University of Southern California Dissertations and Theses

A self-knowledge model of social inference

(USC Thesis Other)

A self-knowledge model of social inference

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Running head: A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 1
A Self-knowledge Model of Social Inference
Minwoo Ahn
Department of Psychology
University of Southern California
August 2018
Master of Arts (PSYCHOLOGY)
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 2
Abstract
Mindreading has been one of the central topics in psychology for a long time. Recently,
there have been a few cognitive models of mindreading proposed, in which perceived
similarity of target to self plays a mediating role in social inference strategies (Ames,
2004a, 2004b; Tamir & Mitchell, 2013). In the current study, I propose a Self-knowledge
Model of Social Inference, where people make social inferences as a function of the degree
of predictability of the perceptual features of a target. Social inference becomes difficult
when the degree of predictability decreases. In the first phase of the study, I assessed
participants’ self-knowledge and used it to create three targets with the different levels of
predictability. I evaluated the accuracy and reaction times of social inferences in the
second phase. The current model accurately predicts the degree of accuracy and the
expected pattern of reaction times of social inferences.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 3
Contents
Abstract 2
Introduction 4
Social Inferential Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
The Self-knowledge Model of Social Inference . . . . . . . . . . . . . . . . . . . . 7
Experiment 10
Phase 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Apparatus and Stimuli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Phase 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Apparatus and Stimuli . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
General Discussion 17
References 20
Appendix A 25
Appendix B 30
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 4
Introduction
Having successful relationships is foundational for individuals’ lives. It is closely tied
to meaningful life indexes, such as wellbeing, happiness, sense of accomplishment, and so
forth. In order to have successful relationships, most people, at some point of life, start
asking how they can accurately infer the mental states of others, and how they can use
those judgments to guide their behavior. Accordingly, the underlying mechanisms involved
in mindreading have attracted many researchers in psychology. Researchers across varying
areas have investigated the topic for a long time, adopting various approaches to pursuing
the answer (Ames, 2004a, 2004b; Aron, Aron, Tudor, & Nelson, 1991; Davis, Conklin,
Smith, & Luce, 1996; Epley & Gilovich, 2006; Epley, Keysar, Van Boven, & Gilovich, 2004;
Fiske & Neuberg, 1990; Frith & Frith, 2006; Glucksberg & Keysar, 1990; Keysar, Lin, &
Barr, 2003; Kruger, Epley, Parker, & Ng, 2005; Ross, Greene, & House, 1977; Ross et al.,
1977; Tamir & Mitchell, 2013; Wimmer & Perner, 1983,?).
Most of them agree that the self plays a pivotal role in social inferences; people use
what they know about their own mind to infer the contents of others’ minds. This has
been repeatedly demonstrated in various cognitive models of social inference (Ames, 2004a,
2004b; Epley & Gilovich, 2006; Tamir & Mitchell, 2013). In this paper, I propose a
self-knowledge model of social inference that postulates that social inferences are a function
of the degree of predictability about identifying the perceptual features of a target. The
current model accurately predicts the accuracy and reaction times of social inferences.
Social Inferential Strategies
Perspective-Taking
For decades, there has been a consensus among psychologists that an ability to take
the psychological perspectives of others has important social consequences. Piaget (1965)
took it as a sign to overcome egocentric biases and observe and tailor one’s behavior to
meet the expectations of the society, Kohlberg (1976) considered it as a major
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 5
improvement in moral reasoning, Batson and Powell (2003) regarded it as a facilitator of
altruistic behaviors in social situations, and Richardson, Hammock, Smith, Gardner, and
Signo (1994) suggested the absence of the ability as a trigger of social aggression among
individuals.
Perspective taking also has been known to affect one’s thought, feeling, and
evaluation of others. Jones and Nisbett (1979) demonstrated that actors are likely to pay
more attention to situational factors that pull and push the target’s behaviors, while
observers often take the dispositions of the target as the explanation of behaviors. Davis et
al. (1996)showed that perspective-taking increases the extent to which a perceiver finds the
target similar to the self and tends to ascribe a great amount of self-descriptive traits to
the target. Galinsky and Moskowitz (2000) demonstrated that perspective taking results in
positive evaluations of a member of an out-group, while decreasing the accessibility of
stereotyping content in the evaluation.
Self-referential Process
There have been repeated reports that people use their own thoughts and feeling to
infer others’ mental states. Perceivers often mistakenly assume that a target could access
their private knowledge (Camerer, Loewenstein, & Weber, 1989; Keysar, Ginzel, &
Bazerman, 1995), they tend to overestimate that they can communicate better with close
friends than strangers (Savitsky, Keysar, Epley, Carter, & Swanson, 2011), and they often
ignore mutual knowledge while comprehending a speaker (Keysar, Barr, Balin, & Brauner,
2000). Perceivers also become self-centered when they don’t have sufficient cognitive
resources to interpret a target’s mental states (Lin, Keysar, & Epley, 2010), and they often
overestimate the extent to which their internal states are more accessible to others than
they really are (Gilovich, Savitsky, & Medvec, 1998).
Recent neurological evidence also reports that people tend to recruit self-knowledge
during social inference. Jenkins, Macrae, and Mitchell (2008) found that ventromedial
prefrontal cortex (vMPFC) activated when people introspected about themselves and when
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 6
they reasoned others perceived similar to self. Mitchell, Macrae, and Banaji (2006) found
that mentalizing about a similar other recruited vMPFC, whereas mentalizing about a
dissimilar other engaged a more dorsal subregion of mPFC. Patients with frontal lesions,
especially right medial frontal lesions showed difficulties with mindreading tasks (Rowe,
Bullock, Polkey, & Morris, 2001; Stuss, Gallup Jr, & Alexander, 2001). More evidence have
been found that the anterior paracingulate cortex, the superior temporal sulcus, and the
temporal poles are uniquely associated with mind reading process (Fletcher et al., 1995;
Gallagher et al., 2000; Gallagher, Jack, Roepstorff, & Frith, 2002; McCabe, Houser, Ryan,
Smith, & Trouard, 2001; Vogeley et al., 2001). These results together imply that perceivers
engaged in introspectin and social inference may share similar neural circuits and cognitive
processes.
Perceived Similarity
Ames (2004a, 2004b) proposed a similarity contingency model, in which perceptions
of similarity (i.e., perceived similarity) moderate two different social inference strategies,
projection and stereotyping, during the estimation of novel attributes of a target. Ames
argued that when perceivers have high levels of perceived similarity to a target, they are
likely to engage in introspection of their own attributes and ascribe those to the target.
When perceivers have lower levels of perceived similarity to a target, they engage in higher
levels of stereotyping, attending to implicit beliefs about particular group membership, not
the individual.
Tamir and Mitchell (2013) proposed an anchor and adjustment model where people
consider the self as a referent and a target as a subject to be compared. People in this
model correct away from the self to account for dissimilarities between self and other. The
adjustment process occurs in a serial fashion, and in their data it was present only when
the target was considered to be similar to the self. Epley et al. (2004) also found that
estimates of others’ perceptions were consistent with one’s own but differed in a manner
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 7
consistent with serial adjustment. They found slower reaction times when people evaluated
dissimilar others, compared to similar others.
In summary, the self is crucial and prevalent during social inferences. Social inferences
are those decisions in which the principles of (dis)similarity are involved. The processes of
making social inferences, however, remain poorly understood. Here, I introduce the
Self-knowledge Model of Social Inference, in which the degree of predictability about the
perceptual features of a target dictates the accuracy and reaction times in social inferences.
The Self-knowledge Model of Social Inference
I argue that social inference is inherently subjective and is the outcome of the degree
of predictability about identifying the perceptual features of a target. In short, a perceiver
encodes and evaluates perceptual cues of the target during an observation stage. In this
stage, the self likely becomes a referent, due to direction of comparison effects (Holyoak &
Gordon, 1983; Karylowski, 1989; Srull & Gaelick, 1983; Tversky, 1977) and its distinctive
availability, accessibility, and saliency in the memory retrieval process of person-perception
(Chambers, Windschitl, & Suls, 2003; Higgins, 1996; Schwarz, 1998, 2007; Solomon &
Vazire, 2016; Tversky & Kahneman, 1973; Vazire, 2010; Windschitl, Kruger, & Simms,
2003). When the perceiver is presented perceptual cues of the target, the associated
features in the observer are activated. The perceiver then assesses the degree of similarities
of those cues with respect to their personal beliefs, preferences, attributes and so on. Since
the encoded cues are inherently noisy, and the retrieved psychological constructs likely
overlap, the outcome includes some extent of uncertainty. Therefore, a degree of
uncertainty about identifying the perceptual similarities of the features is inherent in social
inference task. Throughout this process, the perceiver accumulates evidence over time,
constructing the perceived similarity of the target. Due to the subjective nature of the
evaluation process, individuals may not hold the same levels of perceived similarity about
the same target.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 8
I note that perceived similarity is influenced by many features (i.e., religious beliefs,
political backgrounds, group membership, traits, and so on). Those features in the brain
are represented both dimensionally and categorically, and they are processed and organized
based on similarities and differences in their properties and relations. Features in a
dimensional representation are bounded within a finite-dimensional class, in which a
feature at one end of the class is directly opposite of the feature at the other end. For
instance, an introvert directly differs from an extrovert, honesty is the opposite of
dishonesty, and humble goes against condescending. In this case, the observer can use the
dimensional information to contrast features between the self and the dissimilar target for
making inferences.
On the other hand, features in a categorical representation do not have such
dimensional information. Instead, features in this classification are grouped within a
non-dimensional class and contrasted with multiple options in the class. For instance, a
target’s dislike of basketball can be an objection to sports in general, but it could also be a
sign of the target’s liking of a different type of sports. Or the target’s statement that they
do not like Catholicism, doesn’t tell you whether they are against all religions or instead
whether they accept some other religion. In this case, one feature in the class is not
directly opposite other features in the class. Thus, manipulating the two different types of
representations would bring about varying results, especially when the target is dissimilar
to the self. In the current study, I manipulated dimensional representations (i.e., traits) to
generate targets with different levels of perceived similarity.
We propose that social inferences are made through sequential sampling processes,
such as random walk and diffusion procedures (Ratcliff and Rouder, 1998; see Figure 1). In
a random walk implementation, the observer encodes a stimulus and samples from his or
her associated mental representations of both the self and the target. The observer then
calculates a similarity measure, the extent to which the observer finds the encoded and
retrieved representations similar. Since the mental representations in comparison are
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 9
inherently noisy and overlap, a degree of uncertainty about identifying similarity between
the two is inherent in the task. With successive sampling, the perceiver adds the similarity
measure to the sum of the previous samples. This measure randomly walks toward either a
positive or a negative boundary over time (i.e., in the current experiment, ‘yes’ or ‘no’ as to
whether the target will perform the action described in a social inference task). When the
sum of the similarity measure reaches one of the two thresholds, a response is executed..
Importantly, it is the degree of uncertainty that drives different patterns of drift
toward the decision boundaries in the random walk implementation. If the features in the
assessment process are either similar or dissimilar in their properties and relations, the
observer quickly accumulates evidence during the sampling process. The observer then
makes quick and accurate response in the social inference task. In contrast, the observer is
slower and more error-prone when the features being compared are ambiguous.
Since social inferences are those decisions in which the principles of (dis)similarity are
involved, one relies on the most available and accessible resource to evaluate a target. The
perceiver, therefore, likely uses what he/she knows about him/herself to infer the mental
states of others. Importantly, the observer relies on self-knowledge for both similar and
dissimilar targets, due to their large predictability. As such, the observer directly applies
personal knowledge to the similar target, whereas the observer indirectly applies it to the
dissimilar target. For dissimilar targets, the observer mentally reverses self-knowledge to
construct the mental representation of the dissimilar target and uses it to infer the target’s
behavior. Thus, he/she would spend more time making a judgment on the dissimilar
target, and its accuracy may decrease. The perceiver would engage in different types of
strategies (i.e., stereotyping), only when the target is considered ambiguous. This is
assumed that the observer would start finding an alternative to resolve the issue. Given the
automatic properties of stereotyping, it could be a strong potential in social inferences.
A Random Walk Model of the Self-knowledge Model of Social Inference predicts that
the degree of uncertainty about identifying the perceptual similarities dictates the RT and
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 10
the accuracy of social inferences. When the degree of uncertainty is small, a perceiver likely
relies on self-knowledge to make inferences on both similar and dissimilar targets. The
small degree of uncertainty leads to smaller RTs and the larger accuracy of prediction rate.
The perceiver, however, will be subject to error-prone decisions, when the degree of
uncertainty is large (i.e., a target with ambiguous perceived similarity). Therefore, the
large degree of uncertainty leads to larger RTs and the smaller accuracy of social inferences.
In summary, I present the Self-knowledge model of Social Inference, in which a
perceiver subjectively constructs perceived similarity by sampling and assessing the
perceptual features of the targets. The speed and accuracy for making correct social
inferences is dictated by the degree of predictability involved in perceived similarity. To
assess the validity of the current model, I assessed self-knowledge of participants (Phase 1),
and testified the relation between the degree of uncertainty and RT and the accuracy of
social inferences (Phase 2).
Experiment
The current experiment consists of two phases; In Phase 1, I assessed participants’
self-knowledge using a personality inventory, and in Phase 2, I introduced a target and
asked the participants to evaluate the target’s personalities and to infer the mental states
of the target in a social inference task. I then asked the participants to infer their own
mental states about the same social inference task.
Phase 1
In Phase 1, I used the HEXACO personality questionnaire (Lee & Ashton, 2004) to
assess the traits of participants (i.e., self-knowledge). These estimates are the raw data 1)
that were used to generate three targets with different levels of perceived similarity (i.e.,
similar, dissimilar, and ambiguous), and 2) that were used to estimate the RTs and the
accuracy of a social inference task in Phase 2.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 11
Participants
63 undergraduate students at University of Southern California volunteered and
received course credit.
Apparatus and Stimuli
All stimuli were presented on a 24-in. LED color monitor with a 72-Hz refresh rate
controlled by Windows 7. The resolution of the monitor was 1, 920× 1, 200 pixels.
For stimuli, 60 statements from the HEXACO personality inventory were presented
one by one on the screen (i.e., the full version of the personality inventory includes 100
statements). These statements are designed to tap into six broad dimensions of
personality: Honesty-Humility, Emotionality, Extroversion, Agreeableness versus Anger,
Conscientiousness, Openness to Experience. Each scale includes four facets (for details,
please see Appendix).
On each trial, one statement was presented as text at the center of the screen. Below
the question was a visual rating scale that had five values: highly disagree, disagree,
neutral, agree, highly agree. Below the scale was a text box that accepts a participant
response. To submit a response, participants needed to click one of the values of the scale
and the accept box afterwards. The statement was presented in white, and both the scale
and the text box were in dark gray. The remainder of the computer screen was black.
Procedure
Participants were tested in a small dark room that contained a single computer. The
room was kept dark during the experiment.
Participants were instructed that they would be presented with a series of statements
about themselves. They were instructed to read each statement and to decide how much
they agreed or disagreed with the statement, using a scale below the statement. The
statement was randomly selected, without replacement, from the 60 statements. After they
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 12
chose one of the values on the scale, they clicked the accept box below the scale to advance
to the next trial.
The experiment consisted of seven practice trials and 60 experimental trials. The
practice trials were designed to teach participants the format of the experimental trials,
without informing him/her about the stimuli in the experimental trials. In each trial, the
participant’s response was recorded.
Results and Discussion
The goal of Phase 1 was to assess the participant’s scores on the HEXACO
personality inventory. These were the raw data used in creating and controlling targets
with different levels of perceived similarity in Phase 2.
Generating Three Targets
To manage the degree of perceived similarity of the target to the self, I manipulated
the extent to which each target had similar and dissimilar features of the participant. To
be specific, a similar target was assigned the identical traits of the participant. That is, the
similar target had the same responses on each item that participant judged in Phase 1. A
dissimilar target was assigned the opposing traits of the participant (i.e., the opposite
responses on each item of the personality inventory). Finally, an ambiguous target was
assigned half similar and half dissimilar features of the participant. I expected that the
participant in the ambiguous condition would experience the highest degree of uncertainty
due to the contradictory features of the target, whereas he/she would experience the lowest
uncertainty in both dissimilar and similar targets. The participant was trained about the
target’s features through a computerized personality task in the first block of Phase 2.
To evaluate the validity of the current model, I assessed the relation between the
degree of uncertainty in perceived similarity and RT and the accuracy of social inference in
Phase 2.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 13
Phase 2
Phase 2 consisted of three blocks. In the first block, I presented a computerized
personality task in which participants were introduced to one of three targets and were
taught about the target’s traits. In the second block, the participants were presented a
social inference task, where they were asked to infer the mental states of the target they
judged in the first block. After the task, the participants were promptly asked to rate the
extent to which they found the target similar or dissimilar to themselves. Finally, in the
third block, they were presented another social inference task, in which they estimated
their own mental states on the same items as were used in the second block. Since the
participants needed to learn about three different targets separately, I repeated the first
two blocks per target. The third block was presented only once at the end of Phase 2. The
sequence was intended to minimize the potential of experimenter bias.
I expected that participants would construct perceived similarity during judgments
on each target through the computerized personality task, and that the degree of
predictability involved in each target would affect the participants when they made social
inferences in the second block.
Participants
The participants from Phase 1 continued participating in Phase 2.
Apparatus and Stimuli
The apparatus used in Phase 2 is the same as that in Phase 1.
1st block: A Computerized Personality Task
I created a computerized personality task, in which a participant learned about a
target via a conversational format for presenting items of the HEXACO personality
inventory (see Figure 2). On each trial, I randomly sampled one statement from the
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 14
personality inventory and presented it to the participant. The participant was asked to
judge how descriptive the statement was of the target. The participant used the same
visual rating scale as Phase 1. After the judgment, he/she received feedback from the
target as to whether the judgment was accurate. If it was correct, the participant received
positive feedback (i.e., "Exactly. I find it quite boring to visit an art gallery"); if incorrect,
negative feedback (i.e., "Wrong. I find it quite interesting to visit an art gallery"). The
correct responses of the targets varied across participants, since the targets were generated
based on participants’ responses in Phase 1. A similar target was intended to have the same
responses to the statement of the HEXACO personality inventory as the participant; if the
participant submitted ‘highly disagree’ on item A, the similar target was also designed to
have the same response (i.e., ‘highly disagree’) on item A. Thus the participant received
more positive feedback as he/she judged that the target would be responding similarly to
the items. On the other hand, the participant likely received more negative feedback from
a dissimilar target, as he/she judged the target to be responding similarly to the items. An
ambiguous target was assigned half similar and half dissimilar responses of the participant;
thus, the participant would experience some degree of uncertainty, having difficulty
inferring the mental states of the target in the second block (i.e., social inference task).
For stimuli, I randomly sampled 30 statements from the stimuli used in Phase 1 (i.e.,
60 HEXACO personality inventory), without replacement. The questionnaire thus included
the same item as in Phase 1. The items this time described the target, instead of the
participant. The sampling repeated every time the participant had a new target.
2nd block: A Social Inference Task
To assess the influence of the degree of uncertainty involved in perceived similarity on
RT and the accuracy in a social inference task, I picked the 30 HEXACO personality
statements that were not presented in Phase 1 and modified them into social inference
questions. Typical examples of social inference questions were: "Do you think Bob cleans
his office or home quite frequently?", "Do you think Bob, if he wants something from a
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 15
person he dislikes, will act very nicely toward that person in order to get it?", "Do you
think Bob often sets ambitious goals for himself when he works?". On each trial, I
randomly sampled one question from the pool and presented it to participants. The
participants were asked to answer the question as quickly and accurately as possible by
clicking one of two dialogue boxes at the top of the screen. One of the boxes indicated ’yes’
and the other ’no’ (i.e., the position of the two boxes were counterbalanced). The mouse
cursor always started from the bottom center each trial. Once the participants clicked one
of the boxes, the screen proceeded to the next trial and the mouse cursor started from the
bottom center. After 30 trials, a global similarity question was presented (i.e., How similar
do you find Bob to you?). The participants answered the question using a 10-point Likert
scale, one is extremely dissimilar and ten is extremely similar.
3rd block: A Social Inference Task - Self-version
In the third block, I presented the 30 social inference questions, which were the same
questions as the second block. This time, however, the participants were asked to infer
about themselves, not the target.
Procedure
Participants were tested in a small dark room that contained a single computer. The
room was kept dark during the experiment.
The procedure of the third block remained the same as the second block, except that
the questions were about the participant, not the targets. The first and the second block
repeated each time the participants were presented with a new target. The third block was
presented only once after the first two block repeated three times. After the participants
finished the third blocks, they were greeted and were discharged.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 16
Results
I removed 0.9% of the individual trials because they were RT outliers (i.e., RT > 10
s). I then analyzed the remaining data.
I analyzed participants’ ratings on the global perceived similarity question as a
manipulation check. The participants had a mean rating of 2.6 out of 10 (SD = 1.26) for
the dissimilar target, a mean rating of 4.3 out of 10 (SD = 1.5) for the ambiguous target,
and a mean rating of 8.2 out of 10 (SD = 1.7) for the similar target. Repeated Measures
ANOVA revealed that there was significant difference between each target, F(2,124) =
233.7, p < 0.001.
The Self-knowledge Model of Social Inference accepts the premise that people anchor
on self and project or reverse what they know about themselves to make inferences on
others. To verify this assumption, I analyzed 1) the proportion of self-knowledge used in
the first block of Phase 2, and 2) the absolute difference between self-knowledge and the
participants’ judgments on the targets in the first block of Phase 2. I expected a positive
linear relation between perceived similarity and the proportion of self-knowledge and a
negative linear relation between perceived similarity and the absolute difference measure.
The results showed the expected patterns of behavior for both, the proportion of
self-knowledge in the first block of Phase 2, F(2, 124) = 117.9, p < .001 (Figure 5, and the
absolute difference between self-knowledge and the judgments on the targets, F(2, 124) =
112.5, p < .001 (Figure 4).
The current model predicts larger RT and the smaller accuracy of prediction rate for
a high degree of uncertainty (i.e., an ambiguous target) and smaller RT and the larger
accuracy of prediction rate for a small degree of uncertainty (i.e., similar and dissimilar
targets). As such, I expected the highest accuracy and the smallest RT in the similar
target, the second highest accuracy and the second smallest RT in the dissimilar target,
and the lowest accuracy and the largest RT in the ambiguous target. The results revealed
that there was a significant effect of perceived similarity on the accuracy, F (2, 124) =
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 17
129.4, p < 0.001 (Figure 7), but not on RT, F (2, 124) = 1.45, p = 0.24 (Figure 6).
Discussion
The results of Phase 2 were consistent with the majority of my hypotheses.
Participants anchored on and used self-knowledge to evaluate targets’ traits, and the
degree of uncertainty involved in perceived similarity dictated the accuracy of social
inferences. The current model, however, failed to find significant results in the RT measure.
One possibility is that the RT was contaminated by the reading time involved in the
current stimuli (i.e., sentences). As such, I found the expected pattern of RT (Figure 6).
This would improve if future studies use stimuli that do not require the participants to
spend much time in reading.
General Discussion
In this paper, I present the Self-knowledge model of Social Inference. This model
postulates that people anchor on the self and use what they know about themselves to
infer the mental states of others. The current model argues that the degree of
predictability of a target’s features predicts the accuracy of social inferences and the
amount of time spent on the task. To assess the validity of the current model, I estimated
self-knowledge in Phase 1, and assessed the relation between the levels of perceived
similarity and the dependent measures in Phase 2. The current model accurately predicted
the accuracy, but not the RT in the social inference task. The failure of the RT measure
was suggested to be the result of the contamination by reading time inherent in the current
stimuli. Except for the RT measure, all the other measures used in the current analysis
strongly support the hypotheses of the current model and provide a few advantages
compared to the contemporary models of social inference.
First, the current model provides a novel method that can accommodate individual
differences in controlling the levels of perceived similarity. Previous investigators have used
a number of different manipulations to generate similar or dissimilar targets, but all of them
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 18
need to screen and select participants who fit the pre-designed targets. The targets were,
also, generated by manipulating features that do not include clear perceptual boundaries,
such as beliefs, group membership, hobbies, and so on. This type of manipulation often
provides the participants with ambiguous perceptual cues, rather than dissimilar. For
example, one could recognize a statement about the target’s dislike of basketball as
objection to sports in general, but also it could mean that the target likes different type of
sports, while still remaining a huge fan of sports. One also could be enthusiastic about
politics having different political views from the participants. In the current research, I
manipulated the perceptual features of the targets (i.e., traits) and asked participants to
assess them using a measurement with diametrical ends (i.e., dissimilar vs. similar). Both
similar and dissimilar targets in the current experiment were distinguished from the
ambiguous target. This resulted in significant differences in my dependent measures.
Second, the current manipulation technique provides explicit evidence for
self-knowledge as a primary source during social inferences. Participants anchored on the
self while judging targets and making predictions on the targets in Phase 2. If the self did
not anchor social inferences, participants would not show such accurate prediction rates for
both dissimilar and similar targets. This result provides further support for the concurrent
models of perceived similarity.
Finally, the current model proposes a new approach to the literature of perceived
similarity. That is, the current model posits that the decision mechanism associated with
social inferences is a sequential sampling process, and self-knowledge is the main source for
both similar and dissimilar others. As predicted, the participants only showed lower
accuracy and larger RT in the ambiguous target, whereas they showed larger accuracy and
smaller RT in both similar and dissimilar targets. This evidence suggests that identifying
perceptual (dis)similarities is a central feature in social inferences, and that the
accumulated evidence from the sequential sampling process predicts both the accuracy and
RT in social inferences. These findings are expected to enhance our understanding about
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 19
social inferences.
Although the current study provides significant support for the validity of the
Self-knowledge Model of Social Inference, there are some limitations to the current work.
First, although I contend that a degree of uncertainty is central during social inferences, I
did not directly assess the degree of subjective perceptual uncertainty associated with the
stimuli used in Phase 1 and Phase 2. It may be valuable for future work to develop a new
method and use it to assess the degree of uncertainty in each statement.
Second, I made a hypothesis about RT in social inference, but the data did not follow
the prediction. It is proposed that the reading time involved in the current stimuli
contaminated the dependent measure. Thus, future work may benefit from designing an
experiment where participants are presented stimuli that minimize or eliminate reading
time.
Third, although my predictions result from the current model, I did not simulate my
data using either a diffusion or random walk model. Given that the nature of such models
aims at assessing the parameters of the underlying perceptual distributions, simulating
those models in this study is beyond the purpose of the study. Future work may expand
the current work by assessing the specific parameters involved.
The data from the current experiment supports the validity of the Self-knowledge
Model of Social Inference. People make social inferences as a function of the degree of
predictability of a target’s features. They use their own experiences as a guide during the
inference stage, and depending on the levels of perceived similarity of the targets to the
self, they directly and indirectly apply self-knowledge to the targets. Importantly, social
inference becomes difficult when the degree of predictability decreases. People seem to
engage in different types of strategies while inferring the experiences of the ambiguous
target. These findings not only provide further support to the previous research that posits
self-knowledge anchors social inferences, but also reveal the cognitive mechanisms involved
that have not been studied in social inferences.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 20
References
Ames, D. R. (2004a). Inside the mind reader’s tool kit: projection and stereotyping in
mental state inference. Journal of personality and social psychology, 87(3), 340.
Ames, D. R. (2004b). Strategies for social inference: a similarity contingency model of
projection and stereotyping in attribute prevalence estimates. Journal of personality
and social psychology, 87(5), 573.
Aron, A., Aron, E. N., Tudor, M., & Nelson, G. (1991). Close relationships as including
other in the self. Journal of personality and social psychology, 60(2), 241.
Batson, C. D., & Powell, A. A. (2003). Altruism and prosocial behavior. Handbook of
psychology.
Camerer, C., Loewenstein, G., & Weber, M. (1989). The curse of knowledge in economic
settings: An experimental analysis. Journal of political Economy, 97(5), 1232–1254.
Chambers, J. R., Windschitl, P. D., & Suls, J. (2003). Egocentrism, event frequency, and
comparative optimism: When what happens frequently is “more likely to happen to
me”. Personality and Social Psychology Bulletin, 29(11), 1343–1356.
Davis, M. H., Conklin, L., Smith, A., & Luce, C. (1996). Effect of perspective taking on
the cognitive representation of persons: a merging of self and other. Journal of
personality and social psychology, 70(4), 713.
Epley, N., & Gilovich, T. (2006). The anchoring-and-adjustment heuristic why the
adjustments are insufficient. Psychological science, 17(4), 311–318.
Epley, N., Keysar, B., Van Boven, L., & Gilovich, T. (2004). Perspective taking as
egocentric anchoring and adjustment. Journal of personality and social psychology,
87(3), 327.
Fiske, S. T., & Neuberg, S. L. (1990). A continuum of impression formation, from
category-based to individuating processes: Influences of information and motivation
on attention and interpretation. Advances in experimental social psychology, 23,
1–74.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 21
Fletcher, P. C., Happe, F., Frith, U., Baker, S. C., Dolan, R. J., Frackowiak, R. S., & Frith,
C. D. (1995). Other minds in the brain: a functional imaging study of “theory of
mind” in story comprehension. Cognition, 57(2), 109–128.
Frith, C. D., & Frith, U. (2006). The neural basis of mentalizing. Neuron, 50(4), 531–534.
Galinsky, A. D., & Moskowitz, G. B. (2000). Perspective-taking: decreasing stereotype
expression, stereotype accessibility, and in-group favoritism. Journal of personality
and social psychology, 78(4), 708.
Gallagher, H. L., Happé, F., Brunswick, N., Fletcher, P. C., Frith, U., & Frith, C. D.
(2000). Reading the mind in cartoons and stories: an fmri study of ‘theory of mind’in
verbal and nonverbal tasks. Neuropsychologia, 38(1), 11–21.
Gallagher, H. L., Jack, A. I., Roepstorff, A., & Frith, C. D. (2002). Imaging the intentional
stance in a competitive game. Neuroimage, 16(3), 814–821.
Gilovich, T., Savitsky, K., & Medvec, V. H. (1998). The illusion of transparency: biased
assessments of others’ ability to read one’s emotional states. Journal of personality
and social psychology, 75(2), 332.
Glucksberg, S., & Keysar, B. (1990). Understanding metaphorical comparisons: Beyond
similarity. Psychological review, 97(1), 3.
Higgins, E. T. (1996). Knowledge activation: Accessibility, applicability, and salience.
Holyoak, K. J., & Gordon, P. C. (1983). Social reference points. Journal of Personality
and Social Psychology, 44(5), 881.
Jenkins, A. C., Macrae, C. N., & Mitchell, J. P. (2008). Repetition suppression of
ventromedial prefrontal activity during judgments of self and others. Proceedings of
the National Academy of Sciences, 105(11), 4507–4512.
Jones, E., & Nisbett, R. (1979). The actor and the observer: Divergent perceptions of their
behavior. Edward E. Jones et al., Arrriburion: Perceiving the Causes ofBehavior, 25.
Karylowski, J. J. (1989). Trait prototypicality and the asymmetry effect in self-other
similarity judgments. Journal of Social Behavior & Personality.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 22
Keysar, B., Barr, D. J., Balin, J. A., & Brauner, J. S. (2000). Taking perspective in
conversation: The role of mutual knowledge in comprehension. Psychological Science,
11(1), 32–38.
Keysar, B., Ginzel, L. E., & Bazerman, M. H. (1995). States of affairs and states of mind:
The effect of knowledge of beliefs. Organizational Behavior and Human Decision
Processes, 64(3), 283–293.
Keysar, B., Lin, S., & Barr, D. J. (2003). Limits on theory of mind use in adults.
Cognition, 89(1), 25–41.
Kohlberg, L. (1976). Moral stages and moralization: The cognitive-developmental
approach. Moral development and behavior: Theory, research, and social issues,
31–53.
Kruger, J., Epley, N., Parker, J., & Ng, Z.-W. (2005). Egocentrism over e-mail: Can we
communicate as well as we think? Journal of personality and social psychology,
89(6), 925.
Lee, K., & Ashton, M. C. (2004). Psychometric properties of the hexaco personality
inventory. Multivariate behavioral research, 39(2), 329–358.
Lin, S., Keysar, B., & Epley, N. (2010). Reflexively mindblind: Using theory of mind to
interpret behavior requires effortful attention. Journal of Experimental Social
Psychology, 46(3), 551–556.
McCabe, K., Houser, D., Ryan, L., Smith, V., & Trouard, T. (2001). A functional imaging
study of cooperation in two-person reciprocal exchange. Proceedings of the National
Academy of Sciences, 98(20), 11832–11835.
Mitchell, J. P., Macrae, C. N., & Banaji, M. R. (2006). Dissociable medial prefrontal
contributions to judgments of similar and dissimilar others. Neuron, 50(4), 655–663.
Piaget, J. (1965). The moral judgment of the child (1932). New York: The Free.
Richardson, D. R., Hammock, G. S., Smith, S. M., Gardner, W., & Signo, M. (1994).
Empathy as a cognitive inhibitor of interpersonal aggression. Aggressive Behavior,
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 23
20(4), 275–289.
Ross, L., Greene, D., & House, P. (1977). The “false consensus effect”: An egocentric bias
in social perception and attribution processes. Journal of experimental social
psychology, 13(3), 279–301.
Rowe, A. D., Bullock, P. R., Polkey, C. E., & Morris, R. G. (2001). Theory of
mind’impairments and their relationship to executive functioning following frontal
lobe excisions. Brain, 124(3), 600–616.
Savitsky, K., Keysar, B., Epley, N., Carter, T., & Swanson, A. (2011). The
closeness-communication bias: Increased egocentrism among friends versus strangers.
Journal of Experimental Social Psychology, 47(1), 269–273.
Schwarz, N. (1998). Accessible content and accessibility experiences: The interplay of
declarative and experiential information in judgment. Personality and Social
Psychology Review, 2(2), 87–99.
Schwarz, N. (2007). Attitude construction: Evaluation in context. Social Cognition, 25(5),
638.
Solomon, B. C., & Vazire, S. (2016). Knowledge of identity and reputation: Do people
have knowledge of others’ perceptions?
Srull, T. K., & Gaelick, L. (1983). General principles and individual differences in the self
as a habitual reference point: An examination of self-other judgments of similarity.
Social Cognition, 2(2), 108.
Stuss, D. T., Gallup Jr, G. G., & Alexander, M. P. (2001). The frontal lobes are necessary
fortheory of mind’. Brain, 124(2), 279–286.
Tamir, D. I., & Mitchell, J. P. (2013). Anchoring and adjustment during social inferences.
Journal of Experimental Psychology: General, 142(1), 151.
Tversky, A. (1977). Features of similarity. Psychological review, 84(4), 327.
Tversky, A., & Kahneman, D. (1973). Availability: A heuristic for judging frequency and
probability. Cognitive psychology, 5(2), 207–232.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 24
Vazire, S. (2010). Who knows what about a person? the self–other knowledge asymmetry
(soka) model. Journal of personality and social psychology, 98(2), 281.
Vogeley, K., Bussfeld, P., Newen, A., Herrmann, S., Happé, F., Falkai, P., ... Zilles, K.
(2001). Mind reading: neural mechanisms of theory of mind and self-perspective.
Neuroimage, 14(1), 170–181.
Wimmer, H., & Perner, J. (1983). Beliefs about beliefs: Representation and constraining
function of wrong beliefs in young children’s understanding of deception. Cognition,
13(1), 103–128.
Windschitl, P. D., Kruger, J., & Simms, E. N. (2003). The influence of egocentrism and
focalism on people’s optimism in competitions: when what affects us equally affects
me more. Journal of personality and social psychology, 85(3), 389.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 25
Appendix A
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 26
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 27
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 28
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 29
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 30
Appendix B
Figure 1. The visualization of the Random Walk implementation of the Self-knowledge
model of Social Inference. A perceiver encodes, assesses, and accumulates evidence
throughout the sequential sampling process. The drift rate indicates the degree of
uncertainty about identifying the perceptual similarities of features of both the observer
and a target. The observer is expected to show large reaction times and low accuracy in
the social inference task as the target becomes ambiguous. Importantly, the observer uses
self-knowledge for both similar and dissimilar targets in the decision process. The perceiver
executes a response when the accumulated evidence hits one of the decision thresholds.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 31
Figure 2. The time line of a trial for a computerized personality task. "a" The screen shots
of a computerized personality task. "b" The participant was presented with a statement
and was asked to judge the statement. After the submission of a response, the participant
received feedback from the target, as to whether his/her judgment was correct.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 32
Figure 3. The time line of a trial for a social inference task. The participant was presented
with a question and was asked to make inferences in relation to a target. After the
submission of a response, the screen proceeded to the next trial.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 33
Figure 4. The absolute difference between the participants’ self-knowledge and their
judgments on the targets in the first block of Experiment 1-b. As expected the dissimilar
target produced the largest difference in their responses.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 34
Figure 5. The proportion of self-knowledge used in the first block of Experiment 1-b.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 35
Figure 6. Reaction time (RT) from the second block of Experiment 1-b.The failure of
obtaining a significant result in the measure is assumed by the reading time involved in the
current stimuli.
A SELF-KNOWLEDGE MODEL OF SOCIAL INFERENCE 36
Figure 7. The accuracy of social inference in the second block of Experiment 1-b. The
degree of uncertainty involved in perceived similarity predicts the lowest accuracy in the
ambiguous target and the highest accuracy in the similar target.

Abstract (if available)

Abstract Mindreading has been one of the central topics in psychology for a long time. Recently, there have been a few cognitive models of mindreading proposed, in which perceived similarity of target to self plays a mediating role in social inference strategies (Ames, 2004a, 2004b

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Bayesian hierarchical and joint modeling of the reversal learning task

PDF

Performance and attention novelty slows hedonic adaptation during habit formation

PDF

The spread of moral content in online social networks

PDF

Using Pavlovian cues to boost self-control

PDF

Reframing parental psychological control: the role of insecure attachment

PDF

Psychological distance in the public’s response to terrorism: An investigation of the 2016 Orlando nightclub shooting

PDF

Interpersonal coping responses during adolescence: implications for adjustment

PDF

The role of primary care in the use of specialty mental health services: an investigation of utilization patterns among African Americans and non-Hispanic Whites

PDF

Depression and anxiety symptom outcomes in adolescent users of smoked, vaporized, edible and blunt cannabis

PDF

People can change when you want them to: changes in identity-based motivation affect student and teacher Pathways experience

PDF

Preparing for natural disasters: investigating the effects of gain-loss framing on individual choices

PDF

Making appropriate decisions for nonnormal data: when skewness and kurtosis matter for the nominal response model

PDF

Fathers’ social-cognitive network connectivity: associations with prenatal father-infant attachment and postpartum parenting behavior

PDF

Nice by nature? A twin study of the development and physiology of prosocial personality

PDF

Cross-ethnic friendships, intergroup attitudes, and intragroup social costs among Asian-American and Latino-American youth

PDF

Dynamics of victimization, aggression, and popularity in adolescence

PDF

Meta-accuracy as a moderator of the association between social experience and emotional adjustment during childhood

PDF

Psychological correlates of certainty strength: measuring the relative influence of evidence, opinion of personal contacts & importance to self-identity

PDF

Choice biases in making decisions for oneself vs. others

PDF

Effects of childhood adversity on physiology and health in emerging adulthood

Asset Metadata

Creator Ahn, Minwoo (author)

Core Title A self-knowledge model of social inference

School College of Letters, Arts and Sciences

Degree Master of Arts

Degree Program Psychology

Publication Date 08/06/2018

Defense Date 08/03/2018

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

Tag impression formation,OAI-PMH Harvest,person-perception,self-knowledge,social inference

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Read, Stephen (committee chair), John, Richard (committee member), Monterosso, John (committee member)

Creator Email minwooah@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-55242

Unique identifier UC11671665

Identifier etd-AhnMinwoo-6666.pdf (filename),usctheses-c89-55242 (legacy record id)

Legacy Identifier etd-AhnMinwoo-6666.pdf

Dmrecord 55242

Document Type Thesis

Format application/pdf (imt)

Rights Ahn, Minwoo

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 a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA