0 a) N othing, because t h is i s h is b u sin e ss.
X 0 b) Encourage him to b rin g h is new frien d s home with him.
X 0 c) Warn him about what people w i l l th in k o f him.
X 0 d) Forbid him to a s so c ia te w ith them.
26. A l6-year old g irl will graduate from high school in two years. She is very
much in lo v e w ith a boy who wants to marry h er. He t e l l s h er, " If you love
me, you w i l l marry me now. I f we d o n 't g e t married now, i t ' s q u its." What
should she do?
X 0 a) ;iu it sch ool and marry him.
X 0 b) Marry him s e c r e tly and sta y in sch o o l.
X 0 c) Marry him openly and sta y in sch o o l.
X 0 d) F in ish high sch ool and fo r g et about him.
27. A 17-year o ld g i r l i s in lo v e w ith an unhappily-m arried man whom she i s s e c r e t!
see in g re g u la r ly . He wants to continue in th is way. What should she do?
X 0 a) Stop se e in g him im m ediately, and fo r g et him.
X 0 b) Have a h e a r t-to -h ea r t ta lk w ith h is w ife .
X 0 c) Have him g e t a d ivorce and marry him.
X 0 d) Continue to see him s e c r e tly .
28. A ll your frie n d s have begun to smoke. Your parents have forbidden you to
smoke. What would you do?
X 0 a) Not smoke, but continue running around w ith your fr ie n d s.
X 0 b) D iscu ss i t w ith your parents and abide by whatever d e c isio n they reach,
X 0 c) Begin smoking and t e l l your parents th a t you have.
X 0 d) Give up running around w ith th o se fr ie n d s.
29. Your parents o b ject to the person you have been d atin g on the grounds th a t
L s person i s n ot good enough fo r you. What would you do?
Stop d a tin g th e person.
Have your date come to your house and t a lk i t over w ith your p a ren ts.
Continue d atin g th e person, but hide the fa c t from your p aren ts.
Continue d a tin g the person w ith your p a r en ts’ knowledge, and l e t them do
whatever th ey want t o .
X 0 a)
X 0 b)
X 0 c)
X 0
a)
5
I
X 0
a)
X 0 b)
X 0 c)
X 0
a)
X 0
e)
30. You lea rn th a t your b e s t frien d s have begun to smoke marihuana, what would
you do?
X 0 a) Not smoke marihuana, but continue running around w ith your fr ie n d s.
X 0 b) Stop running around w ith them, but keep what th ey are doing a s e c r e t.
X 0 c) Stop running around w ith them, and rep ort them to the p o lic e .
X 0 d) Try marihuana once, ju s t to see what i t i s l i k e .
31. Your new sch ool i s b ein g planned, and th ere are funds fo r on ly one o f the
fo llo w in g item s. As a stu d en t, which would you v o te to have constructed?
X 0 a ) Gymnasium.
X 0 b) Auditorium.
X 0 c) Swimming p o o l.
X 0 d) L ibrary.
32. You have won a quiz show, and you have your ch oice o f the fo llo w in g p r iz e s ,
t x worth about $3000. Which would you choose?
A new ca r.
A t r ip to Europe.
A c o lle g e sch o la rsh ip .
A $3000 spending account a t th e department sto r e o f your ch o ic e.
B u ild in g a recrea tio n cen ter in your community w ith th e money.
33• Which o f th e se a c ts on th e p art o f your frien d s o f the same sex would be most
li k e l y to make you sto p running around w ith them?
X O a) C r itic iz in g you behind your back.
X 0 b) Taking your stead y date away from you.
X 0 c ) T e llin g you l i e s .
X 0 d) Taking your b elon gin gs w ithout your knowledge.
3U. What behavior o f your parents in fro n t o f your fr ie n d s would embarrass you
th e most?
X 0 a) quarreling w ith each o th er.
X 0 b) C r it ic iz in g you.
X 0 c ) Being drunk.
X 0 d) Bragging to them about what a fin e person you a re.
35. Which p la ce i s th e most fun fo r dates?
X 0 a) M ovies.
X 0 b) Dances.
X 0 c ) Beach.
X 0 d) P a r ties a t someone's home.
3 6 . Which would you con sid er the g r e a te st m isfortune th a t could happen to you?
X 0 a) Being unable t o do p assin g sch o o l work.
X 0 b) Being unpopular and d is lik e d by other boys and g i r l s your age.
X 0 c) Being p h y sic a lly d isa b le d .
X 0 d) B eing so poor th a t you could n ot a ffo rd decen t c lo th e s or a n ice home.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Economic Differentiation And Social Organization Of Standard Metropolitanareas In The United States: 1950

PDF

Normative values of selected law enforcement officers and adult male offenders

PDF

Consensus Of Role Perceptions In A Welfare Planning Council

PDF

Reference Group Theory, Selection, And The Images Of Professions

PDF

The Effect Of Differential Treatment On Attitudes, Personality Traits, And Behavior Of Adult Parolees

PDF

Social Components Of Housing Cost In The Western Metropolis

PDF

Attitudes Of Ministers And Lay Leaders Of The American Baptist Conventionof The State Of Washington On Selected Social Issues

PDF

Ecology Of Negro Communities In Los Angeles County: 1940-1959

PDF

Referential Dissociation And Response To Stress

PDF

An Empirical Examination Of The Relationship Of Vertical Occupational Mobility And Horizontal Residential Mobility

PDF

Status Consistency Among The Clergy

PDF

Delinquency And Rationalization: A Study Of The Delinquent Act

PDF

Interpersonal Relations In Ethnically Mixed Small Work Groups

PDF

Differential Fertility Behavior And Values In Rural And Semi-Urban Costa Rica

PDF

Selected Social Psychological Factors Related To Viewers Of Television Programs

PDF

Role Expectations Of American Undergraduate College Women In A Western Coeducational Institution

PDF

Familism, Suburbanization, And Residential Mobility In A Metropolis

PDF

The Selfish Self: A Social Psychological Study Of Social Character

PDF

Social Changes In Selected Institutions Of The Ussr With Special Reference To The Family

PDF

The Career Business Executive As A Definitive Occupational Type

Asset Metadata

Creator Udry, Joe Richard (author)

Core Title The Construction And Empirical Test Of A Theory Based On Selected Variables In Small-Group Interaction

Degree Doctor of Philosophy

Degree Program Sociology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag OAI-PMH Harvest,Sociology, general

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Locke, Harvey J. (committee chair), Sabagh, Georges (committee member), Thorpe, Louis P. (committee member), Van Arsdol, Maurice D., Jr. (committee member), Vincent, Melvin J. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-77408

Unique identifier UC11357977

Identifier 6004489.pdf (filename),usctheses-c18-77408 (legacy record id)

Legacy Identifier 6004489.pdf

Dmrecord 77408

Document Type Dissertation

Rights Udry, Joe Richard

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA