University of Southern California Dissertations and Theses

The Effects Of Pretraining In Auditory And Visual Discrimination On Texting In First Grade Boys

(USC Thesis Other)

The Effects Of Pretraining In Auditory And Visual Discrimination On Texting In First Grade Boys

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content This dissertation has been 64—13,490
microfilmed exactly as received
CATTERALL, Calvin Dennis, 1925-
THE EFFECTS OF PRETRAINING IN AUDITORY
AND VISUAL DISCRIMINATION ON TEXTING IN
FIRST GRADE BOYS.
University of Southern California, Ph.D., 1964
Education, psychology
U n iversity M icrofilm s, Inc., A n n Arbor, M ich igan
TEE EFFECTS OF PRETRAINING IN AUDITORY AND VISUAL
DISCRIMINATION ON TEXTING IN FIRST GRADE BOYS
by
Calvin Dennis Catterall
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Educational Psychology)
June 196
UNIVERSITY O F S O U T H E R N CA LIFO RN IA
G R A D U A T E S C H O O L
U N IV E R S I T Y PA R K
L O S A N G E L E S 7. C A L I F O R N I A
This dissertation, written by
.....................C a l v i n . D e n n is .. C a tt e r.al 1.........................
under the direction of his.....Dissertation C o m
mittee, and approved by all its mem bers, has
been presented to and accepted by the Graduate
School, in partial fulfillment of requirements
fo r the degree of
D O C T O R O F P H I L O S O P H Y
Dean
Da^.....J.uneJ l..1 .9 6 4 ..........................................
DISSERTATION CO M M ITTEE
....
Chairman
TABLE OF CONTENTS
Chapter
I. PROBLEM
Page
X
Introduction
Purpose of Present Study
Scope and Limitations
Organization of the Chapters
II. LITERATURE 8
Review of the Literature
Summary and Critique of the Literature
with Special Reference to Reading
III. PROCEDURE
Purpose
Subjects
Stimulus Material
Discrimination Pretraining
Familiarization
Texting Training
Inquiry
IV. RESULTS................................. 57
Equivalence of Groups
Stimulus Words
Discrimination Pretraining
Texting Training
Correct Texting Responses per Word
Verbal Report of Discrimination Cue
V. SUMMARY AND CONCLUSIONS................. 7^
Summary
Discussion and Conclusions
BIBLIOGRAPHY................................... 86
ii
LIST OF TABLES
Table Page
1. Mean and SD of Verbal IQ and CA (in months)
for Pretraining Groups (N * 15).......... 58
2. Analyses of Variance of CA (in months) and
Verbal I Q ............................... 58
3. Analysis of Variance of Correct Textual
Responses for Six Pretraining Groups . .. 61
b . RHOS between Approximated Verbal IQ and
Total Correct Texting Responses .......... 62
5. Correct Texting Responses by Stimulus Word • - 71
6. Analysis of Variance of Mean Number of
Correct Texting Responses for the
Four Stimulus Words..................... 71
7- Cues Ss Reported Having Used in Identifying
Words................................... 72
iii
Figure
1.
LIST OF FIGURES
Correct Texting Responses for Eight
Discrimination Pretraining Groups . • .
Page
6 5
iv
CHAPTER I
PROBLEM
Introduction
A substantial portion of the national effort in edu
cation is expended in helping individuals learn to read and
gain information from the printed -word. The importance of
this skill in a culture characterized by a rapid advance in
all areas of knowledge needs no defense. It is probably
not possible to attain a precise science of reading. Im
provements can unquestionably be made on the techniques used
in teaching reading, and valuable new ones can be devised.
Progress must be made in the description of the total read
ing process and in more accurate prediction of reading pro
gress. Before this can happen the major part of the vari
ance associated with instruction and measurement in reading
must be accounted for.
The specific types of intelligence which influence
reading have not been completely documented. The generali
zation that reading requires a mental age of six-and-a-half
years is frequently used but not all of the factors are
known which are operating in beginning reading (Holmes,
1
2
1962). There Is substantial interest in teaching children
to read at a lower chronological and mental age (Rowan,
1961; Durkin, 1962; Fowler, 1 9 6 2). The common practice
of basing measures of expectancy for reading primarily on IQ
should be questioned. More information is needed to help
answer the following questions: What is the relationship
between the ability to read and to communicate effectively?
What part does reading play in helping the individual solve
life’s increasingly complex problems? What is the inter
action between reading, thinking, and general intelligence?
Auditory perception is basic to the process of
speaking and reading in the normally developing child. How
the child associates speech sounds with the written word has
not been clearly defined. The issue is no longer whether
or not to teach phonics. The decisions that need to be
made are: when should phonetic instruction be introduced;
what form should it take; and what type of child will bene
fit the most?
Reading is, at least in part, a complex visual task.
More needs to be known about the role of visual perception,
especially in the early stages of reading. What cues does
the child naturally use and what cues can he be trained to
use to make the visual discrimination necessary in beginning
reading? Measurement of visual perception has been compli
cated by a contamination of errors. A clearly defined dis
tinction, both in theory and in techniques of measurement,
needs to be made between sensory input, the perceptual-
integrational process, and the final motor output. More
information is needed to tell us how perceptual skills can
be developed in the young child.
Whereas both auditory and visual perception are com
plex enough when considered separately, complete description
and prediction in reading awaits an understanding of total
perceptual integration. This integration needs to be more
precisely defined in the developing child as it is affected
by the social-emotional and cultural milieu.
A wide variety of techniques are used to teach read
ing. Changes of emphasis and technique are evident in all
areas of curriculum. New approaches in such areas as
arithmetic and science are usually based on discoveries and
changes in the related field. Instructional innovations in
reading are usually not so clearly associated with research.
The relative effectiveness of the whole-word versus phonetic
approach continues to be debated. Application to the read
ing process of the theories and research techniques associ
ated with operant learning holds promise for providing much
needed answers.
Purpose of Present Study
The purpose of the present study was to analyze the
process under which the beginning reader brings the appro
priate speech response (the spoken word) under the control
i f
of the visually presented verbal stimuli (the printed word).
Skinner (1957) calls this process "texting." The experi
mental design used operant conditioning techniques to inves
tigate the effects of pretraining in two major sense modali
ties. Eight groups were given discrimination pretraining,
familiarization, and texting training on four words. Each
word contained only one vowel and an ascending or descending
configurational cue.
In making texting discriminations children have fre
quently reported using ascending or descending configura
tional cues. For ease of discussion consonants containing
these cues have been referred to as configurational cues.
It was felt that additional information was needed about the
relative values of emphasizing these configurational cues as
opposed to vowel cues in the early texting process. Thus
the visual discrimination pretraining on the vowel cues was
varied in that the Ss* attention was directed to (a) the
consonant-ccnfiguration cue or the (b) vowel cue.
To provide further information about the role of
auditory perception in texting, auditory discrimination pre
training was varied in three ways: (a) the letter-name of
the configurational or vowel cue was given; (b) the letter-
sound of the configurational or vowel cue was given, or (c)
the configurational or vowel cue was visually emphasized
without auditory cueing. The emphasis of many reading pro
grams on early phonic instruction points to the need for
further research in this area. Because of the importance
of interaction between auditory and visual perception, the
integration of these two sense modalities in the reading
task was also investigated.
Scope and Limitations
Since boys have traditionally demonstrated more
difficulty with beginning reading than girls, the population
was delimited to first grade boys with average ability.
The study was also designed to provide information about the
application of reinforcement theory to the beginning texting
response in young boys. Whenever possible procedures were
used which would facilitate application of possible findings
to the school setting. For example, actual words were used
rather than nonsense syllables. Stimuli were typed on
plain white cards and reinforcement was adapted to the needs
of the individual. The Ss responded verbally rather than
pushing a lever, etc.
The findings of the study are limited to the early
stages of reading. Other visual and auditory factors may
be operating at other stages in this complex task. The
data directly apply only to helping the child make the cor
rect texting response, and have only limited application to
the even more important process of helping the child gain
meaning from what he reads. Since all of the Ss were boys,
the results must subsequently be validated with girls. Ss
were seen one at a time; the results and techniques would
have to be adapted to be used with a group in the classroom.
Restricting the population to boys in the middle range of
ability places further limitations on the applicability of
the material to other ability levels.
Organization of the Chanters
Chapter II presents a review and critique of the
literature. The first part of the review briefly summa
rizes recent studies using operant techniques with children.
This brief methodological review has been provided as a
rationale for the controls and procedures used in the pre
sent study. Research designs most directly applicable to
early reading are discussed in greater detail.
The second part of Chapter II briefly summarizes the
literature for both the methodological and content vari
ables. A critique of the literature identifies the need
for further research.
Chapter III defines the purpose of the study with
special reference to the issues which have been raised in
the literature. The hypotheses to be tested are stated
followed by a description of the procedure used in the
study.
Chapter IV describes the statistical treatment of
the data and the results of the study. Chapter V briefly
summarizes the purpose, procedure, and findings of the
7
study. The results are discussed and implications for edu
cational practice and further research are drawn.
CHAPTER II
LITERATURE
Review of the Literature
The review of the literature has been, divided into
several parts. Consideration had to be given to the meth
odological procedures and techniques applicable to the prob
lem and population studied, A brief history and definition
of operant conditioning is provided. A review of recent
reinforcement studies with children has been included be
cause of the importance of this area to operant theory and
procedure. Since the study is concerned with the emerging
function of reading, selected developmental studies related
to this topic have been summarized. Because Skinner's work
and operant learning have influenced programmed learning and
auto-instructional devices, a few studies have been cited in
this area. Sex differences in reading indicated the need
for including studies on factors related to the sex variable.
The relationship between texting and speech led to a section
on verbal conditioning. The last part of the literature
review discusses discrimination studies with children.
Discrimination studies directly relating to reading complete
the review.
In brief, the literature has been summarized under
the following headings:
Definition of Operant Conditioning
Amount and Type of Reinforcement
Schedules of Reinforcement
Immediate versus Delayed Reinforcement
Secondary Reinforcement
Aversive Conditioning
Developmental Studies Related to Reading
Programmed Learning and Auto-Instructional Devices
Sex Differences and Sex Role Identification
Verbal Conditioning
Di sc r iminat ion
Discrimination as it Applies to Reading
Definition of Operant Conditioning
Operant (Type II) conditioning should be distin
guished from the classical Pavolvian (Type I) paradigm. In
the latter, the conditioned stimulus is presented in close
temporal proximity with the unconditioned stimulus until the
response, previously associated with the US, is evoked by the
CS. It is generally agreed that this paradigm will not ex
plain the major part of behavior accounted for by condition
ing. The distinction between this and operant or instru
mental (Type II) conditioning was originally made by Skinner
10
(1938, 1953, 1957, 1959). English & English (1998) provide
the following definition of operant conditionings
The complex of organismic processes involved in the
experimental, procedure (or the procedure itself)
■wherein a stimulus, having evoked a response that
brings into view a rewarding stimulus, thereafter is
more likely to evoke that response.
Much of human behavior would appear to be constituted by
what Skinner (1938) calls operant responses. Operants cor
respond to "voluntary responses" in that they seem to be
emitted, rather than elicited by specific stimuli. Of the
many responses available to the individual, only one is cor
rect. What is learned is the emittance of the new response
to the old stimulus.
Hively (1959) pointed out that although Skinner*s
work has often been referred to as a learning theory, this
is not technically true. Rather, it is a functional analy
sis aimed at discovering relationships between dependent
variables in the behavior of organisms and the independent
variables in their environment. One of Skinner* s major
contributions has been his emphasis on a detailed behavioral
analysis of the responses to be acquired and the programming
of stimuli to shape and maintain behavior.
Operant techniques, i.e., experimental procedures
for increasing or decreasing the probability of specific
behavior, have long been employed with infrahuman organisms..
Hilgard (1956) stressed the continuing need for validating
experiments with human subjects in real life situations.
11
A 1 1 1 1 1 1 1 1 1 6 1 of. studies have been reported on operant learning
with adults. More recently, a few investigators have em
ployed these techniques with children. For example,
Longstreth (1962) used preschool children in a discrimina
tion design involving a runway similar to the mazes tradi
tionally used with rats. His results, in which faster
speeds were elicited when the stimulus was paired with
reward, substantiate previous findings with infrahuman Ss.
The sections which follow summarize the current, pertinent
literature on operant techniques with children.
Amount and Type of Reinforcement
Hively (1959) suggested that there is growing evi
dence that some sort of contingency of reinforcement is
basic to most of that change in behavior of all organisms
which we call learning. The literature on factors affect
ing reinforcement in children has been organized as followss
amount and type of reinforcement, immediate versus delayed
reinforcement, schedules of reinforcement, secondary rein
forcement, and aversive conditioning.
Terrell (195$) studied the effectiveness of token
incentive, immediate reward, and promised reward on
and 9 year old children on a size discrimination task. He
suggested that his results showing more trials required for
delayed incentive, may have differed from earlier findings
because of (a) a difference in the type of population and
12
(b) the task may have been interesting enough to negate the
value of promises. Using 9 year old boys, Miller and Estes
(1961) investigated the effects of different levels of mone
tary incentive ($.01 and $.5 0) and W achievement upon the
acquisition of a visual discrimination habit. The results
of previous studies were substantiated by the finding of no
difference between the two degrees of incentive. An unex
pected finding was the clear inferiority of the reward
groups. This may have been accounted for by a tendency on
the part of the Ss to become preoccupied with the money.
Bijou & Sturges (1959) summarized Information per
taining to the use of consumables and manipulables as posi
tive reinforcers in experimental studies with children.
They pointed out that the only way to find out whether or
not an object is a positive reinforcer, i.e., stimuli which
strengthen the behavior followed, is to submit it to an ex
perimental test. Brackbill & Jack (1958) reported the use
of individually determined reinforcers (the child chooses
toy, candy, etc.) on a simple discrimination task. These
findings suggest that this technique will reduce error vari
ance frequently found in analysis of variance designs with
children. The investigators feel that this is accounted
for by S’s increased motivation and report better discrimi
nation responses with a combined reward-punishment tech
nique.
Campbell (1961) varied the amount and type of rein
13
forcement (using fixed ratio, fixed interval, and variable
interval schedules) with preschool children and obtained
records similar to those generated by lower organisms, but
with more variability. He found that alteration in the
amount or type of reinforcement produced rate increases on
a simple lever pulling task. His work indicated that the
presence of the E in the cubicle seemed to suppress the
rate and proportion of response strength.
Hicks (1962) studied the effects of various social
reinforcements on the achievement behavior of male kinder
garten children. On a model building task, he found thats
(a) Positive reinforcement is significantly superior to
negative reinforcement; (b) intermittent and continuous
reinforcement have approximately the same effect with a
slight difference in favor of continuous reinforcement;
(c) punishment produces a significant reduction in the
amount of achievement behavior evoked; (d) intermittent
reinforcement produces approximately the same amount of
achievement behavior as continuous reinforcement, but takes
significantly less reinforcement to be elicited. Since he
had hypothesized that continuous reinforcement would have
less value than intermittent reinforcement, Hicks suggested
that any positive reinforcement is sufficient if the task
is interesting enough to the S. It would appear from
these data that continuous positive reinforcement is the
recommended procedure during the initial stage of training.
lb
Findings from the control group indicated that the lack of
overt social reinforcement had a negative value for some of
the Ss and a positive value for others.
In an extension of a study of verbal reinforcement
by Buss & Buss (1956), Curry (i 9 6 0) found, with fifth and
sixth grade children, that verbal reinforcement combina
tions of Right-Wrong and Nothing-Wrong led to faster learn
ing than did Right-Nothing. Since her findings are not in
agreement with Terrell & Kennedy (1957), who had included a
light flash along with the verbal reinforcement, Curry
suggested that the differential effectiveness of the verbal
reinforcement combinations may depend upon the absence of
other information about the correctness of a response.
Schedules of Reinforcement
Studies dealing with schedules of reinforcement have
been of maoor importance in operant conditioning studies on
infrahumans. Fattu, Auble, & Mech (1955) were among the
first to investigate partial reinforcement with preschool
children. Bijou (1955) found that a variable ratio inter
mittent distribution was more resistant to extinction than a
continuous schedule with a small group of nursery school
children. Long, Hammack, May, & Campbell (1958) investiga
ted fixed ratio, fixed interval, and variable interval
schedules with 200 children from ^ to 8 years. Children
under 3 were not successful Ss because they refused to
15
remain alone in experimental cubicles for more than a few
moments. Working with small children necessitated that the
amount of time be kept short, which tended to prevent the
schedules from having an opportunity to start controlling.
Nonetheless, they found almost as much experimental control
could be gained over the behavior of children as that of
lower organisms. Changes from a fixed ratio to intermit
tent reinforcement seemed to be more difficult to establish
in children than in animals. The results of Meyers (i 9 6 0)
with preschool Ss (ages 3-5 to 6-3) supported previous find
ings that partial secondary reinforcement during condition
ing produces greater resistance to extinction than continu
ous secondary reinforcement.
immediate versus Delayed Reinforcement
Studies of delay of reward have produced fairly con
sistent results with animals, but not with children.
Lipsitt, Castaneda, & Kemble (1959) reported that children
(*+-2 to 5 - 8 years) responded significantly more slowly to a
light stimulus associated with delay than to one associated
with an immediate reward. In another experiment Lipsitt &
Castaneda (1958) found no significant difference in the qpeed
of response to the two lights. When the children were
given a choice as to which stimulus was to be presented, the
immediately reinforced one was chosen significantly more
often. Rieber (1961) concluded that the major effect of
16
delay of reward on speed of response with kindergarten chil
dren was to facilitate the association of competing re
sponses with the conditioned stimulus. Terrell & Ware
(1961) supported the findings of Lipsitt & Castaneda (1958).
They reported that kindergarten and first grade children
learned significantly more quickly (on size and form dis
crimination problems) -when rewarded immediately following
correct responses than when rewarded after a seven second
delay. Using a similar task and population, Ware & Terrell
(1 9 6 1) reported that Ss learned significantly more quickly
under immediate than delayed reinforcement. They siggested
that delay in reinforcement might have either thwarted goal-
directed behavior (thereby causing frustration) or have been
caused by boredom of the Ss.
Brackbill & ICappy ( 1 9 6 2) suggested that the favor
able results frequently associated with immediate reinforce
ment might be due to the fact that usually only tests of
acquisition (immediate recall) are used rather than those
requiring retention knowledge. Third grade boys learned a
series of discriminations under reinforcement delay inter
vals of 0-, 5- 5 or 10 seconds. Recognition and relearning
were tested one day or eight days after original learning.
Retention was facilitated by delay during acquisition for
the short retention interval. The facilitative effect
faded when retention was tested after the longer interval.
17
Secondary Reinforcement
As would be expected, humans differ from lower ani
mals in their greater responsiveness to secondary rein
forcers. The literature in this area was surveyed to de
termine the applicability of this type of reinforcement for
early texting.
Bijou & Sturges (1959) summarized some of the uses
of secondary reinforcers with children. Meyers (i9 6 0)
found that tokens can be established as strong secondary re
inforcers with preschoolers (ages 3 - 5 to 6-3 ) in a study
concerned with the extinction of an operant response follow
ing partial and regular, primary and secondary reinforcement
procedures. Leiman, Meyers, & Meyers (1961) used the
traditional secondary reinforcement paradigm and found it
can be established in fifth and sixth grade children.
Essentially the same results were obtained by Baer (1962)
using a "talking puppet" with nursery school children.
Aversive Conditioning
Previous work with animals has shown avoidance
schedules to establish an operant response which is regular
in rate, efficient in avoiding the aversive stimulus, and
durable over a long period of time. Punishment can be
defined in two ways: (a) presentation of a negative rein
forcement, and (b) withdrawal of a positive reinforcement.
Investigators have been generally unwilling to use
strong aversive conditioning techniques -with children,
Hammack (i9 6 0) and Baer (i9 6 0) have studied avoidance con
ditioning with preschoolers (*+ - 6 years). Baer used an
“escalator1 1 schedule on the withdrawal of positive rein
forcement (turning off a cartoon movie) and found that re
sponses came increasingly under control of the escape con
tingency of the schedule. Later (1961) Baer investigated,
with a group of 6 year olds, the effects of the same proce
dure on bar-pressing for peanuts. He concluded that the
withdrawal of showing a cartoon appeared to act as a nega
tive reinforcer. During later sessions, control Ss showed
considerable spontaneous recovery of the response, while
punishment Ss did. not, even though no longer punished.
Robinson & Robinson (1 9 6 1) sought a stimulus which
would bet (a) definitely unpleasant to most children with
out producing severe fright, (b) precisely reproducable,
(c) apparent to the child even if his close attention was
not directed to it, and (d) continuous or repetitive so that
the child could avoid it by making the appropriate response.
Results indicated that nursery school children (3£ to 51-
years) found a loud, high frequency, pure tone sound inter
rupting the playing of children’s records to be an effective
stimulus, even though there was little verbalized "insight”
into the reason for the absence of the noise.
Developmental Studies Related to Reading
Operant conditioning and related techniques have
been used to investigate emerging factors, i.e., response
differences in relationship to maturation, age, etc. Se
lected studies -which have application to the developmental
tasks required in early reading have been summarized.
Koch & Meyer (1959) pointed out that learning-set
experiments have long been worked with animals. They
investigated the relationship between rates in inter-problem
transfer in visual pattern discrimination in 3 3 preschool
children (29 to 66 months). Rates of transfer were shown
to be related to mental age as measured by the Binet. The
task proved to be too simple at a mental age above 90 months.
From this point downward, for a range of some 50 months,
learning set formulation varied with intelligence test
results.
Using two developmental levels (*+ to 5, and 8 to 9
years), Meyer & Seidman (I960) required Ss to learn the con
cept “larger than" by responding to pairs of blocks. They
concluded that the developmental level is an important vari
able in the determination of reinforcer effectiveness when
the conditions defining reinforcement are ambiguous. In a
similar experiment, Johnson & Zara (I960) also used the typi
cal transposition design with 30 nursery school children.
Ss who learned a double discrimination reached criteria sig
nificantly more rapidly and succeeded more readily on a
20
"near," "far," and "very far" transposition.
In a different type of problem, Ghent (I960) found
an almost linear relationship from 3 to 7 years of age in
identifying whether or not four tachistoscopically presented
figures were right-side-up. The 3 year olds recognized the
right-side-up orientation more frequently than any other
orientation; 5 to 7 year olds recognized the figures equal
ly well in various orientations.
Re-analyzing data from previous studies, Stevenson
& Weir (1961) indicated that the effect of reinforcement and
non-reinforcement of a single response differs significantly
as a function of chronological age. Whereas simple S-R
concepts were adequate to explain the conduct of younger Ss
(3 years) after reinforcement of a single response, another
unit of response must be used to understand the older S*s
behavior (up through age 9). They suggested that older
children may be guided by verbal hypotheses and that rein
forcement has its greatest effect in the confirmation of
these hypotheses. Reese (1962) reports that a variety of
experimental paradigms indicate that young children evidence
less mediational behavior than older children. Many vari
ables have been shown to influence such behavior. He felt
that it Is not yet clear whether the emergence of media
tional behavior is a voluntary or involuntary process, I.e.,
modifiable or unmodiflable.
Weir (1962) used operant conditioning to study ex
21
pectancy or set In children at two age levels (5 to 7 and 9
to 13). Groups selected from both age levels were given
identical variations in instructions and reinforcement
schedules. Different instructions did not produce a signi
ficant difference but the younger Ss chose the reinforcing
knob more than the older. Weir felt this supports the fact
that older Ss entered the task with the expectancy or set of
getting 100 per cent response which was not present in the
younger Ss.
Gibson, Gibson, Pick, & Osser (1962) report a devel
opmental study of the discrimination of letter-like forms In
167 children, through 8 years. A series of normative
letter-like forms, together with carefully controlled vari
ations were developed. Ss were required to select all of*
the stimuli that "looked like" a normative standard from a
large number of variants and other related stimuli. Sym
metrical patterns were confused less with their variants
than were asymmetrical ones. More errors were made when the
standard contained a closed loop. Standards with all
curved lines were associated with more errors than those
containing only straight lines. Gibson.et al. suggest that
visual discrimination of letter-like forms improves with age
because children learn "dimensions of difference" which are
critical for differentiating letters. These authors feel
that there is little or no evidence that reading readiness
tasks transfer to reading. Variables which are significant
22
for letter discrimination Instead of pictures or objects
■would be of "greater potential transfer value," The as
sumption has usually been made that features which have
helped the individual discriminate between objects will
transfer to graphic discriminations, Learning takes place
for the features which distinguish one letter from another
but slowly, if at all, for non-critical features.
Kendler & Kendler (1959» 1962). studied developmental
differences in problem solving. In one study (1959) they
found that slow kindergarten children, like infrahuman ani
mals, demonstrated nonreversal shifts. Like older Ss, more
capable kindergarten children did experience a shift. The
authors suggest that kindergarten children may be in the pro
cess of developing a mediating response which aids them in
the process of making such a shift. Later (1962), they
investigated the cue functions of words for children at two
age levels. Their evidence indicated that as the child
grows he shifts from a simple S-B to a mediational response.
They stress the importance of studying the developmental
process and its interaction with verbal mediational behavior
at different vertical levels.
ffrogrammed Learning and Auto-Instructional Devices
Although not directly applicable (and consequently
an extensive review of the literature has not been included),
some of the research on programmed instruction is pertinent.
23
There has been general support for Skinner1s position of the
need for careful programming of instructional material.
Much -work has been directed to the response that a S must
make in a learning task, i.e., -whether an overt response is
necessary or whether thinking about or reading the correct
answer is sufficient. On a program which called for a
choice between two responses on each frame, Suppes &
Ginsberg (i9 6 0) found that 5 to 6 year olds required to
overtly correct their errors learned more than children who
were merely informed whether or not they had responded
correctly.
In a study with junior high students, Goldbeck (I 960)
found that on easy items Ss who read the frames learned more
than those required to write their responses. When the
time required to learn was taken into account, this superi
ority was even greater. In two other experiments with the
same age group, Goldbeck & Campbell (1962) compared overt,
covert, and reading response modes and found that: (a) the
overt group performed below the covert and reading groups at
the intermediate level of difficulty; (b) the reading group
surpassed the other groups on both immediate and 1 0-week
retention; and (c) the reading groups learned most per time
spent.
Since these older Ss had probably already "learned
to learn," replication studies are needed with younger Ss.
Keislar & McNeil (1 9 6 2) used a teaching machine with two
matched groups of first and second grade children. Both
the overt response group (who selected the correct answer
before the program advanced) end the nonovert group (who
merely observed the program) learned significantly more than
an uninstructed control group. However, the difference be
tween these two groups was not significant. Using slight
ly older Ss (first, second, and third grade), McNeil &
Keislar (1961) found a higher correlation between achieve
ment and motivation in situations involving greater individ
ual interaction with the procedures employed than in situa
tions involving less individual confrontation in the learn
ing act. They concluded* (a) although all children did
some learning, higher and lower MA's seemed to learn more;
(b) the lack of difference in the responding groups may have
been a function of the type of final criterion, a uses test;
and (c) boys seemed to do better than girls on this type of
task even though they did not differ in their initial know
ledge or readiness for the program. These authors conclude
that sex is an important variable.
Sex Differences and Sex Role Identification
Other studies have reported a differential sex
response to tasks. The greater incidence of reading prob
lems in boys than girls has been well documented by Gates
(1961) on a population of 66k6 boys and 6U68 girls in grades
two to eight. He found that girls scored significantly
25
higher In each of the twenty-one comparisons on the Gates
Reading Survey tests.
Alpern (1961) designed an experiment to show the
interaction between the (a), masculinity-femlninity of E,
(b) the sex of the S (56 children), and (c) the sex of E
(56 college students). He found thatt (a) short briefings
for the Es did not prevent the reinforcement procedures from
effectively producing learning; (b) male Ss performed with
a greater percentage of correct responses and took fewer
trials to criterion than female Ss; (c) Ss took fewer trials
to criterion when E was of the opposite sex; (d) regardless
of sex, Ss voluntarily persisted In the learning task more
with masculine than with feminine Es.
Another study reported first by Epstein (1962) and
later by Epstein & Liverant (1963) was directed toward the
effect of sex-role identification on the verbal conditioning
of children (135 boys, ages 5 to 7)» Sex role identifica
tion was operationally defined as a preference for objects
which represent a sexual role. The child was given plastic
dolls and was reinforced with "good” whenever he said
"father" or "mother." The measure of conditioning was the
difference between the first 15 non-rewarded trials and the
last 15 rewarded trials. Ss with a strong masculine ego-
ideal were conditioned more effectively by a male than a
female E. The male E was a more effective reinforcing
agent for the high-masculine in contrast with the low-
26
masculine Identifier. Both high and low masculine identi
fiers conditioned more to "father” than to "mother." This
research points to the fact that the prediction of verbal
conditioning in children may be maximized by considering the
S's psychological characteristics, the situational factors,
and the response content.
Verbal Conditioning
Although the studies previously mentioned by Buss &
Buss (1956), Curry (i9 6 0), Terrell & Kennedy (1957)> Meyer &
Seidman (i9 6 0) have all utilized verbal conditioning tech
niques, they have been primarily directed to other problems.
Using adults, Krasner (1958) was one of the first to use
operant conditioning to control verbal response with rein
forcement. Rheingold, Gewirtz, & Ross (1959) found that
vocalizations of 3-month-old infants could be reinforced by
smiling, clucking, and lightly squeezing the infant's abdo
men. Salzinger (1959) and Salzinger, Seizinger, Portnoy,
Eckman, Bacon, Deutsch, & Zubin (1962) report similar find
ings with older Ss. They suggest that reinforcement meth
odology would permit investigation of a class of words as
they naturally appear in the child's vocabulary. It was
found, in children from 5 to 7 years, that speech rate
increases as a result of the administration of reinforce
ment, and stabilizes or decreases as reinforcement is with
held.
27
Baer & Goldfarb (1962) conducted a developmental
study of verbal conditioning using boys and girls at the
third, sixth, and tenth grade levels. They felt that using
children as Ss provides a unique opportunity to study rein
forcement history, because children at different ages have
had different amounts of experience with the same verbal
habit and with reinforcement. As a function of development
and maturation, children can be expected to vary in the type
of reinforcement to which they respond. Baer &. Goldfarb
found that boys decreased and girls increased in response to
conditioning with age. Epstein (1 9 6 2) and Epstein &
Liverant (1 9 6 3) reported that the adult male has less poten
tial as a reinforcer and encounters less appetite for rein
forcement in the adolescent male, while the reverse is true
for the adolescent female. They found a shift from same-
sex to cross-sex reinforcer potential and appetite-for-
reinforcement which occurs at approximately the age of
puberty. Certain stages in the child*s life, typified by
puberty and sex role changes, are more important from the
standpoint of reinforcement. Epstein & Liverant suggested
that their procedure showed potential for the Investigation
of personality development. It was concluded that normal
children of differing ages and sex should not be combined
in one population but should be treated as independent
groups in verbal conditioning experiments.
Eritekson (1961) found performance improved over
trials in the verbal conditioning of sixth graders. No
difference was found between the non-social (marble) rein
forcement as compared with a social reinforcement ("good®).
A short period of social deprivation (ho availability of
social reinforcers) increased the efficiency of sex rein
forcement relative to a comparable period of social satia
tion (3 0 social reinforcers presented in a 15 minute
session). Staats & Staats (1962) summarized the applica
tion of operant conditioning principles to the acquisition
of speech and reading and suggested that both behaviors are
quite similar.
Discrimination
The variables which contribute to the child1s abil
ity to make early, appropriate discriminations in learning
have also been investigated by operant conditioning tech
niques. Jeffrey (1958a) required kindergarten children to
respond "high” or "low" to varied piano tones, with equivo
cal results. In another study (1958b) he found that
labeling right-left oriented stimuli was generally difficult
for b year olds. However, learning to press 'buttons
oriented in the same direction as the figures was readily
attained by this age group. Jeffrey concluded that many
things which are difficult for children can be made more
simple. This concept has been incorporated in auto-
instructional programs and is consistent with the introduc-
29
tion of complex material into childrens education at an
increasingly early age.
Bight-left discrimination findings throw light on
the rather typical confusions between the letters b, d, p,
and q. In an early study, Davidson (1935) reported that
whereas 50 per cent of kindergarten children could make the
up-down discrimination necessary in p and q, only 10 per
cent could make the right-left discrimination in b and d.
It would appear to be essential to determine whether or not
the children (a) could not see the difference, or (b) did
not attach any importance to it. The above mentioned
study of Jeffrey (1958b) demonstrated that kindergarten
children could be trained to make right-left discriminations
and could associate this difference with stick men at least
two years before the normal 7£ year figure that Davidson
reported. Hendrickson & Muehl (1962) cite this as an
example of the necessity of accepting normative studies with
caution. Such statistics do not tell us when functions can
be taught if conditions are effectively arranged. Ad
dressing themselves to the directionality problem, Hendrick
son & Muehl investigated the relative effectiveness of
attention and motor response with kindergarten children.
The two groups who focused their attention on the difference
in directionality were significantly better in learning the
names for b and d than a control group (who discrim
inated color). No significant differences were associated
30
with the motor response group.
Other effects of reinforcement on discrimination
have been reported. The essentially favorable findings of
Terrell & Kennedy (1957) and Curry (I960) of the effects of
verbal reinforcement on discrimination have already been
cited. Kennedy, Turner, & Lindner (1962) found blame a
strong inhibitor on a visual discrimination task with high
school Ss. Erickson & Lipsitt (i9 6 0) found delay of rein
forcement to be an insignificant variable on a discrimina
tion task with fourth grade children. Miller & Estes
(1961) found no difference between $.01 and $ .5 0 levels of
incentive on visual discrimination in 9 year old boys. Of
those Ss who were able to properly discriminate between the
stimuli (two drawings of a face, the only difference being
the height and spacing of the eyebrows), less than half were
able to verbalize the relevant clue. There was no signifi
cant difference in errors or trials for Ss who could iden
tify the relevant clue as compared with those who could not.
Cantor (1955) provided preschool children pretrain
ing on distinctive names for similar visual stimuli (line
drawings of two girls). Learning of motor responses to the
same stimuli in a transfer task was facilitated. Using the
same stimulus, Norcross & Spiker (1957) and Spiker & Nor-
cross (1962) have supported the finding that possession of
verbal labels for visual stimuli in a learning task will
produce a superior performance on a transfer task in which
31
the same stimuli were associated with nonverbal responses*
Price (1 9 6 1) found that children given pretraining on a
meaningful relational name will perform better on a trans
position test than will children who received no pretraining
or who learned nonsense names.
Stevenson & Weir (1959) investigated response shift
as a function of overtraining and delay of transfer in 6 to
8 year olds. The discrimination problem was matching
colored pegs with colored lights* The degree of training,
criterion, 3 6, or 7 2 overtraining trials, was not a signifi
cant variable. The 2k hour delay between training and
transfer proved to be more efficient than immediate trans
fer*
Discrimination as it Applies to Reading
Mowrer (1951 *) referred to a theory of sentence con
ditioning, a process wherein we learn to use words both
spoken and written to become signs of real experience. He
suggests that using the non-oral method of associating the
word with an experience may ignore the fact that the spoken
words are valuable for meaning. Whatever else the child
uses to figure out a word, he must eventually pronounce it
(even if only under his breath) in order to get its meaning.
Skinner (1957) used the term "textual behavior" to
refer to an operant vocal response "under the control of a
nonauditory verbal stimulus" (printed symbols). By this
32
process the Individual brings the speech response under
control of appropriate visually presented verbal stimuli—
reading* Separating texting from “comprehension1 1 or “un
derstanding" permits a focus on the variables involved in
making the appropriate vocal responses to the printed word.
Staats & Staats (1962) reviewed the application of
operant conditioning principles to the acquisition of both
speech and reading. They felt that although a full analy
sis of reading must include the principles of classical and
operant conditioning in the development of word meaning, the
use of the term texting concerns only the operant condition
ing of a vocal response to a visual verbal stimulus. Al
though there is a great similarity between speaking and
reading, many more individuals have difficulty with reading
because: (a) speech is a more gradual process; (b) speech
has stronger reinforcers (acquisition of food, etc.); and
(c) these reinforcers are more individually applied. In
understanding reading in the school situation, consideration
must be given tos (a) the onset of a reading program which
may be quite sudden and involve a great deal of aversiveness*
(b) There are only weak sources of reinforcement for many
children in the training situation. (c) The reinforcers
which are available are not made Immediately contingent upon
the many reading responses involved. (d) If adequate rein
forcement for the child is not available in the reading pro
gram, the behaviors which are prerequisites for the acqulsi-
33
tion of reading may extinguish. It is necessary to inves
tigate how the reinforcers commonly available to the school
can he used to produce specific texting behavior.
Bloomer (1959, 1961) investigated the concepts of
meaning and the reading and spelling difficulty of words.
He found that the number of meanings that a word has, its
concreteness, and the frequency of its occurrence are all
related to spelling and reading difficulty. The concrete
ness of a word was related to its reading difficulty but not
to spelling, indicating the differences between these two
processes. The relationship between the frequency and
learning difficulty is not a linear function; greatest
learning occurs in the initial trials. The frequency of
occurrence of a word related most to reading and spelling
difficulties. Bloomer suggested that this was indicative
of the fact that the more familiar the word, the more mean
ing it has. He concluded that the teaching of reading will
be more effective if greater emphasis is placed upon words
which have specific and concrete meanings and also if
greater emphasis is given to making the words meaningful to
children.
Several studies have been directed toward the role
of reactive inhibition, a term introduced by Hull to refer
to a negative drive state, analogous to fatigue, which comes
about as a consequence of activity. Adopting the measure
of reactive inhibition traditionally employed with adults,
Lynn (i960) used an inverted number printing task with chil
dren between 8 and 11 years. The findings give some sup
port (a low but significant negative correlation) to a rela
tionship between accumulative reactive inhibition and skill
in reading. In an extension of this work, Otto & Fredricks
(1963) hypothesized that good readers generate less reactive
inhibition than poor readers over a period of time. Their
data obtained from 10 good and 10 poor readers each, from
grades four, five, six, seven, and eight, supported this
hypothesis. These authors suggest that, at least in the
early stages, school learning such as simple addition and
reading takes place by instrumental conditioning•
In a 3 x 3 x 3 factorial design, Otto (1961) used
three levels of reading proficiency, three grade levels
(two, four, and six) and three modes of reinforcement (audi
tory, visual-auditory, and kinesthetic-visual-auditory) to
study acquisitions and retention of paired associates. His
conclusions weres (a). Good, average, and poor readers, in
that order, require increasingly more trials to master a
list of paired associates. Average and poor readers might
be expected to profit more from auto-instructional devices,
(b) Both reading level and grade level must be considered
important variables. (c) There was no significant inter
action between the three modes of reinforcement and reading.
Additional sensory cues did not seem to be of value.
35
(d) Once they have mastered It, poor readers retain a list
of paired associates after 2b hours as well as good readers*
This final conclusion does not support the commonly held
idea that poor readers do not retain what they have learned
as well as good readers.
Other investigators have directed their attention to
the discriminated cues used in texting. The results of
Arnoult & Lewis (i9 6 0) showed that increased exposure time
provided little improvement in discrimination accuracy with
tachistoscopically presented pairs of nonsense figures re
quiring a judgement of “same1 1 or "different." It appeared
that the longer the child had to look at the figure, the
more likely he was to find "differences" in identical stimu
li. It is somewhat equivocal whether or not these findings
suggest that a wholistic perceptual approach is most effec
tive.
La Pray (1962) tested several hypotheses based on
the linguistic approach to first grade reading. In opposi
tion to the linguistic position, she founds (a). The word
configuration is exceedingly important to reading in first
grade children. (b) The best readers were highest in
recognizing known words through configurations. (c) Stu
dents who were low in configuration were not significantly
helped by the addition of a single letter at the beginning
or end of a word.
Spiker (1956) investigated the factor of distinc-
36
tiveness and equivalence of cues with, children and found
that stimulus differentiation pretraining from one visual
stimulus to another provides a positive transfer to a dif
ferent response. Later (I960) he used an associative
transfer paradigm similar to those used in pretraining
studies. When the initial task (visual) in pretraining and
transfer were relatively dissimilar, interference was set
up from the first to the second responses.
Muehl (I960) investigated the effects of different
types of visual discrimination pretraining on the perform
ance of kindergarten children in learning to read a vocabu
lary list. He found that children who received pretraining
in matching the same words that appeared in a later vocabu
lary list learned to “read1 1 the words significantly better
than children who received similar pretraining with dif
ferent words or with geometric forms. Muehl suggested
that it was the response to single letters or details of
letters that appeared to account for his findings. He
hypothesized that transfer to the learning task was a re
sult of acquisition during the pretraining of a set of
observing responses to distinctive letters in the stimulus
words. The beginning reader must be able to associate word
names, or meanings, with their printed symbols which must,
in turn, be distinctive and stable stimuli. More needs to
be known about the visual discrimination cues used in the
texting task. Muehl's results indicate that his Ss used
37
single letters (most often the first or last) or details of
these letters to make the distinction. The child often
would not have known the name of the letter. Such evidence
would support the large body of knowledge about the impor
tance of visual-perceptual skills in early reading.
In another experiment by Muehl (1961) nonsense
words were used with 60 kindergarten Ss divided into three
pretraining groups;, (a) the word had a relevant shape and
a relevant letter; (b) the word had an irrelevant shape
but a relevant letter; and (c) the three letters which were
subsequently used in the discrimination task were presented
separately. No difference was found between the first and
second group. This was consistent with the assumption that
children discriminate among similar length words of differ
ent shape on the basis of specific letter differences.
The group using the three letters separately did better on
the pretraining, but not the reading task response. Muehl
suggested that although letters included in words are more
difficult to discriminate than those presented singly,
careful visual discrimination training will overcome this
difficulty.
In another study, Staats, Staats, & Schutz (1962)
used three groups of 12 kindergarten Ss. The discrimina
tion material progresses through the training with little
opportunity for failure and reinforcement was contingent
upon correct response. A familiarization procedure per-
mitted S to "buy" trinkets with tokens. In discrimination
pretraining, Group I used the same words later seen in the
criterion task. Group II received the letters that were in
the criterion words. Group III received no discrimination
pretraining. All were given the same texting task immedi
ately following the pretraining. Although learning took
place, an analysis of variance did not show a significant
difference between the groups. A Duncan multiple range test
showed significant differences between Group I and Groups II
and III. This supported the typical finding that stimulus
discrimination pretraining results in positive transfer to
another task when the same stimuli are presented in the la
ter task. The authors concluded that pretraining on letters
outside of the complex of the word was not helpful. There
was a lack of significant correlation between discrimination
and texting and the Columbia Mental Maturity Scale and
texting. This would indicate that other behaviors are in
volved in these tasks. It should be noted that the S was
not required to actually say the letter names. ’ ’ Thus, while
the results question the relevance of pure discrimination
pretraining on alphabet components or words in facilitation
of later texting behavior, it may be that texting training
on these units would prove valuable for transfer to the
whole word. Further research relating texting of letter
units to the later texting of entire words thus would appear
to be fruitful." (Staats, Staats, & Schultz, 1962, p. 3 6.)
39
One other variable, the knowledge of letter sounds,
has been investigated. Durrell (1958) reported on an ex
tensive study based on over 2,000 first grade children.
September tests which provided best prediction of June word
recognition and reading performance were: writing letters
dictated, naming letters, identifying letters named, and
learning rate for words. Durrell suggested that most read
ing difficulties could be prevented by an instructional pro
gram which provides early instruction in letter names and
sounds followed by applied phonics and techniques to devel
op a sight vocabulary and silent reading skills. Mental
age, as measured by the Otis Quick-Scoring Tests of Mental
Ability, had a low correlation with reading achievement and
letter and word perception skills. Robinson & Wright
(1959) question the methodology and statistics upon which
these conclusions are based.
As an extension of Durrell*s and his own earlier
work, Muehl (1962) divided 87 kindergarten children into two
groups for pretrainings (a) a “relevant" group learned
names for three letters which subsequently were the critical
stimuli for three pictures and (b) an "irrelevant" group
learned names for letters that did not appear in the words
of the transfer task, matching a picture with the nonsense
word. The reduced variability and restricted range in the
"relevant" group indicated interference with the "reading"
task. Muehl suggested that these results may have been
to
■brought about by any one of several factorsr (a) the label
ing response may not relate to the identifying response when
the S sees the stimulus in the word; (b) kindergarten chil
dren may not be able to utilize the information provided for
them by letter names; and (c) the value of the letter name
may be in its phonetic attack on the word. If this is the
case, the use of letter names (which do not always corre
spond with the phonetic sound) might have resulted in the
observed negative transference.
Summary and Critique of the Literature
with Special Reference to Reading
A summary of the literature indicates that the use
of single level, fixed ratio, immediate reinforcement tends
to bring about greatest facilitation of learning in chil
dren. Both positive and negative reinforcers seem to bring
about conditioning, with a slight advantage in favor of
positive.
Child growth and development studies have indicated
that both the level of developmental and perceptual organi
zation are important. It would appear that older children
make more use of mental set and verbal hypotheses in learn
ing. These factors must be taken into account in helping
children to learn to read.
Careful analysis of the sequence of learning and
programming of material tends to maximize learning. The
ability level of the learner is an important variable. The
bl
fact that slow and average readers take more trials to learn
(Otto, 1961) may mean these groups will ultimately show
greater relative gains from auto-instructional devices than
will fast learners. The necessity for an overt response is
being questioned but good interaction with the material is
desirable.
The results of certain previous studies must be
questioned in view of the recent findings that sex is an
important variable for which adequate control has not always
been applied. Although boys have greater difficulty with
reading, there is some evidence (McNeil & Keislar, 1961;
Alpern, 1961; and Otto, 1961) that operant learning tech
niques can be developed which will be of more help to boys
than girls. The sex of both the subject and the examiner
and the interaction between the two have been shown to be
important variables. The psychological characteristics of
the learner, the situational factors in which the learning
takes place, and the response content need to be better
understood if prediction of learning is to become more
accurate.
Verbal conditioning has been demonstrated at an
early age. Operant conditioning techniques have proven
helpful in studying the emerging verbal function. There is
a similarity between the process of speaking and reading.
Fewer problems tend to be associated with speech because its
acquisition is more gradual, and stronger, individually-
h2
applied reinforcers are more frequently available.
Discrimination studies have indicated that early
training and careful programming will produce responses
(i.e., right-left discrimination) which would normally not
develop until later. Visual perceptual differences appear
to be related to success in reading. Even though children
can make a correct response in a visual discrimination task
they may not be able to verbalize how they were able to do
it.
A low, significant, negative correlation has been
found between reactive inhibition and reading (Lynn, I960).
Once they have mastered it, poor readers retain a list of
paired associates as well as good readers (Otto, 1 9 6 1).
Research utilizing the traditional motor response
(e.g., lever pulling) has frequently made it difficult to
apply the findings of this research to the reading task.
Although it has insured experimental accuracy and facili
tated replication, emphasis on laboratory apparatus has
reduced the applicability of findings to the school situa
tion. A complete operant learning procedure in reading
awaits more information about factors affecting textual
discrimination and the effects of individual differences on
this process.
Discrimination research has shown that long exposures
tend to cause Ss to report non-existent differences (Arnault
& Lewis, i9 6 0). Interference results if the initial task
^ 3
in pretraining and transfer are relatively dissimilar
(Spiker, 1956). Pretraining in matching words produced
better results than matching geometric fonns (Muehl, I960).
Transfer to the learning tasks may result from observing
responses to distinctive letters in the stimulus words.
Single letters, usually first or last, tend to be used to
make the necessary distinctions (La Pray, 1962; Muehl,
1961, 1962). Careful visual discrimination training can
help children make the difficult but important discrimina
tion between letters appearing in words as opposed to those
appearing singly (Muehl, 1961).
Visual discrimination pretraining on single
letters later appearing in words did not produce a signifi
cantly different response than pretraining on words (Staats,
Staats, & Schutz, 1 9 6 2). This suggests that the Ss were
responding to the configuration of the entire word rather
than to specific parts. In the same design, alphabet pre
training did not result in more correct responses. This
may be due to the fact that in pretraining the S was only
required to make a visual discrimination and did not learn
the names of the letters. These authors suggest the possi
bility that pretraining on either the letter-names or the
letter-sounds might produce more transfer to the texting
task. Muehl®s extension of this work (1 9 6 2) indicated that
learning letter names in nonsense words interfered with the
later association of these words with pictures. Again,
this would not appear to he the same task that is Involved
in actual texting, i.e., learning a letter name may not help
make a correct visual discrimination (nonverbal) tout it may
have a positive transfer to a texting response. Children
seem to rely on configurational cues which tend, in the long
run, to toe unreliable. In short, although results have
generally substantiated the importance of toeing able to
discriminate the letter within the configuration of the
word, they have been equivocal in regard to pretraining for
the letter-name versus the letter-sound. Phonetic based
approaches to reading frequently stress the importance of
early knowledge of vowels and vowel sounds. There Is need
for further research on the effectiveness of this approach.
CHAPTER III
PROCEDURE
Purpose
The present study extends the stimulus pretraining
research by comparing the effects of three varieties of the
auditory and two of the visual dimension on early texting
responses in children. Muehl (I960, 1961, 1962). and
Staats, Staats, & Schutz (1 9 6 2) indicated that discrimina
tion of the letter within the configuration of the word is
an important part of the early texting response. Two vari
ations in the visual dimension were studieds (a), consonant-
configurational cues and (b) vowel cues. Research has
demonstrated that configurational cues facilitate transfer
to the initial texting task. Vowels provide even more
basic information in the reading task. It was hypothesized
that pretraining on the vowel cue would produce more total,
correct texting response.
The work of Durrell (1958) and Muehl (1962) raise
questions about the effectiveness of pretraining on letter-
*5
1 + 6
sounds. Muehl ( 1 9 6 2) used pretraining on letter-names of
three letters (two of which had configurational cues) and
found this interfered with later texting of nonsense syl
lables. Muehl suggested that his results may have been
accounted for by the fact that he used the letter-name
rather than the letter-sound.
The present study investigated the effects of pre
training on three variables in the auditory dimension:
(a) the letter-name, (b) the letter-sound, and (c) no audi
tory pretraining. It was hypothesized that pretraining on
the letter-sound as it appears in the stimulus word, would
have greater transfer effect to later texting responses than
would letter-naming. Contrary to Muehl*s findings (1962)
but consistent with educational practice, it was also
hypothesized that both letter-naming and letter-sounding
would be more effective than no auditory pretraining.
There is more differentiation in English between letter-
names and letter-sounds in vowels than in consonants, which
make up the primary configurational cues. It was hypoth
esized that simultaneous training on both letter-sound and
vowel cues would produce a facilitating interaction effect.
In summary, the study was designed to test the
following hypotheses:
1. Auditory pretraining on letter-names in conjunction
with visual pretraining on configurational cues would pro
duce significantly more correct texting responses than would
^7
training in letter-names with vowel cues.
2. Auditory pretraining on letter-sounds in conjunction
with visual pretraining on vowel cues would produce signifi
cantly more correct texting responses than would training in
letter-sounds with configurational cues.
3. Auditory pretraining on both letter-names and
letter-sounds would produce significantly more correct text
ing responses than no auditory training.
Auditory pretraining on vowel cues would produce
significantly more correct texting responses than would
auditory pretraining on configurational cues.
5. Visual pretraining on vowel cues would produce sig
nificantly more correct responses throughout the total pro
cess of texting training.
6 . Visual pretraining on configurational cues would
produce significantly more correct responses in the initial
stage of the texting task.
The experiment was conducted during regular school
hours in small rooms near the class which the Ss attended.
Ss were seen one at a time and were seated across the table
from E. All of the discrimination pretraining, familiar
ization and texting training was done by the investigator.
Subjects
The 120 Ss were all enrolled in one of the 32
schools in the Cupertino (California) School District which,
* + 8
at the time of the data collection, had 1 8 ,8 8 6 students
enrolled, kindergarten through eighth grade. This primari
ly urban district includes all of one community and parts of
five others. The area is essentially residential and is
made up of families mostly from the upper and lower middle
class. "Non-Anglos," primarily Americans of Mexican,
Japanese, and Chinese descent, compose about one per cent of
the population. Teachers were asked not to refer children
of bilingual background or from families which did not,
broadly speaking, represent the middle class.
The study was delimited to investigate discrimina
tion pretraining in the beginning textual response. The
population was therefore restricted to students in the first
grade. School records verified that all of the 120 Ss were
born between December 2, 1956 and December 1, 1957; normal
first grade age for that school year in California schools.
The data were collected during November and December, 1963;
the chronological ages ranged from 5-11 to 7-0 years. Al
most all Ss had previously had a year in kindergarten as
well as the eight to fourteen weeks in first grade.
Since boys have more difficulty with beginning read
ing than girls and since sex has been demonstrated to be an
important variable in discrimination studies, the population
was further delimited to first grade boys. No S was
included who had previously been referred for psychological
evaluation.
IQs were not available for students in the first
grade. The Vocabulary Test from the L-M Binet was used to
obtain an approximation of verbal intelligence# The
majority of these vocabulary tests were administered by the
investigator but in some schools they were given by other
qualified pupil personnel workers# The test helped to
establish rapport with the S and served to introduce him to
the experimental situation. The range of the approximated
verbal IQs was restricted to 85 to 115# All teacher nomi
nations were screened from the 1 5 schools in the district
vjith the lowest mean intelligence scores. Out of this
nominated group, 87 boys who scored too high on the Vocabu
lary Test and 11 who scored too low were not included in
the study. The 120 Ss finally chosen had a mean approxi
mated verbal IQ of 102.5 with a standard deviation of 9.0.
The most comparable scores available in the district
were from the California Test of Mental Maturity administer
ed to all fifth grade students in October of that school
year. Median Total CTMM IQ for this group was 111. Thus
although the experimental group was in the normal range when
compared with national norms, they were below the district
mean# The fact that boys experience more difficulty with
reading than girls has been cited. These factors tended to
produce a slower rate of development in the Ss than other
first grade students. In heterogenously grouped schools,
they tended to be dn the lowest reading groups. In 12 of
the 1 5 schools where variations of homogeneous groupings
were used, they tended to he grouped together in the lowest
classes. A wide variety of instructional techniques were
used in the 1 5 schools. To cancel out the effects of these
instructional differences, qualified Ss were asked to pick a
slip of paper which assigned them randomly to eight experi
mental groups (N = * 1 5).
Since red letters were used in some of the stimulus
words, color hlindness was ruled out by the use of one plate
from the Pseudo-Isochromatic Test developed by Dvorine
(1953). A normal older child or mature person sees a num
ber 23 on this plate as a figure on a dotted background;
red-green and totally color-blind Ss cannot see any number.
Differentiating the number proved too difficult for most of
the Ss in this study. In a modification of the usual tech
nique, Ss were asked to use their finger to follow the line
with the "different color." Nine boys were eliminated
because possible color-blindness could not be ruled out.
Throughout the experiment no attempt was made to
follow precisely a set of instructions. This was done to
help the S know what was expected of him, to gain and main
tain rapport, and to simulate the natural flexibility of the
teacher.
51
Stimulus Material
To more closely approximate the school setting,
actual words were used rather than nonsense syllables. All
contained four letters; none had a commonly accepted,
strong emotional connotation. Stimulus words were chosen
that would be understood by a majority of Ss when a defini
tion was furnished them even though they might not be able
spontaneously to produce such a definition. Since all Ss
received the same stimuli, no assumption of equal difficulty
was necessary. To provide information about the relation
ship between the S*s familiarity with a word (Bloomer, 1959*
1 9 6 1) and his texting response, words were chosen of differ
ing levels of familiarity and difficulty.
To insure adequate variation and control of the
con sonant-configurational cues, the following factors were
considered; (a) Wo %«/ord contained more than one major con
figurational cue. (b) Wo major configurational cue was
used more than once and no word contained a double conso
nant. (c) Words were chosen which maximized the difference
between the letter-name and letter-sound of the configura
tional cue. (d) The four major types of configurational
cues were used, one each beginning and ending ascender and
descender.
To insure adequate variation and control of the
vowel cues, the following factors were considered: (a) Wo
word was used with either two vowels or a diphthong.
52
(b) The same vowel did not appear more than once in the four
words, (c), Short vowels were selected to maximize the
difference between letter-names and letter-sounds.
All groups received the same wordss draw, must,
grin, and crop. Rinsland (1 9^5) analyzed samples of
writings and recorded conversations of 7 > 9 5* * first grade
children. Tabulations identified 5*099 different words in
the 353*87*+ total words used at this grade level. In this
list, ”drawM appeared 2 7 3 times, placing it in the second
group-of-100 most frequently used words. “Must" appeared
8 7 times and was in the fifth group-of-100 words. "Grin"
was not used at all at this grade level, and “crop" was used
only once. Rinsland suggests that the definiteness of
knowledge about each word tends to vary with its frequency
of occurrence.
In order to facilitate replication in a typical
school setting, all stimulus material was typed on plain,
white 5 x 8 inch cards with a standard primary typewriter.
The procedure consisted of four parts: (a), discri
mination pretraining, (b) familiarization, (c) texting
training, and (d) inquiry.
Discrimination Pretraining
Three stimulus words appeared on each of the 12
discrimination pretraining cards. They were evenly spaced
horizontally and centered vertically. Each of the four
selected words appeared singly, to be systematically discri
minated with each of the other three ■words, which appeared
twice* The position of the correct alternative on the
card was varied across the series of 1 2 cards, to prevent a
position preference* Three sets of cards were used: one
with the configurational cue in red, one with the vowel in
red, and one with all letters in black. Three groups saw
the red configurational cue and three the red vowel cue.
A seventh group saw the stimulus words all in black and an
eighth group received no discrimination pretraining; each
group had I? Ss.
For the seven discrimination pretraining groups,
the 1 2 discrimination cards were randomly gone through once,
E directed Srs attention to the configurational or vowel
cue appearing in red (or the whole word for the Discrimina
tion Pretraining Control Group) by pointing and/or saying,
"Look at this." After the appropriate instructions were
given, S was requested to, "Show me the one that is differ
ent (or not the same)." Instructions for the experimental
groups were as follows:
1. For the Letter-lIame-Configurational Group (LNC) E
pointed to the configurational cue and indicated, "This is
a(n) _____ (giving the name of the letter).”
2. For the Letter-Sound-Configurational Group (LSC) E
pointed to the configurational cue and indicated, "This says
(giving the letter sound)." In three of the four
■words the configurational cue was part of a consonant "blend*
B pronounced the consonant as it was used in the blend with
out actually saying the complete phoneme*
3* For the Configurational-Control Group (CC) E pointed
to the red configurational cue and asked S to indicate which
one was different.
b . For the Lett er-Kame-Vowe 1 Group (I1TV) E pointed to
the vowel appearing in red and indicated, “This is a(n) ___
(giving the name of the letter)*"
5* For the Letter-Sound-Vowel Group (LSV) E indicated,
“This says ____ (giving the sound of the vowel as it was
used in the word)."
6 . For the V o \ r e 1-Control Group (VC) E pointed to the
red vowel cue and asked S to indicate which one was differ
ent*
7* For the Discrimination-Pretraining-Control Group
(DPC) all letters of the stimulus words appeared in black*
E pointed to each word and said, "Look at this"; S was
asked to find the word that was different.
8 . The Kon-Discrimination-Pretraining Group (NDP) was
given no practice in discrimination.
Familiarization
Familiarization, texting training, and inquiry were
the same for all eight groups. The familiarization pro
cess immediately followed the discrimination task for the
55
seven pretraining groups and was the first task for the NDP
Group. ITeither configurational nor vowel cues appeared in
red on the four cards in the familiarization deck. The
words were presented one at a time on 5 x 8 inch cards in
ascending order of difficulty as determined by a pilot
study; draw, must, grin, and crop. E directed S’s atten
tion to the word on the card and indicated, "This word says
. Can you tell me what ____ means?"
After the S had attempted to define the four words,
the familiarization cards were gone through again. This
time E said, "This word says _____• Will you say ____?“
Texting Training
The texting training task, a modification of the
paired associates method, immediately followed. The text
ing training deck consisted of forty 5 x 8 inch cards.
Each word was presented randomly ten times. Instructions
were, "Nov; the game changes. I am going to put these cards
(indicating the Ho cards) in front of you and show than to
you one at a time. I want to see if you can guess the
word before I say it. Try to say the word before I do."
The cards were placed in a pile at reading distance in
front of S. Each card was turned over as it was removed
from the top of the pile so that only one word was exposed
at a time. The E said, "This word says ____," and paused
to give S a chance to respond. If there was no response
56
after approximately five seconds, E provided the -word for
S. If S gave the correct response, he was given verbal
reinforcement by E who said "good" or "right." or repeated
the word. If S gave an incorrect response, E indicated,
"No, that says ." and gave the correct response.
Inquiry
After the training was ended, an inquiry was con
ducted to see if the S knew what cues he had used to iden
tify the word. The same cards and order were used as had
been in the familiarization process, S was asked, "Do you
remember what this word Is?" S*s response was reinforced
as in the training process. E then asked, "How do you
know that this word says ____?" If Wo cues were given,
E asked, "Which helps you the most?" Inquiry continued
until E had enough information to tabulate S’s response.
CHAPTER IV
RESULTS
Equivalence of Grout)s
There were 1? Ss in each of eight discrimination
pretraining groups. The groups v r e re analyzed for varia
tion of age and level of ability. Means and Standard
Deviations of the chronological age and approximated Verbal
IQ are provided in Table 1. Analyses of variance were
computed to determine the possibility of group differences
which might affect the results. Because of the restric
tion of the age and level of ability in the Ss and follow
ing the practice of Boneau (i9 6 0), tests for homogeneity
of variance were assumed unnecessary. Table 2 presents
the results of these analyses for CA and IQ. The results
of these tests indicate that the groups did not differ
significantly from each other on CA or IQ.
Stimulus Words
During the familiarization process9 each of the 120
Ss were asked to define the four stimulus words. Accept?-
57
58
Table 1
Mean and SD of Verbal IQ and CA (in months)
for Pretraining Groups (N = 15)
Group
Verbal
M
IQ
SD
CA in
M
months
SD
LNC 1 0 5A 7 8.05
77-0 2.90
LSC 1 0 0 . 20
8.53 77.> f 2 . 3 6
CC 1 0 2 .0 7
7.69
77.6 2 . 9 6
LKV
106.07
8.90
77 A 2.87
LSV 100.53
8.97
7 8 . 0 3 . 1 0
VC
101.93 9.05 77.8
3.17
DPC
100.67
9 . 6 7
76.5 3.55
NDP 1 0 3 .^ 0
9.03 79. 3.59
Table 2
Analyses of Variance of CA (In months)
and Verbal
IQ
Source of Variance DF SS F
Groups— CA
7 67.39 <1
NS
Error 1 1 2 11*4-9.61
Groups— IQ
Error
7 581.11 ! - 1.02 NS
112 91^5.86
59
able definitions were given as follows: draw, 114-; must,
grin, and crop, 28. The percentage of correct
responses ranged from 95 per cent (draw) to 2 3 per cent
(crop). These figures are consistent with Rinsland*s find
ings (19*+5) previously reported. The words would appear to
represent different levels of difficulty as measured by the
ability of the Ss in this population to give adequate
definitions.
Discrimination Pretraining
During the discrimination pretraining task Ss in
seven of the eight groups were asked to identify the “word
that is different." Although more difficult than most
reading readiness material, this task proved to be quite
easy for this population. Out of a possible score of 12,
the mean number of correct responses for the 105 Ss receiv
ing this pretraining was 11.62 with a standard deviation of
. 8 3. The group that did not receive either a configura
tional or vowel cue (DPC), with a mean of 11.1 3 and SD of
1.087? had the least number of correct responses. The
difference between DPC and the group with the highest mean,
however, was not significant.
Texting Training
The major part of the experiment involved testing
six hypotheses on a 2 x 3 factorial design. In order to
maximize pooling effects, only the data from six of the pre
60
training groups were used in the main analysis of variance.
Data from the control groups, DPC and NDP, were treated
separately. Each S*s ho texting responses were divided
into five blocks of eight trials to study the effects of
texting over trials. A Lindquist (1953) Type I analysis
of variance separated out the effects of pretraining
experiences in these six groups over trials. A modifica
tion of Analysis of Variance Program BIMD 11 was used on an
IBM 709b Computer at the Western Data Processing Center,
School of Business Administration, University of California,
Ins Angeles. The data were run twice in order to examine
the effects of replication for both Ss and trials. The
first run was a subjects analysis (one variable) over
trials (replicates) based on total scores. The second
dealt with the major experimental variables, the two levels
of the visual and the three levels of the auditory variable,
for each of the 90 Ss. These data were incorporated into
a single analysis of variance. The results of this analy
sis are presented in Table 3*
The appropriate test (Table 3 ) for the total sum of
squares within subjects and between subjects was the within
subjects error term. This pooling resulted in a signifi
cant F (.01) for the means between subjects. This would
support the typical finding that large differences among
individuals were maintained in spite of the relatively con
stant experimental treatment.
61
Table 3
Analysis of Variance of Correct Textual Responses
for Six Pretraining Groups
Source of Variance DF SS F
Between Subjects
Visual 1 20.056 2.4-3 ns
Auditory 2
26.4-13
1.60 NS
Visual x Auditory 2 4-2.004-
2.55
NS
Error 84- 692.826 • •
Total Between Subjects
89 781.299*
3.52 <.01
Within Subjects
Trials 4 - 2 6 3 .3 5 6 36.46 < .0 0 1
Trials x Auditory 8
7.631 <1 NS.
Trials x Visual 4 - 9.400 1.30 NS
Trials x Visual x
Auditory 8
IO.0 3 9 <1 NS
Error 336
606.773
• •
Total Within Subjects 360
897.199* 1.35
NS
* Tested by within subjects error term.
62
Table b
RHOS Between Approximated Verbal IQ and
Total Correct Texting Responses
Group Rho
LUC
•05
LSC .28
CC -.11
LUV .56*
LSV
Jh
CM
•
1
VC -.38
DPC -.11
UDP
.17
* p. 0 5
63
In an attempt to determine the relationship between
IQ and texting, Rank-difference coefficients of correlation
were computed between the ability of the Ss and their number
of total correct texting responses. These results appear
in Table *+. Although the ability range was restricted,
IQs in the population were free to vary as much as 30
points. It would appear that sources of variance other
than that measured by the Binet Vocabulary Test must still
be identified to account for the major part of these indi
vidual differences.
The F of 36.*t6 (significant at the .001 level)
associated with the means between trials (Table 3) indi
cates that learning did take place. The mean total texting
response for the 90 Ss was 19-50. This would mean that the
average S responded correctly *+9 per cent of the time. In
the main analysis of variance the means of the five blocks
of eight trials were as followss Block I, 2.^7; Block II,
3 . 8O-5 Block III, *+.26; Block IV, lf.31 * - ; and Block V, b . 63.
As would be expected, these data do not approximate
a straight line. The sum of squares associated with equa
tions for the line of best fit were tested for significance,
using the within subjects error term. Both the linear and
quadratic equations were significant at the .0 1 level and
the cubic equation was significant at the .05 level. This
would indicate that a simple linear estimation did not fit
the data and that the texting task used in this study is
6b
complex in nature. This was also substantiated by observa
tion. Many of the Ss would start texting training with few
successes. After they began to improve, they often reached
a plateau or period of regression when they tended to be
come confused and miss words that they had previously texted
correctly. Toward the end of training, they would usually
begin to make more correct responses again. The uneven
ness of the learning curves for all eight groups can be seen
in Figure 1.
Hypothesis 1 suggested that pretraining on letter
names would produce more correct texting responses when
presented in conjunction with configurational cues. Hy
pothesis 2 suggested that a similar facilitative interaction
would be found between letter-sounds and vowel cues. Since
the A x V interaction term shown in Table * + was not signi
ficant, both hypotheses were rejected.
Adapting a procedure reported by Spiker & Norcross
(1962), t tests were used to study the differences between
the means of Block V, the last eight trials. Group GC was
significantly higher than LNC and NDP at the .05 level of
confidence. The difference between CG and LSG just missed
significance at this level. This would support the
previous findings that children make effective use of
configurational cues in the early texting task.
Since the learning curves (Figure 1) show CC and VC
to have the highest scores on Block V, it was decided to see
Mean num ber of correct textu al
o
A
( 0
0
c VC V
c LWVv
L
LSC c c LSVv v
DPC * * * * * * ' * * NDP
ft e ft • o * *
- V i
X X
in nr ic s:
BLOCKS OF EIGHT TRIALS
. 1 .--C o r r e c t texting resp on ses for e ight discrim ination pretraining groups.
66
•whether or not these two auditory control groups were
different from all four groups which had received auditory
pretraining (LNC, LSC, LNV, and LSV). The t between the
means of these combinations of groups indicate that the Ss
who did not receive auditory pretraining did significantly
better (.05) on the final Block of the learning task. This
supports Muehl''s (1 9 6 2) findings that training on letter-
names interfered with later texting.
Hypothesis 3 stated that pretraining on letter-
names and letter-sounds would produce more correct responses
than no auditory pretraining. Not only must this be
rejected, but these findings indicate the opposite effect;
learning letter-names produces texting interference. Thus
the lack of significance in the auditory main effect would
extend previous findings reported with kindergarten children
to this first grade population.
Hypothesis stated that auditory pretraining on
vowel cues would produce significantly more correct texting
responses than would auditory pretraining on configurational
cues. The insignificant A x V interaction effect in the
analysis was not used to test this hypothesis because its
results were confounded by the auditory control groups (CC
and VC). Total texting response differences were studied
between the Ss who received auditory training on vowel cues
(LSV and LNV) and those receiving auditory training on con
figurational cues (LSC and LNC). This difference, at the
67
•01 level of confidence, supports the hypothesis that audi
tory pretraining on vowel cues produces more total correct
texting responses than does auditory training on configura
tional cues.
Hypothesis 5 suggested that visual training on
vowel cues would produce significantly more correct texting
responses throughout the texting training. The T x V in
teraction term which included the two auditory control
groups (CC and VC) was the appropriate measure with which to
test this hypothesis. The lack of significance of this mea
sure lead to a rejection of the hypothesis. To test further
the significance of differences between total texting re
sponses (rather than for trials as in the major variance
analysis), t tests were computed between the means of the
groups. Group CC produced significantly more correct text
ing responses thans LNC (.02) , LSC (.05), and DPC (.05).
LNV was significantly higher than LNC (.05). Thus the
superiority of the CC group previously found in Block V,
remains true over the total texting response. The superi
ority of the vowel cue over the configurational cue (LNV
versus LNC) was also mentioned in testing hypothesis h .
Hypothesis 6 suggested that visual pretraining on
configurational cues would produce significantly more
correct responses in the initial stage of the texting
response. Examination of Figure 1 indicates that no sig
nificant variation was produced in Block I by the different
pretraining experiences* Group DPC, -which did not receive
red letter cues, had the lowest mean score over the first
eight trials (Block I) hut this difference was not signifi
cant. The V x T interaction in the main analysis was not
significant and, if the segment of learning measured in this
study is considered total learning, the hypothesis must be
rejected. On the other hand, all five Blocks might be
considered the initial stage of a larger texting process.
As has already been demonstrated, the performance of group
CC would support previous findings that configurational
cues facilitate transfer (Muehl 0-960, 1961, 19623 and
Staats, Staats, & Schutz O.9 6 2 3 ) . The reduced number of
total texting responses associated with auditory training on
configurational cues has also been discussed. Thus, al
though the hypothesis as stated is rejected, these data do
not directly negate previous findings.
Another analysis of variance was made between the
means on total texting responses for all eight groups.
Since the F test was not significant, the Duncan multiple
range test (Edwards, i9 6 0), a procedure previously used by
Staats, Staats, & Schutz (1962), was employed. The only
group which was significantly lower than the rest of the
groups on this statistic was the LNC (.05)• In view of the
lack of significance in the main analysis, this finding
would appear to be of limited value.
To study the effects of the red letter appearing in
69
the discrimination pretraining task, a control group which
saw all of the stimulus words in black (DPC) was incorpo
rated into the design. The following facts about this
group were noted: it had the lowest mean score on the dis
crimination task; it had the lowest mean score for Block X;
it was among the four lowest groups on Block V; and it
had the second lowest mean total texting score. It was
significantly lower (t equal to . 0 5) than the highest group
(CC) on total texting response. In no other case, however,
were these differences significant. In summary, DPC
appeared to experience some difficulty at several points
during discrimination and texting training. It cannot be
said within the normal level of confidence, however, what
effect the additional cue appearing in red has on the text
ing response.
During the experiment it did not appear that the
discrimination task was providing appropriate transfer to
the training task. Control group NDP, which did not
receive any pretraining, was incorporated to further inves
tigate this factor. Although the learning curve of NDP
shows a more exaggerated drop for Block V than any of the
other groups, the observed difference was not significant.
It must be concluded that, for this population, the dis
crimination pretraining did not facilitate transfer to the
texting training task. Although this finding substantiates
that of Staats, Staats, & Schutz (1962) it is at variance
with the findings of Muehl (I9 6 0) and others
70
Correct Texting Responses per Word
The total number of correct texting responses per
word, based on 10 exposures of each word to the 120 Ss,
appear in Table 5.
An analysis of variance between the mean number of
correct texting responses for the four stimulus ^rords is
presented in Table 6 . The lack of significance associated
with the F test Indicates that texting did not significantly
vary with the difficulty of the word. This is an unusual
finding in light of the marked difference in the level of
difficulty of the words both in the general first grade
population (Rinsland, 19^5) and in the differential ability
of the Ss to define the word adequately (see Chapter III).
These results would tend to suggest that knowledge of the
word does not, as is commonly thought, necessarily facili
tate a correct texting response.
Verbal Report of Discrimination Cue
The S*s verbal report of what cues he used to dis
criminate a texting response is generally considered to lack
reliability. The S may make a discrimination upon a vari
ety of changing cues. He may have difficulty in verbaliz
ing the method used. He may say what he has been taught
but be actually using different cues. There is also the
possibility that the E may have difficulty in understanding,
71
Table 5
Correct Texting Responses by Stimulus Word
Word
Total Mean
(all Ss) (per S)
SD
(per S)
draw 626 5 - 2 1 7
2 . 8 2
must
555 *t.625 2.39
grin 551 i f. 592
2 .¥f
crop
5*f5 ^.5^2
2.57
Table 6
Analysis of Variance of Mean Number
of Correct Texting Responses
for the Four Stimulus Words
Source of Variance DF SS F
Groups— Words
Error
3 52.88 2.68 NS
k - 7 6 3131.60
72
Table 7
Cues Ss Reported Having Used in Identifying Words
Stimulus Words Grand
Total by
draw must grin crop Total
Category
(N=120) (N=120) (N=120) (N=120) (^*+80)
Letter-Sound
Consonant
—
Beginning 2 2 1 b
9
Middle 1 1
End 1 1 2
Middle Vowel 1 2 1 2 6 18
Letter-Hame -
Consonant
Beginning
13 31
16
19 79
Middle 6 6
5 3
20
End 11 6 4
7
28
Middle Vowel
5 3
10
19 37
l6* f
Visual Configuration
Consonant
Beginning
23 15
36 16 90
Middle 2 2 b
End
15 13
11 16
55
Middle Vowel b 1
5
2 12 161
Whole Word
31
28 22 22 1 0 3 103
Undetermined 8 11 6 9 3^ 31 *
73
recording, and interpreting what the S reports. The data
are reported in Table 7.
The responses concerning the cues the Ss reported
having used were analyzed. The breakdown of the h80 total
responses was as follows: individual letters were reported
to have been used 3^3 times, the whole word was used 103
times, and no specific cue could be identified 3* + times.
The breakdown of the individual letters indicates that: the
name and shape of the letter was reported to have been used
l6h - times; the visual-configuration of a letter (no name
could be given) was used 1 6 1 times; and the letter-sound
was identified 3*+ times. Ss reported using 178 beginning
and 85 ending consonants. Of the four major consonant-
configurational cues used in the stimulus words, the "g"
in grin and the ud" in draw were reported to have been used
the most and the 1 11" in must was used the least. This age
population apparently reports receiving the most help from
configurational cues when they appear at the beginning of
the word. The fact that there was no significant differ
ence (Table 6 ) between the number of correct texting
responses for any of the four words raises questions as to
the actual value of either the cues or the S’s report.
CHAPTER V
SUMMARY AND CONCLUSIONS
Summary
Purpose
The present study extends the research upon the
effects of discrimination pretraining in the early texting
response of children. Texting is the process of bringing
an appropriate vocal response under the control of a
visually presented verbal stimulus, i.e., "reading aloud."
In the visual domain, differences between consonant-
configurational cues and vowel cues were investigated. In
the auditory area, differences between letter-names,
letter-sounds, and no auditory pretraining were studied.
A factorial design also made possible an analysis of the
variance associated with the interaction between the visual
and auditory variables.
Procedure
The Ss were 120 first grade boys in the average IQ
range (85-115) and who were of normal first grade age. Ss
7b
were randomly assigned to eight groups of 15 and were
provided individual discrimination pretraining, familiari
zation, and texting training on four stimulus words: draw,
must, grin, and crop. All stimulus material was typed on
plain white 5 x 8 inch cards with a primary typewriter.
In pretraining, S was required to identify the stimulus
word that was different out of the three presented on each
of 12 discrimination pretraining cards. Three groups
received a configurational cue in red, and three a vowel
cue in red. One of each of these groups received training
on the letter-name of the cue, one on the letter-sound,
and one received no auditory training. Two control
groups were used; one received all stimulus words in
black and another received no discrimination pretraining.
The familiarization process included: hearing E
say the word, defining it, and repeating It after E.
Both the familiarization and texting training were the
same for all Ss. Each word was shown 10 times in random
order, k O trials in all. In training, a modified oper
ant learning technique was used. S was encouraged to say
the word before E. He was given immediate knowledge of
the accuracy of his response and appropriate verbal rein
forcement.
76
Findings
1. Visual pretraining on configurational cues without
auditory training produced the greatest amount of texting
transfer for this population.
2. For a combined auditory-visual approach, training
on vowel letter-names and letter-sounds produced more
correct texting responses than did similar training on
configurational cues.
3. Texting interference was associated with auditory
training on configurational cues, the value of which
appeared to he primarily visual.
There was no significant difference between train
ing on letter-names and letter-sounds.
5. Although the discrimination task did not appear to
facilitate texting, significant learning did take place in
all eight groups.
6 . The number of correct texting responses did not,
as was anticipated, vary with word difficulty.
7. There was no consistent relationship between the
number of correct texting responses and IQ. A very
significant difference in texting was found within Ss
despite appropriate experimental control.
8 . The verbal report of the Ss about the cue they had
used to identify the words did not appear to relate direct
ly to their texting response.
77
Discussion and Conclusions
Discrimination Pretrainine Groups
Several conclusions can be made about the effects
of discrimination pretraining on the early texting response
in the age group studied. Greatest transfer effect in
early texting is associated with visual pretraining on con-
figurational-consonant cues (CC) without concomitant audi
tory training. The superiority was seen both in the last
block of training trials and over the total texting task.
Similarly, the number of correct total texting responses was
significantly higher (.0 5) for the two groups which received
visual training on configurational (CC; and vowel (VC) cues
than the four groups who received the same visual cues in
combination with auditory training (LSC, INC, LSV, LNV).
This would tend to support earlier findings that children
primarily use configurational cues in early reading.
Contrary to common belief, beginning readers appear
to have difficulty integrating auditory-visual stimulation
in the texting task. The results of Muehl(1962) wherein
pretraining on letter names tended to interfere with word
recognition in kindergarten children are thus corroborated.
The letter names in Muehl* s study were designed to have a
mediational value rather than be a direct cue in the
picture recognition task. The present findings extend
this work to letter-names and-sounds which are direct cues
to beginning texting responses. Both studies suggest the
78
possibility that phonetic-visual instruction at this early
stage not only does not facilitate, but tends to interfere
with reading.
The data were analyzed to examine the effect of
increasing certain cue values by making them red. Six
groups received either a vowel or configurational cue in
red. A control group which received the stimulus words
entirely in black (DPC) made fewer correct discriminations
than any of the six groups. The mean number of correct
texting responses of DPC was among the lowest of the groups
at several points on the learning curve. Since none of
these differences was significant, however, the value of
the red cues in texting is equivocal. In order to read,
children need to be helped to make meaningful and continued
visual fixations. The transfer effect of techniques which
increase the stimulus value of cues needs further study.
The instructions given to three of the groups in
discrimination pretraining did not direct the S*s attention
to any specific part or letter in the word. One group saw
a configurational cue in red (CC), one a vowel cue in red
(VC), and the other saw all words in black (DPC). The
difference between the CC and DPC was significant at the
•05 level but the difference between VC and DPC did not
quite reach this level. This would tend to support the
theory that children discriminate on the basis of word
segments, such as specific letter configurations, rather
79
than the entire word- More information is needed about
the effects of pretraining on different word segments.
Although just visual training on configurational
cues was best (.0 1), the results indicate that a combination
of visual and auditory training on vowel cues (LSV and LNV)
produces more correct texting responses than does such
training on configurational cues (LSC and LNC). Further
more, the results would not indicate a greater transfer
effect associated with letter-sounds as opposed to letter-
names, even though the stimulus words were chosen to maxi
mize such differences. Thus the data would not support a
phonetic program which relates sounds to printed symbols,
especially configurational consonants. These findings
would support, however, a primarily visual approach using
phonic word attach shills, especially on vowels. in early
reading. Programmed reading texts that require students
to mahe meaningful visual discriminations to material based
on phonetic principles would appear to be using such an
approach.
Out of the six groups in the factorial design,
pretraining on the letter-name of a configurational cue
(INC) produced fewer (.0?) correct texting responses than in
the other five groups. The letter-sound configurational
group (LSC) was also low, although not significantly so.
The value of the configurational cue at this early stage of
reading, therefore, would appear to be essentially visual.
80
Interference with the transfer task takes place when train
ing is introduced on the letter-names or letter-sounds of
these configurational cues.
Measurement of Learning
It is also important to note that the reinforcement
and operant learning procedures produced a significant
amount of learning over a short period of time. The de
sign of the experiment does not permit direct comparison
with other teaching techniques. However, the Ss used in
this experiment were considered by their teachers to he
making very slow progress in reading. In some instances,
several different approaches had already been used with
little success. Considering the level of difficulty of
the words, an impressive amount of learning took place for
such a brief training period. Texting training which
utilizes immediate reinforcement and knowledge of results
merits further study. The ease with which difficult words
were texted suggests the possibility that this approach
might be especially valuable in helping children text
abstract words that have been found difficult in sight-
vocabulary techniques. These present findings lend
support to the statement of Staats and Staats (1962) that
information is needed on how to use the reinforcers commonly
available to the school to bring out desired texting
responses.
81
Observations during tbe experiment indicate that
certain factors reduced learning for many of the Sss
X. There was some indication that prior knowledge of
the meaning of the word tended to interfere with the texting
response. A S, for example, would respond "picture" to the
stimulus word "draw" 5 the partially correct association
would tend to result in a repetition of the wrong response.
2. As has been traditionally noted, when incorrect
responses were given they tended to be repeated. Ways of
avoiding the reinforcement of incorrect responses are
needed.
3 . Occasionally a S would associate the verbal rein
forcement ("right" or "wrong") with a stimulus word.
Reinforcement wherein only the correct word was repeated
might avert this difficulty.
h * It was difficult to vary the amount and speed of
reinforcement to maintain maximum attention and rapport.
Arranging of material so as to enable each S to produce
essentially only correct responses would tend to reduce the
need for such individual variation.
A large number of the Ss appeared to have difficulty
in focusing their visual attention on the stimulus word.
This factor would be closely associated with the variable
of motivation. Techniques are needed which will help the
child to make meaningful visual fixations and to increase
visual attention span. Once accomplished, this would tend
82
to facilitate the association of facts that the child knows
about a word with the appropriate visual imagery.
The Textual Response as it
Relates to Word Difficulty
One of the unexpected findings coming out of this
study was the fact that there was no significant difference
in the Ss* ability to text difficult as opposed to easy
words. Level of difficulty was measured by the S*s ability
to define the words as well as research indicating the num
ber of times they were spontaneously used by this age group.
Observation during the experiment suggested that a large
number of the Ss made little or no attempt to relate their
understanding about the word to the texting task. This
might be related to the above suggestion that knowledge of
the meaning of the word occasionally caused interference.
The child must eventually gain meaning from the
written word if it is to give him information. However,
these results suggest a possible emphasis on the separate
process of texting as the child makes his first adjustment
to reading (i.e., learning to read). This process could
proceed with a minimum of interference associated with the
many word meanings. After the child has experienced suc
cess from such a texting response-pat tern, other processes
could be utilized to help him gain the necessary meaning
from his reading (i.e., reading to learn). In recent years,
83
vocabulary in beginning readers has generally been care
fully controlled. The experience chart and the indepen-
dent-reading movements have raised questions about the
necessity of this control. More information is needed
about the intricate relationship between texting and word
meaning•
Value of Discrimination Pretraining
Another finding of interest is the apparent ease
with which the Ss performed the discrimination task. The
task did not appear difficult enough to necessitate careful
discriminations. Mistakes appear to have been generally
associated with lack of interest on the part of the Ss.
There is a discrepancy about the transfer value of
discrimination pretraining in this and other studies.
Spiker (1956, i9 6 0), Norcross & Spiker (1957), Spiker &
Norcross ( 1 9 6 2), and Muehl (1959, 1961, 1962) have reported
positive transfer with preschool and kindergarten Ss.
Staats, Staats, & Schuta (1962) report more equivocal
results with kindergarten subjects. Previous designs have
all used operant motor responses or matching rather than
texting. Hone has used the number of variables and con
trols incorporated in this design. No significant differ
ence was found between the groups which received discrimina
tion pretraining and the one that did not (NDP). Transfer
to a texting response may be influenced by different
8 3 4 -
factors than to a purely motor response. The discrepancy
may also be accounted for In the different developmental
levels of the populations studied. Discrimination which
brings about transfer at the kindergarten level may have
less effect on a first grade population.
Gibson, et al. (1962) have pointed out that helping
the child pay attention to distinctive features in letters
could be incorporated in reading readiness tests. "There
is little or no evidence that these experiences transfer to
reading. But if the typical matching tasks used variables
which are significant for letter discrimination (instead of
pictures or objects), there would certainly be greater po
tential transfer value" (p. 9 0 5)* The discrimination task
used in this study is more complex than that typically used
in reading readiness training. It is possible that more
difficult readiness material might produce more transfer
effect. Information about the amount of transfer effect
of readiness material is urgently needed. Once more is
known about this process, the material can be programmed and
presented in such a way as to maximize learning.
In general, this research points to the continuing
need for more complete identification of the variance asso
ciated with prediction in the reading process. The fact
that the approximated verbal IQ used in this study did not
relate to the number of correct texting responses would
support previous findings (Muehl, 1962; Staats, Staats, &
Schutz, 1962). If the IQ range had not been partially
restricted (85 to 115) in this experiment, more variation
of scores and consequently a correlation might have heen
obtained. The importance of the changing process of visual
perception as a possible source of variance has already heen
referred to. It has been suggested that perceptual
differences provide better prediction for success in early
reading than the IQ (see Catterall & Weise, 1959). The
importance of reinforcement techniques in studying motiva
tion has also been stressed. It vrould seem that the meth
odology reported in this study could provide valuable
criterion measures of learning which might facilitate
further investigation of these important variables.
BIBLIOGRAPHY
Alpern, G. D. Children’s performance in a socially rein
forced learning task as a function of two characteris
tics of the adult reinforcer. Dissertation Abstr.,
1 9 6 1, 2 1, 1993.
Antonitis. J. J., & Barnes. G. W. Group operant behaviors
An extension of individual research methodology to a
real life situation. J. genet. Psychol.. 1961, 98,
95-111.
Arnoult. M. D., & Lewis, J. T., III. Form discrimination
durxng brief exposures. Percent, mot. Skills. I960,
11, 259-260.
Baer, D. M. Escape and avoidance response of pre-school
children to two schedules of reinforcement withdrawal.
J. exp. Anal. Behav.. I960, 3, 155-159.
Baer, D. M. Effect of withdrawal of positive reinforcement
on an extinguishing response in young children. Child
Develpm.. 1961, 32, 67-7*+.
Baer, D. M. A technique of social reinforcement for the
study of child behaviors Behavior avoiding reinforce
ment withdrawal. Child Develpm.. 1962, 33, 8^7-858.
Baer, P. E., & Goldfarb, G. E. A developmental study of
verbal conditioning in children. Psychol. Rep., 1962,
10, 175-181.
Bijou, S. W. Patterns of reinforcement and resistance to
extinction in young children. Child Develpm., 1955*
26, 1 6 1-1 6 8.
Bijou, S. W., & Sturges, P. T. Positive reinforcers for
experimental studies with children: Consumables and
manipulatables. Child Develpm.. 19599 30, 151-170.
Blackwood, R. 0. Operant conditioning as a method of train
ing the mentally retarded. Dissertation Abstr.. 1963,
23, 297^ .
86
87
Bloomer, H. H. Stimulus properties of words: Variables
related to spelling difficulty. Unpublished doctoral
dissertation, University of Southern California, 1959*
Bloomer, B. H. Concepts of meaning and the reading and
spelling difficulty of words. J. educ. Res.. 1961, 5b9
178-186.
Boneau, C. A. The effects of violations of assumptions
underlying the t test. Psychol. Bill.. I960 , 57,
b 9 - 6 b ,
Brackbill, Yvonne, & Jack, D. Discrimination learning in
children as a function of reinforcement value. Child
Develnm.. 1958, 2 9, 185-190.
Brackbill, Yvonne, & Kappy, M. S. Delay of reinforcement
and retention. J. comp, physiol. Psychol.. 1962, 55,
lV-18.
Bullock, D. H. Some aspects of human operant behavior.
Psychol. Ree.. I960, 10, 2*+l-258.
Buss. A. H., & Buss, Edith H. The effect of verbal rein
forcement combinations on conceptual learning. J. exp.
Psychol.. 1956, 52, 283-287.
Campbell, B. J. Some effects of altering amount and type of
reinforcement on operant behavior of preschool children.
Dissertation Abstr., 1961, 21, 2791 *.
Cantor, G. N. The effects of three types of pretraining on
discrimination learning in pre-school children.
J. exp. Psychol.. 1955, *+9, 339-3^2.
Catterall, C. D., & Weise, P. A perceptual approach to
early reading difficulties. Calif. J. educ. Res..
1959, 10, 212-219.
Chansky, N. M. Learning: A function of schedule and type
of feedback. Psychol. Rep.« , I960 , 7 , 362.
Curry, Carolyn. Supplementary Report: The effects of
verbal reinforcement combinations on learning in
children. J. exp. Psychol.. I960, 59, b 3 b .
Davidson, Helen. A study of the confusing letters. B, D,
P, and Q. J^ genet. Psychol.. 1935, *+7, ^58-^68.
88
Durkin* Dolores. Reading instruction and the five-year-old
child. In J. A. Figurel (Ed.), Challenge and Experi
ment in Reading. New Yorkt Scholastic Magazine. 1962.
Bp. 23-27.
Durrell, D. D. (Ed.) Success in first grade reading. J.
Educ. Bost.. 1958, Ito, 1-U8.
Dvorine. I. Dvorlne Pseudo-Isochromatic Plates. Second
Edition, Baltimore, Marylands Waverly Press, 1953*
Edwards, A. L. Experimental design in -psychological
research. New York: Rinehart, I960.
English, B. , & English, Ava C. A comprehensive dictionary
of psychological and psychoanalytical terms. New Yorks
Longmans, Green, 1958.
Epstein, R. Verbal conditioning in children as a function
of sex-role identification, internal control, and need
for approval. Dissertation Abstr.. 1962, 22, 2*+62.
Epstein, R., & Liverant, S. Verbal conditioning and sex
role identification in children. Child Develpm.,
1963, 3^> 99-106.
Erickson, Marilyn A. Effects of social deprivation and
satiation on verbal conditioning in children.
Dissertation Abstr.. 1961, 22, 1722.
Erickson, Marilyn T., & Lipsitt, L. P. Effects of delayed
reward on simultaneous and successive discrimination
learning in children. J. comp, physiol. Psychol..
I9 6 0, 53, 256-260.
Fattu, N., Auble, D., & Mech* E. V. Partial reinforcement
in a bar-pressing situation with preschool children.
J. genet. Psychol.* 1955, 8 7, 251-255.
Fowler, W. Teaching a two-year-old to read; An experiment
in early childhood learning. Genet. Psychol. Monogg..
1 9 6 2, 6 6, 181-283.
Gates* A. I. Sex differences in reading ability. Elem.
Sch. J.. 1961, 61, ^31-^.
Ghent, Lila. Recognition by children of realistic figures
presented in various orientations. Canad. J. Psvehol.-
I960, m, 21+9-256. — ’
89
Gibson, Eleanor J*, Gibson, J. J.. Pick, Anne D.. & Osser.
H. A developmental study of the discrimination of let-
ter-like forms- J- comp, physiol. Psychol.. 1962, 55*
897-906.
Gladis, M. Grade differences in transfer as a function of
the time interval between learning tasks. J. educ.
Psvchol.. I960, 5 1 , 191-191 h.
Gleisman, D. Understanding in reading from the viewpoint of
sentence psychology. Read. Teacher. 1959, 13? 22-28.
Goldbeck, R. A. The effect of response mode and learning
material difficulty on automated instruction. Report
No. 1, September, 1960, Contract No. Nonr-3077 (00),
American Institute for Research, Pittsburgh.
Goldbeck, R. A., & Campbell, V. N. The effects of response
mode and response difficulty on programmed learning.
J. educ. Psychol.. 1962, 5 3 , 110-118.
Green, E. J., & Squier, R. W., Jr. The rate of learning
a discrimination in relation to some predictive
measures. Amer. J. Psychol. T 1962, 75? 29*1—297.
Hammack, J. T. Investigation of operant avoidance behavior
in children. Dissertation Abstr.. I960, 20, 2918-2919*
Hendrickson, Lois N., & Muehl, S. The effect of attention
and motor response pretraining on learning to discrimi
nate B and D in kindergarten, children. J. educ.
Psvchol.. 1962, 5 3 , 236-2*1-1.
Hicks, D. J. The effects of various reward and punishment
conditions on achievement behavior. Dissertation
Abstr., 1962, 23, 92?*
Hilgard, E. R. Thearies of learning. New Yorks Appleton-
Century-Crofts, 195©•
Hively, ¥. Implications for the classroom of B. F. Skinner*s
analysis of behavior. Harv. educ. Rev.. 1959, 29,
37-*+2.
Holmes, J. A. When should and could Johnny learn to read?
In J. A. Figurel (Ed.), Challenge and Experiment in
Reading. New York: Scholastic Magazine. . 1Q62.
Pp. 237-2*fl.
Jeffrey, W. E. Variables in early discrimination learnings
X. Motor responses in the training of a left-right
discrimination. Child Develpm.. 1958, 29, 269-275.
Jeffrey, W. E. Variables in early discrimination learnings
II. Mode of response and stimulus difference in the
discrimination of tonal frequencies. Child Develpm..
1958, 29, 531-538.
Jenson, A. E. Reinforcement psychology and individual
differences. Calif. J. educ. Res.. 1962, 13, 17^-178.
Johnson, R. C., & Zara, R. C. Relational learning in young
children. J. comp, physiol. Psvchol.. I960, 53,
59^-597.
Keislar, E. R., & McNeil, J. D. A comparison of two
response modes in an autoinstructional program with
children in the primary grades. J. educ. Psvchol..
1 9 6 2, 53, 132-1^3.
Kendler, H. H.« & Kendler, T. S. Vertical and horizontal
processes in problem solving. Psvchol. Rev.. 1962,
69, 1-1 6.
Kendler, T. S., & Kendler, H. H. Reversal and nonreversal
shifts in kindergarten children. J. exp. Psvchol..
1 9 5 9 , 5 8 , 5 6 - 6 0 .
Kennedy, W. A., Turner, A. J., & Lindner, R. Effectiveness
of praise and blame as a function of intelligence.
Percept, mot. Skills. 1962, 15, 1^3-l^.
Koch, Margaret B., & Meyer, D. R. A relationship of mental
age to learning-set formation in the preschool child.
J. comp, phvsiol. Psychol., 1959, 52, 387-389.
Krasner, L. Studies of the conditioning of verbal behavior
Psychol. Bull., 1958, 55, 1^-170.
Kurtz, K. H. Discrimination of complex stimuli! The
relationship of training and test stimuli in transfer
of discrimination. J. exp. Psvchol.. 1955, 50,
283-292.
LaPray, Margaret H. An investigation of the linguistic
approach to beginning reading with respect to word-
perception. Dissertation Abstr.. 1 9 6 2, 22, 3118.
Leiman? A. H., Meyers, J. L.. & Meyers, Nancy A. Secondary
reinforcement in a discrimination problem with children
Child Develpm.. 1961, 3 2, 3^9-353.
91
Lindquist, E. F. Design and analysis of experiments in
-psychology and education. Boston: Houghton Mifflin,
1953.
Lipsitt, L. P. Simultaneous and successive discrimination
learning in children. Child Develpm.. 1961, 32,
337-3^7.
Lipsitt, L, P., & Castaneda, A, Effects of delayed reward
on choice behavior and response speeds in children.
J. comp, -physiol. Psychol.. 1958, 51* 65-67.
Lipsitt, L. P., Castaneda. A., & Kemble, Joan D. Effects
of delayed reward pretraining on the discrimination
learning of children. Child Develpm.. 1959, 30,
273-278.
Long, E. R. An investigation of operant conditioning tech
niques for children. Privately issued terminal
progress report of Small Grant No. M-2007, Univ. of
North Carolina, 1958.
Long, E. R. The use of operant conditioning techniques in
children. In S. Fisher, (Ed.), Child research in
psvchopharmacology. Springfield, 111.: Charles
Thomas, 1959- Pp. 111-12^.
Long, E. R., Hanmack, J. T., May, F., & Campbell, B. J.
Intermittent reinforcement of operant behavior in
children. J. exp. Anal. Behav.T 1958, 1, 315-339.
Longstreth, L. E. Incentive stimuli as determinants of
instrumental response strength in children. J. compar.
physiol. Psvchol.. 1962, 55, 398-i +01.
Lynn, R. Individual differences in introver si on-extro
version, reactive inhibition, and reading-attainment.
J. educ. Psvchol.. I960, 51, 318-321.
McNeil. J. D.. & Keislar, E. R. Individual differences
and effectiveness of auto-instruction at the primary
grade level. Calif. J. educ. Res.. 1961, 12, 160-16^.
Meyer, W. J., & Seidman, S. B. Age differences in the
effectiveness of different reinforcement combinations
on the acquisition and extinction of a simple concept
learning problem. Child Develpm.. I960, 31,
Meyers, Nancy A. Extinction following partial and continu
ous primary and secondary reinforcement. J. exp.
Psvchol.. I960, 60, 172-179.
92
Miller. Louise B., & Estes, Betsy W. Monetary reward and
motivation in discrimination learning. J. exp.
Psychol.. 1961, 61, 501-504.
Mowrer, 0. H. The psychologist looks at language. Amer.
Psychologist. 1954, 9* 660-694.
Muehl, S. The effects of visual discrimination pretraining
on learning to read a vocabulary list in kindergarten
children. J. educ. Psvchol.. I960, 51, 217-221.
Muehl? S. The effects of visual discrimination pretraining
with word and letter stimuli on learning to read a ward
list in kindergarten children. J. educ. Psychol..
1961, 52, 215-221.
Muehl, S. The effects of letter-name knowledge on learning
to read a word list in kindergarten children. J. educ.
Psvchol.. 1962, 53, 181-186.
Muehl, S. The effects of letter-name knowledge on learning
to read a word list in kindergarten children. In J. A.
Figurel (Ed.) Challenge and Experiment in Reading.
New York; Scholastic Magazine. 19627 Pp. 128-133*
Norcross. Kathryn J., & Spiker, C. C. The effects of type
of stimulus pretraining on discrimination performance
in preschool children. Child Develpm., 1957, 28,
79-«4.
Otto, W. R. The acquisition and retention of paired asso
ciates by good, average, and poor readers. Disserta
tion Abstr.. 1961, 21, ^8 9.
Otto. W. R., & Fredricks, R. C. Relationship of reactive
inhibition to reading skill attainment. J. educ.
Psychol.. 1 9 6 3, 9 * , 227-230.
Price, L. E. The effect of type of verbal pretraining on
transposition in children. Dissertation Abstr.. 1961,
21, 2017.
Reese, H. W. Verbal mediation as a function of age level.
Psychol. Bull.. 1 9 6 2, 59, 5 0 2-5 0 9.
Rheingold. H. L., Gewirtz, J. L., & Ross, H. W. Social
conditioning of vocalizations in the infant. J. cornu.
physiol. Psychol.. 1959, 52, 6 8-7 3. —
Rieber, M. The effect of CS presence during delay of re
ward on the speed of an instrumental response, J, of
exn. Psychol., 1961, 61, 29Q-291f.
RjLnsland, H. D. A basic vocabulary of elementary school
children. Hew Yorks Macmillan, 19*+5*
Robinson, Helen, & Wright, B. News and Comments. Elem.
Sch. J.. 1959, 8 , 419-^36.
Robinson, Nancy M., & Robinson- H. B. A method for the
study of instrumental avoidance conditioning with
young children. J- comp, physiol. Psvchol.. 1961,
5*t, 2 0 -2 3 .
Rowan, Helen. fTis time he should begin to read. Car
negie Corp. N.Y. quart.. 1961, 9, 1-3.
Salzinger. K. Experimental manipulation of verbal behavior
A review. J. gen. Psvchol.. 1959, 61, 6j>-9l f.
Salzinger, Suzanne, Salzinger, K., Portnoy. Stephanie,
Eckman, Judith, Bacon, Pauline M.. Deutsch, M., &
Zubin, J. Operant conditioning of continuous speech in
young children. Child Develpm.. 1962, 33, 683-695*
Sechrest, L. Implicit reinforcement of responses. J.educ
Psychol.. 1963, 9 + , 197-201.
Skinner, B.F. Behavior of organisms, an experimental
analysis. New Yorks Appleton-Century-Crofts, 1938.
Skinner, B. F. Science and human behavior. New Yorks
Macmillan, 19!?3«
Skinner, B. F. Verbal behavior. New Yorks Appleton-
Century Crofts, 1957*
Skinner, B. F. Cumulative records. New Yorks Appleton-
Century-Crofts, l£59.
Spiker, C. C. Experiments with children on the hypothesis
of acquired distinctiveness and equivalence of cues.
Child Develpm.. 1956, 27, 253-261.
Spiker, C. C. Associative transfer in verbal paired-
associate learning. Child Develpm.. I960, 31, 71-87.
Spiker, C. C., & Norcross. Kathryn J. Effects of
previously acquired stimulus names on discrimination
performance. Child Develpm.. 1962, 3 3, 859-86*t.
9*+
Staats, A, W., & Staats, Carolyn K. A comparison of the
development of speech and reading "behavior with
implications for research* Child Develpm.. 1962, 33,
831-8^6.
Staats, Carolyn K*, Staats. A. W.. & Sehutz, R. E. The
effects of discrimination pretralning on textual
behavior. J. educ. Psvchol.. 1962, 53, 32-37*
Stevenson, H. W., & Weir, M. W. Response shift as a
function of overtraining and delay. J. comp, physiol.
Psychol., 1959, 52, 327-329.
Stevenson, H. W., & Weir, M. W. Developmental changes in
the effects of reinforcement and non-reinforcement of
a single response. Child Develpm.. 1961, 32, 1-5.
Suppes, P., & Ginsberg, R. Application of a stimulus
sampling model to childrens concept formation of
binary numbers with and without an overt correction
response^ Technical Report No. 35, December I960,
Institute for Mathematical Studies in the Social
Sciences, Stanford, California.
Terrell, G«, Jr. The role of incentive in discrimination
learning in children. Child Develpm.. 1958, 2 9,
231-236.
Terrell, G., Jr., & Kennedy. W. A. The discrimination
learning and transposition in children as a function
of the nature of the reward. J. exp. Psychol.. 1957,
53, 257-260.
Terrell, G., & Ware, R. Role of delay of reward in speed
of size and form discrimination learning in childhood.
Child Develpm.. 1961, 32, ^09-^15.
Ware, R., & Terrell, G. Effects of delayed reinforcement
on associative and incentive factors. Child Develpm..
1961, 3 2, 789-793.
Weir, M. W. Effects of age and instructions on children* s
probability learning. Child Develpm.. 1962, 3 3,
729-735.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

An Exploratory Study Of The Relationship Between Laboratory Sensitivity Training And The Self-Perceptions And Success Of Student Teachers

PDF

The Effect Of Subject Sophistication On Ratio And Discrimination Scales

PDF

Sound Blending: The Relationship Of Inter-Letter Interval, Consonant Category, And Stimulus Word Frequency To Word Identification

PDF

Variables Differentiating Mexican-American College And High School Graduates

PDF

The Effects Of Monetary Rewards On Reading Level, Attitude Toward Reading, And Self-Concept Of Ability And School Achievement

PDF

The Visual Fusion Threshold (Vft) Test As A Measure Of Perceptual Efficiency In Kindergarten And First Grade, And As A Possible Predictor Of Later Reading Retardation

PDF

Children'S Resistance To Competing And Distracting Stimuli In The Classroom

PDF

The Appropriateness Of Field And Level Of Vocational Choice As Related Toself-Concepts, Intelligence, School Achievement, And Socioeconomic Status

PDF

Classroom Climate And Pupil Characteristics In Special Classes For The Educationally Handicapped

PDF

An Experimental Study Of The Effect Of Anxiety And Stress On The Group Intelligence Test Performance Of Elementary School Children

PDF

Social System And Psychological Models In The Labeling Of Children As Educable Mentally Retarded

PDF

The Relative Efficiency Of Prompting And Confirmation Learning Paradigms

PDF

Concept Formation Ability Of 'Brain-Injured' Children Of Normal Intelligence

PDF

A Study Of The Effects Of Generalized Expectancies Upon Accuracy Of Interpersonal Perception

PDF

Self Concept And Social Class Correlates Of Contrasting Behavioral Subcultures Among Ninth Grade Boys

PDF

Sex-Role Identity, Self-Concept, And Vocational Interests Of Adolescent Hemophiliacs

PDF

The Relationship Of Selected Variables To The Ability Of First-Grade Children To Read And Interpret Maps

PDF

Some Effects Of Preschool Experience On The General Intelligence And Creativity Of Culturally Disadvantaged Children

PDF

Effects Of Success And Failure On Impulsivity And Distractibility Of Three Types Of Educationally Handicapped Children

PDF

Effects Of Preschool Enrollment And Parent Participation On Academic Growth

Asset Metadata

Creator Catterall, Calvin Dennis (author)

Core Title The Effects Of Pretraining In Auditory And Visual Discrimination On Texting In First Grade Boys

Degree Doctor of Philosophy

Degree Program Educational Psychology

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

Tag education, educational psychology,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Meyers, Charles Edward (committee chair), Brown, Charles M. (committee member), Lasswell, Thomas E. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-342913

Unique identifier UC11359088

Identifier 6413490.pdf (filename),usctheses-c18-342913 (legacy record id)

Legacy Identifier 6413490.pdf

Dmrecord 342913

Document Type Dissertation

Rights Catterall, Calvin Dennis

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