 

 

Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

COMPUTER-MEDIATED IMPRESSION FORMATION:
A TEST OF THE STICKY CUES MODEL USING FACEBOOK

presented by

Brandon Lee Van Der Heide

has been accepted towards fulﬁllment
of the requirements for the

Doctoral degree in Communication

 

 

kledaaz;

/ ’ Major Professor’s Signature
27 ¢~*{\1 2001
/ I '

 

 

Date

MSU is an Afﬁrmative Action/Equal Opportunity Employer

 

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K lPrq/Acc&Pres/CIRC/DaleDue indd

 

COMPUTER-MEDIATED IMPRESSION FORMATION:
A TEST OF THE STICKY CUES MODEL USING FACEBOOK

By

Brandon Lee Van Der Heide

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Communication

2009

ABSTRACT
COMPUTER-MEDIATED IMPRESSION FORMATION:
A TEST OF THE STICKY CUES MODEL USING F ACEBOOK
By
Brandon Lee Van Der Heide
This research offers a model of online impression formation that explains how different
impression-bearing cues may carry more or less informational value. This research
considers the possibility that impression-bearing cues have greater informational value
when those cues are distinctive and are task-relevant. This research refers to such cues as
sticky cues. Further, this research suggests that sticky cues may help to describe how
interpersonal and categorical cues vary in terms of the amount of impression-bearing
information they provide to observers. This research reports two original experiments
that varied both the distinctiveness of interpersonal and group cues and the task-
relevance of those cues. This research examined the effects of group cues on judgments
of a target’s intelligence and the effects of interpersonal cues on judgments of a target’s
extraversion. The results were consistent with the sticky-cues hypothesis with regard to
interpersonal cues to extraversion, but only cue distinctiveness (and not cue relevance)
were effective group cues that informed participants’ judgments of a target’s
intelligence. These ﬁndings are discussed in light of other theoretical perspectives on
impression formation in computer-mediated communication and future research

directions are discussed.

This dissertation is dedicated to my wife, Jen Van Der Heide, whose love, patience, and

support made this project (and so many other things) possible.

iii

ACKNOWLEGEMENTS

I would like to thank several people who have made this project possible. First, I
would like to thank my parents, Bill and Marge Van Der Heide, for instilling in me a
special reverence for learning about the world around me, for helping me to develop
conﬁdence that I could achieve most anything at which I worked diligently, and for doing
all they could to make sure that any barriers between me and my dreams would be
surmountable.

I would also like to thank my doctoral committee members, Dr. Chuck Atkin, Dr.
Frank Boster, and Dr. Cliff Lampe. I would like to thank Chuck for his wisdom. Chuck
was an irreplaceable asset not only because of his intellectual prowess, but also for his
wisdom when it came to making tough job decisions. Chuck cared so deeply about the
success of the graduate students in the department—that may be what I learned most
from him. Chuck, I am grateful for the care you took of my fellow students and me. 1
may have gotten my degree without you ensuring that things would work out, but I doubt
I would have enjoyed it so very much. I am also grateﬁJl to Frank because his door was
always open. Whether it was statistical consultation, methodological advice, or other
theoretical guidance, I always left Frank’s office or classroom knowing more about being
a scientist than I did when I arrived. The ratio of things learned to time spent with Frank,
as many students can attest, is very high. I believe this ratio to be both a reasonably
reliable and exceptionally valid indicator of quality; Frank you are among the best and I
am honored to have learned from you. I am also deeply indebted to Cliff for his constant

pressure to get me to be more practical. Cliff, I can say with conﬁdence that you will be

iv

among the most formative of my professional inﬂuences. Your attention to detail, ability
to raise questions, and suggest possible applications of my work (applications that I had
never even dreamed that my work could apply to) will forever put me in your debt.

There are many inﬂuential people in my life, but for the past three years the most
inﬂuential person has been, without a doubt, Dr. Joe Walther. I anticipate that your
inﬂuence, Joe, will live on with me forever. When people who don’t know about
computer-mediated communication ask me about my advisor, it’s hard to know what to
tell them. Lately, I’ve been using this metaphor: “Working with Joe is a lot like learning
to play basketball from Michael Jordan, but you don’t just go to the gym every once and
a while and have him give you tips on how to shoot free throws. What actually happens is
you spend about 70 hours a week with him, he talks not only about the fundamentals of
the game, but also about how to deal with an agent, how to negotiate a contract, he cooks
smoked pork loin for you, buys you beer, emails you at 3am just to see what’s up
(because he knows you’re awake watching game-tape. . .and he’s watching it too), takes
care of your kid when he can, and loves your family like they’re his own. And, when you
blow a key game by making an avoidable mistake, he’s there to chuckle and make sure
you know enough not to blow it again.” I can never repay you, Joe, for all you’ve done
for and given to me. The best I can hope for is that, one day, I’ll be able to do for
someone else a bit of what you did for me. I won’t squander that chance.

I also am indebted to various research team members who made long hours in the
lab seem not quite so long. Most prominently I wish to thank Dave Westerman, Stephanie
Tong, Caleb Carr, and Dave DeAndrea. I also wish to thank Dr. Elizabeth Vander Lei

whose consistent encouragement (read: ever-present chiding) helped me to stretch myself

to become more than I ever dreamed; I don’t know what you saw in me, but thanks for
seeing it.

Finally, I wish to thank my wife and son—Jen and Liam. Jen, your patience and
support has been epic and deserves to be the stuff of legend. You agreed, over coffee, to
put your life on hold so that I could be personally and professionally fulﬁlled. When we
decided to do this, I do not know if either of us grasped the sheer enormity of what we
were undertaking, but I do know that if it were not for your incredible love (and general
husband-management skills) I would never be where I am today. You have allowed me to
be more than I could have ever dreamed that I could become. We’re not quite there yet,
but I’m sensing chocolates (the kind you eat pool-side) in your future. Liam, you don’t
fully understand it yet, but you made many sacriﬁces along the way—most notably time
spent with your Dad. Nonetheless, when I came home from days away from you at
school or at a conference you treated me like I never leﬁ. You truly are the coolest kid

ever; I love you so much.

vi

TABLE OF CONTENTS

LIST OF TABLES ....................................................................................................... ix

LIST OF FIGURES ....................................................................................................... x

CHAPTER 1 .......................................................................................................... ' ........ l
Computer-Mediated Impression Formation: A Test of the Sticky Cues Model

Using Facebook ................................................................................................. 1

F acebook as an Impression Formation Platform ............................................... 2

Factors Affecting Impression Formation .......................................................... 4

Interpersonal Impression Utility ............................................................ 5

Impression Bearing Information Seeking .............................................. 6

Cue Informational Value ................................................................................. 10

Interpersonal Vs. Group Impression Formation .............................................. 15

Testing the Sticky Cues Model ....................................................................... 18

CHAPTER 2 ................................................................................................................ 22

Study 1 ............................................................................................................. 22

Method ............................................................................................................. 22

Procedure and Materials ...................................................................... 22

Participants and Design ....................................................................... 24

Dependent Measures ........................................................................... 24

Induction Check .................................................................................. 25

CHAPTER 3 ................................................................................................................ 28

Results ............................................................................................................. 28

Discussion ....................................................................................................... 29

CHAPTER 4 ................................................................................................................ 32

Study 2 ............................................................................................................. 32

Method ............................................................................................................. 33

Procedure and Materials ...................................................................... 33

Participants and Design ....................................................................... 35

Dependent Measures ........................................................................... 36

Induction Check .................................................................................. 37

CHAPTER 5 ............................................................................................................. 40

Results ............................................................................................................. 40

Discussion ....................................................................................................... 40

CHAPTER 6 ............................................................................................................. 42

' General Discussion .......................................................................................... 42

vii

APPENDICES ............................................................................................................. 52

REFERENCES ............................................................................................................ 66

viii

LIST OF TABLES

TABLE 1: ITEM — FACTOR LOADINGS ............................................ 61
TABLE 2: CONTRAST WEIGHTS FOR PREDICTIONS .......................... 62
TABLE 3: MEANS (AND STANDARD DEVIATIONS) .......................... 63

ix

LIST OF FIGURES

FIGURE 1: STUDY 1 SAMPLE STIMULUS MATERIAL .................................. 64

FIGURE 2: STUDY 2 SAMPLE STIMULUS MATERIAL ................................. 65

Chapter 1
COMPUTER-MEDIATED IMPRESSION FORMATION:
A TEST OF THE STICKY CUES MODEL USING FACEBOOK

Research on interpersonal impression formation in computer-mediated
communication (CMC) has examined a variety of topics including whether or not
impressions form in CMC (Kiesler, Siegel, & McGuire, 1984); whether CMC inclines
people to form group rather than interpersonal impressions (Reicher, Spears, & Postmes,
1995); whether online impressions are like offline impressions (Jacobson, 1999); at what
rate (Walther, 1992, 1993) and under what conditions impressions form online (Walther,
1994); and the breadth and depth of impressions that form in CMC (Hancock &
Dunham, 2001 ). While past research has helped researchers to understand how
individuals make interpersonal judgments about others online, less research has
examined whether all information that informs the impressions people form of one
another online is of equal impression-bearing value.

This research seeks to understand what makes different types of information
more or less useful in helping people to form impressions in CMC. In accomplishing
this goal, this research proposes a model of impression formation that allows for
impressions to form on the basis of particularly useful impression bearing information.
Speciﬁcally, this piece advances the argument that, in some situations, individuals look
for particularly potent cues, which this research coins sticky cues, in order to form an
impression of another person efﬁciently even in an environment such as in CMC where
interpersonal information is less prevalent than in traditional face-to-face interactions.

This work ﬁrst explores some basic assumptions about the nature of impression

formation. Working from these basic assumptions, this work suggests several
dimensions upon which impression-bearing cues are proposed to vary. Finally,
hypotheses about the types of information (or cues) that will have the most effect on
impression judgments are forwarded and tested using stimuli comprised by mockups of
proﬁles from the popular social networking website, Facebook. Results supported the
hypothesis that distinctive cues that were task-relevant together produced the most
extreme extraversion judgments, while distinctive group cues alone produced the
expected intelligence judgments.
F acebook as an Impression Formation Platform

boyd and Ellison (2007) deﬁne social network websites as web-based services
that allow individuals to publicly or semi-publicly share information about themselves
using a pre-formatted proﬁle page, form connections with other individuals, and peruse
these “friends” proﬁles and connections. Facebook is among the most popular of these
social network sites. Research has begun to explore how different cues (i. e., pieces of
information about a person) affect impression formation in this computer-mediated
environment. For example, Walther, Van Der Heide, Kim, Westerman, and Tong (2008)
explored how two elements of individuals’ personal proﬁle page, their ﬁiends’
photograph and the messages their friends left on the profile owners’ page affected
impressions of individuals’ physical attractiveness. Speciﬁcally, this research found that
a proﬁle owner’s friends’ physical attractiveness affects a perceiver’s judgment of the
proﬁle owner’s physical attractiveness—proﬁle owners with physically attractive friends
were judged to be more physically attractive than proﬁle owners with physically

unattractive friends. This research also suggested that what a proﬁle owner’s friends say

about him or her can affect judgments of a proﬁle owner’s physical attractiveness. When
male proﬁle owners’ friends left negatively valenced comments about certain moral
behaviors on their proﬁles, proﬁle owners were judged to be more physically attractive
than if their friends leﬁ positively valenced comments; however, this relationship was
reversed for female proﬁle owners.

While some research has begun to explore how impressions of a person form on
the basis of a F acebook proﬁle, much research into impression formation in CMC has
examined the effects of relatively few isolated cues on impression formation. Facebook
provides an interesting venue to study impression formation in CMC because it allows
researchers to begin to understand how perceivers utilize social network site proﬁles that
contain a large number of cues about a proﬁle owner in order to form impressions of a
target. Online venues such as F acebook allow for a individuals to send large numbers of
messages to perceivers, and they allow for a perceiver of a target proﬁle to assess people
not only on the basis of what they say about themselves but also what others say about
them and the information presented about them as aggregated by the technological
system (c. g., the number of friends one has). Previous research has explored and
enumerated several different types of cues that may exist in social network environments
such as self-generated, other-generated, and system-generated cues (Tong, Van Der
Heide, Langwell, & Walther, 2008; Walther et al., 2008; Walther, Van Der Heide,
Hamel, & Shulman, 2009). This work builds upon these perspectives to help describe
how perceivers select pertinent cues from among the large amount of information

available to form an impression of a target.

Additionally, the impression formation process may be unique on F acebook
because of the sheer amount of information present on proﬁle pages. Traditionally,
impression formation processes have been examined by exploring how an impression of
a target develops quickly (and accurately) on the basis of relatively few cues about the
target (Albright, Kenny, & Malloy, 1988; Ambady & Rosenthal, 1993). Impression
formation in the context of a computer-mediated social network proﬁle raises new
questions about how people form focused impressions of a target in the face of a
veritable wall of cues about a target’s identity.

Factors Affecting Impression Formation

To explore the factors that impact the potency of impression—bearing cues, this
research begins by reviewing some key perspectives from interpersonal communication
that help to explain the types of information that are judged by perceivers to be useful
when forming impressions and how perceivers may go about seeking out particularly
useful impression-bearing cues. Next, this research builds on Gigerenzer and Goldstein’s
(1996) probabilistic mental models theory and Kelly’s (1967, 1973) distinctiveness
information to inform a model of how particularly potent impression-bearing cues affect
impression formation in CMC.

Interpersonal Impression Utility

Do people always ﬁnd interpersonal impressions useful, or do situations vary in
the degree to which people ﬁnd having an interpersonal impression useful? In their
classic text, Miller and Steinberg (1975) argued that interpersonal impressions of others

are not necessary for navigating every social situation. There are some situations where

unique knowledge of a person’s idiosyncratic qualities is useful, while other situations
require only more general sociological or cultural-level knowledge about a person.

Miller and Steinberg (1975) describe three distinct levels of knowledge upon
which people may make predictions about how others will behave: the cultural level, the
sociological level, and the psychological level. From this Miller and Steinberg propose
that the way people interact with others is informed by the predictions people make on
the basis of these different sorts of information. The most general level of information
about others, cultural information, is deﬁned as “the sum of characteristics, beliefs,
habits, practices and language shared by a large group of people” (p. 12). Sociological
level predictions rely on speciﬁc information about a person’s social group membership.
A membership group is “a class of people who share certain common characteristics,
either by their own volition or because of some criteria imposed by the predictor” (p.
1.7). Finally, psychological level predictions presume that one has knowledge about the
unique idiosyncrasies associated with a speciﬁc interaction partner.

If situations vary in the degree to which forming an interpersonal impression is
useful and economical, one can extrapolate that the most reﬁned interpersonal
predictions—psychological level predictions—are not necessary for every situation. An
operating assumption of this research is that the usefulness of forming an impression of
another person varies across situations. For example, if one were seeking a potential
romantic partner on an online dating website, one may be highly motivated to learn
about interpersonal characteristics such as physical attractiveness and likes or dislikes.
If, prior to a vacation, a person consulted an online discussion forum about what the

weather was like in the Cayman Islands in March, it may be sufﬁcient to know that a

person who answered that question was a resident of the Cayman Islands—knowledge
of their music preferences, food allergies, and physical attractiveness may be of little
interest to a perceiver. Let us refer to the degree to which it is useful in an interpersonal
encounter to form a psychological-level impression of another individual with whom
one is interacting as the interpersonal impression value of a given situation. If situations
vary in terms of how useful it is to form an interpersonal impression of others, it seems
logical to ask: When people are in situations for which an interpersonal impression is
useful, do they also seek out speciﬁc impression bearing information in order to form
the impressions needed to navigate these social situations? This work now turns to a
discussion of some ﬁndings that suggest that people may seek information that best
helps them to traverse interpersonal encounters.
Impression Bearing Information Seeking

Previous research has suggested that people form impressions that help them
navigate social situations. Hancock and Dunham (2001) argued that, in CMC, people are
able to form impressions that, although more narrow than face-to-face (F TF)
impressions, are more extreme. In making this conclusion their research examined
impression formation among college students who were completing an object-matching
task. In a one-shot interaction, one participant was instructed to describe a series of
tangrams to another participant who sought to identify the described tangram from an
array of possible solutions. Their research examined the effect of communication mode
on impression extremity (e. g., how extremely introverted or extraverted a perceiver
rated a target) and impression breadth (e. g., how many items per personality

characteristic a perceiver made some judgment rather than indicating that had

insufﬁcient information to make a judgment.) Hancock and Dunham found that
perceivers rated each targets’ traits (Openness, conscientiousness, extraversion,
agreeableness, and neuroticism) more extremely on Costa and McCrae’s (1991) NEO-
Five Factor Inventory (N EO-FFI) personality measurement in CMC than in FTF
interactions. However, the breadth of impressions formed was greater in F TF
interactions. That is, they found that, when participants were given the option of rating a
partner on each item of the NEO-FFI or indicating that they could not answer the
question, in FTF interactions people were able to indicate judgments on more items of
the NEO-FFI than in CMC. Thus, because participants were able to express a greater
number of impression judgments, Hancock and Dunham concluded that the breadth of
impressions formed was greater for FTF compared with CMC interactions. Although
this effect persisted for judgments of extraversion, neuroticism, and agreeableness, it
was qualiﬁed by a signiﬁcant ordinal interaction such that there was no difference in the
breadth of impression formation in CMC and F TF interactions for conscientiousness and
openness. Hancock and Dunham interpreted this interaction as evidence that
“participants were prepared to make judgments regarding their partner’s openness and
conscientiousness in as much detail in the CMC condition as in the F TF condition” (p.
335).

One plausible explanation of these ﬁndings is that, in order to best accomplish
the matching task, it was beneﬁcial for participants to form an impression about their
partner’s openness and conscientiousness. Miller and Steinberg’s (1975) suggestion that
in some situations it may be useful to have knowledge of a partner’s idiosyncrasies helps

to explain some ﬁndings in impression formation in CMC. In the case of Hancock and

Dunham’s (2001) ﬁndings, determining a partner’s conscientiousness may have helped a
perceiver make a judgment about whether the detail a target was providing was adequate
to determine the shape of the tangram. Speciﬁcally, openness and conscientiousness
may have been more relevant to the task because in order to successfully complete the
task, participants had to trust that their partners were carefully and accurately describing
the tangram to be matched (requiring a conscientiousness judgment) and also may have
required conﬁdence about the aesthetic sensitivity and intellectual curiosity of their
partner (which, according to Costa and McCrae [1991], are components of openness.)
However, it seems unlikely that a partner’s extraversion would be a helpful impression
to form in order to best accomplish the matching task. Thus, conscientiousness may
have been a task-relevant impression while extraversion was not.

Probabilistic mental models. This explanation of Hancock and Dunham’s (2001)
ﬁndings is consistent with Gigerenzer and Kurz’s (2001) adaptation of Brunswik’s
(1956) lens model. Gigerenzer and Kurz argued that when people make judgments based
on information, they do not explore all of the possible information available to them.
Instead, people make judgments based on the best piece of information available to
them. Gigerenzer and Goldstein’s (1996) formal explanation of this take-the-best
heuristic is rooted in probabilistic mental models theory (PMM theory; Gigerenzer,
1993; Gigerenzer, Hoffrage, & Kleinbblting, 1991). Both PMM theory and the take-the-
best heuristic emerge from a line of psychological inquiry positing that when people
make judgments, they are disinclined to engage in the complex process of evaluating,
weighting, and summing each individual piece of decision-relevant information. Rather,

people satisﬁce: They sacriﬁce optimal decision making processes (where an individual

might exhaustively evaluate each decision alternative and be virtually assured of making
a correct judgment) in favor of a more efﬁcient decision-making approach that yields an
adequate decision and avoids gratuitously expending cognitive effort (Simon, 1945,
1956, 1982; cf. Krosnick, 1991). (In fact, Gigerenzer and Goldstein [1996] present
evidence that the take-the-best heuristic, in addition to being efﬁcient, leads users of the
heuristic to exceptionally accurate conclusions on some tasks.)

PMM theory may inform an understanding of the impression formation process
as it occurs in CMC. When impression bearing verbal and nonverbal information is
forced into a text-based channel, the transfer rate of impression bearing information
slows (Ramirez, 2007; Walther, 1992, 1993). Because of this, all other things being
equal, we see generally less broadly developed impressions in CMC than in face-to-face
interaction within a relatively restricted time interval (Hancock & Dunham, 2001).
However, it is possible that when the formation of an impression of another person
facilitates task or social success, people speciﬁcally seek out information that may help
them to form a targeted impression of another person, as was the case for impressions of
conscientiousness and openness in Hancock and Dunham’s (2001) research.

Gigerenzer and Goldstein’s (1996) “take-the—best” algorithm assumes that we are
able to determine which cue is the clearest indicator of an underlying impression. Based
on the arguments above it seems possible that situations vary in the degree to which
forming an impression of another person is useful, and that we seek out the best
information in order to form a satisfactory impression about a discussion partner’s
pertinent characteristics. However, the question remains: How do we decide which

information is the best available information? In order to explore a possible answer to

this question, this research now turns to a discussion of the informational value of
impression bearing cues.
Cue Informational Value

If, as Gigerenzer and colleagues (Gigerenzer & Goldstein, 1996; Gigerenzer &
Kurz, 2001) argue, people satisﬁce by making judgments according to only the best
available cues, there must be some mechanism by which people evaluate the relative
value of a cue. Let us refer to the degree to which an individual cue allows an observer
to discriminate between different values of an underlying impression as the
informational value of a cue. This research proposes that there are several criteria that
shape how different impression-bearing cues affect impression judgments of a target: the
warranting value of the information, the distinctiveness of the information, and the
relevance of the information to impression utility demands. This research explores the
deﬁnition of each of these informational characteristics and considers the degree to
which each characteristic is a necessary or sufﬁcient condition in determining whether
or not a cue has large amounts of informational value.

One factor that may inﬂuence the informational value of an impression-bearing
cue is the warranting value associated with that cue. Walther and Parks (2002) stated
that the warranting value of information is “derived from the receiver’s perception about
the extent to which the content of that information is immune to manipulation by the
person to whom it refers” (p. 552). Walther and Parks argued that individuals prefer to
base judgments on information that has greater warranting value. Similarly, Donath
(1999, 2007) argued that it in virtual communities signals vary in the degree to which

users can manipulate them. Speciﬁcally, Donath suggested that while online people may

10

claim most anything about their personal characteristics, these self-claims may be less
trustworthy than less manipulable information. Drawing on Donath’s work, Walther,
Van Der Heide, Hamel, and Shulman (2009) suggested that when another party gives a
third-person testimonial about an individual, the testimonial is of greater judgmental
value than a ﬁrst-person disclosure. The preference for warranted information is
hypothesized by the warranting principle to lead one to base judgments of a person’s
physical attractiveness on a third-party’s testimony about the target rather than the
target’s disclosure of his or her own physical attractiveness. An initial test of the
warranting principle indicated that people do indeed display a preference for information
with a greater warranting value on some characteristics such as physical attractiveness
(Walther et al., 2009).

The warranting value of information should inﬂuence a cue’s informational
value. That is, when the warranting value of impression bearing information is high, it
stands to reason that such a cue should also be perceived to have greater informational
value to a person who is seeking to make an impression judgment.

Another factor that may inﬂuence whether a cue has high levels of informational
value is the degree to which the information contained in the cue is distinctive. For the
purposes of this research distinctive information can be conceptualized at one end of a
bi-polar scale opposite equivocal information. Equivocal cues contain information that
allows for the possibility of a number of different interpretations. For instance, a
sloppily written email message may easily lend itself to a number of different
interpretations (Lea & Spears, 1992). One might make a dispositional inference from the

badly composed message and judge the message author to be lacking in

11

conscientiousness. Alternatively, one might make the situational inference from this
same message that the author was in a hurry to write the message. Such an equivocal cue
is not likely to have great informational value.

Distinctive cues are similar to Kelly’s (1967) concept of distinctiveness
information. Kelly suggested that the observation of distinctive behavior allows an
observer to make attributions about some perceptive object (including interpersonal
impression judgments). Orvis, Cunningham, and Kelly (1975) used the example of Dr.
Stanton who complimented the work of a student, whose name is Barry. If an observer
of Dr. Stanton knows that he almost never compliments students’ work, the observer
may consequently make the judgment that Barry’s work was good. In this case, Dr.
Stanton’s behavior is distinctive. Conversely, when an observer of Dr. Stanton notes that
he praises almost every student’s work, the observer is less likely to judge Barry’s work
as good. In this case, Dr. Stanton’s behavior is not distinctive. Distinctive cues, as used
in the present research, differ from Kelly’s notion of distinctiveness information in that
they can operate when an observer makes a single observation of these cues. Drawing on
the example above, distinctiveness information is only available to an observer when she
or he is able to observe other instances of Dr. Stanton’s behavior.

Distinctive cues allow observers to make judgments about a target or object
without multiple observations. Kelly’s (1973) subsequent work describes how observers
make attributions on the basis of a single exposure to a stimulus. Kelly suggests that
observers discount some causal attributions about an effect when other rival causal
attributions may be made. A corollary of the discounting principle describes distinctive

cues: When a cue (or in Kelly’s lexicon an effect) may only have arisen from one

12

underlying attribute (a cause), observers will not discount the underlying attribute as a
possible cause of the cue. Thus, the cue is distinctive.

A distinctive one does not easily lend itself to multiple interpretations. For
instance, given that one is familiar with the sex-linked naming practices in the United
States, an email from Bill is a distinctive cue that the originator of the message was
male, while an email from Jennifer is a distinctive cue that the originator of the message
was female. Alternatively, an email from a person named Alex provides equivocal
information, as the name “Alex” is commonly held by both women and men. The
distinctiveness of a one should inﬂuence its informational value. That is, when a cue is
highly distinctive, it should also be perceived to have high informational value to a
person seeking to make an impression judgment.

A ﬁnal factor that may inﬂuence the impressions people form on the basis of a
cue is the task-relevance of that cue to the impression value speciﬁc to the situation.
Cues have a greater effect on judgments when they are more applicable to a particular
judgmental task (Todorov, 2000). In other words, the impression that a perceiver forms
should be more strongly informed by a cue that provides impression-bearing information
about an impression that is useful in helping the perceiver to accomplish his or her goals.
However, a perceiver who encounters a cue that provides impression bearing
information about an impression that is irrelevant to his or her goals should be less likely
to utilize that cue to inform his or her impressions of the target. This should be the case
because a perceiver is, presumably, looking for speciﬁc cues that helps him or her to
accomplish his or her goals. It is possible that the relevance of cues to the impression

value of the situation may also explain why Hancock and Dunham (2001) found that in

13

CMC impressions of the personality dimensions of openness and conscientiousness were
as broadly formed as impressions of a partner’s openness and conscientiousness in F TF
communication. Participants may have focused on those impression-bearing cues that
most clearly helped them to accomplish the collaborative tangram-matching task. Thus,
when a cue provides perceivers with impression bearing information that is relevant to
the impression value of the situation, perceivers should be more likely to attend to this
cue because it helps them to accomplish their task goals. Speciﬁcally, as a cue becomes
more relevant to the impression demands of a situation, it should have a greater effect on
a relevant impression judgment.

To recap, when we form impressions of others, we select cues with greater
amounts of informational value. It is also conceivable that in a situation where forming
an impression of another person might be very useful (there is a high level of impression
utility), there may be some cues with high amounts of informational value. Let us refer
to these highly useﬁrl cues as sticky cues. Sticky cues may be exceptionally important in
CMC. Weisband and Atwater (1999) found that, in CMC groups, liking between task-
group members was positively related to people’s task-related communication
behaviors. In face-to-face groups, however, liking between group members was not
signiﬁcantly related to task-related communication behaviors. Instead Weisband and
Atwater concluded that liking among group members in face-to-face groups was
facilitated by nonverbal behaviors not present in CMC groups. This suggests that
because CMC ﬁlters out many nonverbal cues, users of CMC may be able to focus on
speciﬁc textual information presented in computer—mediated messages. Thus, in CMC,

where people expect to have to seek out salient social information without the aid of

14

traditional nonverbal cues, sticky cues may function as particularly potent morsels of
social information, allowing individuals to form well-developed and efﬁcient, although
perhaps narrow, impressions of others’ characteristics.
Interpersonal vs. Group Impression Formation

Research on impression formation in CMC has explored two distinct streams.
One of these research directions has explored the way that idiosyncratic impressions
form via interactive text-based CMC between communicators over time (see Walther,
1992, 1993). Another direction in impression formation has been to explore the way that
categorical information shapes the process of social identiﬁcation in CMC (see Reicher
et al., 1995). The ﬁrst of these perspectives (collectively referred to as a social
information processing theory [SIPT] approach) focuses primarily on how impressions
form in a zero-history, text-only interactive exchange. The second perspective,
popularized by the social identity model of deindividuation effects (SIDE; Reicher et al.,
1995), has focused on how information about the types of groups or categories a person
belongs to may shape normative behaviors and beliefs that occur in CMC. The sticky-
cues model may improve our understanding of how these categorical / group
impressions form by describing the factors that cause group or categorical impressions
to form most potently. This research now turns to a discussion of the SIDE model and
how sticky cues may affect the formation of categorical / group impressions.

The SIDE model (Reicher et al., 1995) proposes that features common to
mediated communication (such as visual anonymity and physical isolation) reduce the
salience of one’s personal identity and increase the salience of one’s group identity.

Increased group identity salience leads to a greater adherence to the norms associated

15

with a salient group. In line with Turner’s (1982, 1985) self-categorization theory, SIDE
argues that when people personally identify, they think of themselves as individuals with
unique characteristics and idiosyncrasies, but when people socially identify, they think
of other individuals primarily as members of some salient category or group. In visually
anonymous and physically isolated environments such as in CMC, the SIDE model
suggests that people relate to one another in group, rather than interpersonal terms.

F iske and Neuberg (1990) suggested that people form impressions of others in
several ways. People may form categorical impressions of others and then recategorize
based on individuating attributes, or people form impressions of others based on their
individuating attributes. Fiske and Neuberg suggested that instead of choosing just one
of these impression formation strategies, perceivers vary on a continuum based on the
degree to which they utilize individually individuating information about a target. Fiske
and Neuberg argued that the degree to which a perceiver utilizes individuating
information varies according to how motivated a perceiver is to determine individuating
information about a target. Speciﬁcally, Fiske and Neuberg suggested that people ﬁrst
rely on categorical information to form impressions of others. If people remain
unmotivated to form an attribute based impression of a target, they will form a
categorical impression of a target. After this initial categorical classiﬁcation, people may
be motivated to use an attribute based approach or a categorical approach to reﬁne the
initial impression of a target. One factor that may serve to motivate individuals is the
impression value of a given situation.

F iske and Neuberg (1990) argued that categorical impressions initially have

priority over interpersonal attribute based impressions. Speciﬁcally, Fiske and Neuberg

l6

argued that committed outcome dependencies can motivate individuals to form
categorical impressions. A committed outcome dependency occurs when an individual is
motivated by a desire to conﬁrm a speciﬁc type of expectation (rather than form an
accurate impression). For example, if an individual was primed to expect that his or her
fellow group members valued efﬁciency (e. g., Postmes et al., 2001), individuals may be
more motivated to maintain categorical impressions that conﬁrm initial expectations
while turning a blind eye to individual characteristics that may indicate that his or her
partner, in contrast to what one would expect from that person’s group membership, is
not an efficient person. F iske and Neuberg argued that the desire to conﬁrm expectations
may be especially salient when favorable situational outcomes depend on the
conﬁrmation of these expectations.

As Miller and Steinberg (1975) suggested, in many situations categorical
(cultural or sociological-level knowledge) about others is sufﬁcient. In situations where
these categorical impressions are sufﬁcient, it is likely that perceivers will attend to
impression bearing information that conﬁrms their group based expectations about
targets. This research has argued that, in some situations, individuated impressions may
be more useful (i. e., the situation is high in interpersonal impression value). However, it
is also possible that, in some situations, it may be sufficient to form a categorical
impression of another person. For instance, if a person were seeking credible medical
advice on an Internet discussion board, that person may be motivated to form an
impression of an advice-giver’s credibility based on cues indicating that the advice-giver

belonged to the category “physician”. Let us refer to the degree to which it is useful to

17

form a categorical impression of another person as the categorical impression value of a
situation.
Testing the Sticky Cues Model

This work proposes that the same cues can have different degrees of
informational value based on the different social situations that one encounters. When
forming a speciﬁc impression of a personal characteristic (e. g., extraversion) is task-
relevant, cues relating to that characteristic are more likely to inﬂuence impression
judgments on the speciﬁc impression. Conversely, when a speciﬁc impression judgment
is not task-relevant, the same cues that may have strongly inﬂuenced impression
judgments when those judgments were task-relevant may not have a strong inﬂuence.
This is expected to occur because when a speciﬁc impression is task-relevant, an
observer should be more likely to actively seek out information about that impression.
However, when a speciﬁc impression is not task-relevant, an observer should be less
likely to be actively seeking cues that inform the task-irrelevant impression. For
instance, in massively multiplayer online role-playing games (MMORPG) there are
commonly different types of goals. Some goals are social and others are achievement
based (Yee, 2007). If one’s goals for playing an MMORPG are primarily social, one
may try to ﬁnd out how socially attractive (i. e., likeable) others are, focusing primarily
on cues that provide information about a person’s social attractiveness (e. g., similarity).
Meanwhile players with achievement goals may be more interested in how skilled a
slayer of enemies another player is (e. g., Pena & Hancock, 2006; Utz, 2000), making

sociability cues less valuable to the perceiver.

18

This varying degree of information usefulness gives rise to an explanation of the
informational value of a given cue. When one can best complete a task by forming a
speciﬁc impression of a target, an observer should actively seek out information relevant
to the speciﬁc impression in question. Additionally, distinctive cues provide observers
with more impression bearing information than do equivocal cues. For instance, when
individuals make outright disclosures that they have some characteristic (e. g., a person
claims that she or he is extremely extraverted), this disclosure clearly provides
information about an individual’s extraversion; such information is distinctive. Other
cues to extraversion may be less distinctive. For example, extraverted individuals tend to
generate messages that use a preponderance of adjectives (Oberlander & Gill, 2006) and
are more verbose (Marcus, Machilek, & Schﬁtz, 2006) than introverts. An observer may
perceive a target that generates verbose messages including a preponderance of
adjectives to be more extraverted. However, the observer may also attribute such a cue
to other characteristics such as intelligence or inability to self-censor. Such information
is more equivocal than direct disclosures.

According to the warranting principle (Walther & Parks, 2002), judgments of a
target should vary on the basis of cues that an evaluator perceives to be difﬁcult to
manipulate (i. e., verbosity and frequent adjective use) rather than easily manipulable
cues (i. e., direct disclosures). However, Walther et al. (2009) found that direct self-
disclosures of extraversion affected a perceiver’s judgments of a target’s extraversion.
They speculated that the warranting principle might affect judgments especially when a
target is perceived to have something to gain. In the case of extraversion, Walther et al.

argued that there may be little perceived social beneﬁt to portraying oneself as

19

extraverted, while for other perceptive judgments such as physical attractiveness, the
perceived social beneﬁt of being perceived to be physically attractive is quite high:
Thus, the warranting value of information should drive physical attractiveness
judgments. Their data were consistent with this suggestion. The warranting value of
information has been shown to affect impression judgments of a target. The present
research posits that this effect should operate independently from the effect of sticky
cues on impression judgments. Because of this, the present research focuses on the
effect of distinctive cues and how relevant those cues are to situational demands rather
than the warranting value of the cues on impression judgments. The present research
speculates that distinctive cues allow individuals to form more extreme impressions than
do equivocal cues.

According to the deﬁnition of sticky cues, a cue informs impressions most when
it is both relevant and distinctive. Because sticky cues have greater informational value
than other cues, they allow an observer to form a more extreme impression of a target’s
characteristics than he or she would have otherwise formed. Observers of impression
cues (e. g., extraversion) should judge a target to be most extraverted when they are
exposed to a sticky cue to a target’s extraversion, but they should judge a target to be
less extraverted when the cue is not sticky.

This research predicts that an ordinal interaction between cue relevance and cue
distinctiveness on impression formation such that judgments of a target’s characteristics
are greatest when participants are exposed to a distinctive cue that is also task relevant,
but judgments of a target’s characteristics are less extreme when a cue is distinctive but

the task is not relevant, the task is relevant but the cues to the target’s characteristics are

20

not distinctive, or the task is irrelevant and the cue to the characteristic of interest is not
distinctive. In order to test this prediction, two studies are reported. The ﬁrst examined
the effects of relevant and distinctive group cues on intelligence judgments. The second
examined the effects of relevant and distinctive interpersonal cues on extraversion

judgments.

21

Chapter 2
STUDY 1

This research argues that distinctive and relevant group information inform
impression judgments to a greater degree than cues that are either (a) distinctive or
relevant but not both or (b) are neither distinctive nor relevant. Judgments of intelligence
were chosen as the focal impression characteristic because, depending on the
experimental task, judgments of intelligence could be either extremely relevant or
extremely irrelevant. Finally, intelligence was the impression judgment of choice
because group cues (a student’s university afﬁliation), which varied in terms of cue
distinctiveness and task relevance, were readily available.

The ﬁrst study examines the effects of sticky group cues to intelligence
impression judgments. This research predicts that perceivers judge targets to be most
intelligent when group-based intelligence cues are both distinctive and relevant to the
task.

Method
Procedure and Materials

In order to manipulate the relevance of intelligence ability judgments to the
experimental task, half of participants completed a task for which intelligence judgments
are highly relevant: helping to select possible candidates from a pool of people being
chosen to participate as individual competitors on a “college-week” version of a popular
television quiz show (see Appendix A). Participants were informed that the ideal
candidate is someone who demonstrates exceptionally high intelligence. Thus, forming

an impression of the target’s intelligence was task relevant. Alternatively, half of

22

participants were directed to a task for which intelligence judgments are not relevant:
Helping to select possible candidates from a pool of potential overnight hosts for
prospective undergraduate students for a new program being developed by the
Admissions Office of the school in question (either Westlake or Harvard University).
Participants were informed that the ideal candidate for this program is someone who is
very social and extraverted.

Stimuli were comprised by mockups of personal proﬁle pages on the popular
Internet social networking site F acebook (see Figure l). Stimuli were composed to
reﬂect differences in the distinctiveness of the group-based cues to intelligence
represented on the target’s proﬁle. In the distinct group cue to intelligence condition
stimuli depicted the target as being a member of the “Harvard University” network. In
the equivocal group cue to intelligence condition stimuli will depict the target as being a
member of the network “Westlake University”. These cues were chosen as distinctive
and equivocal group cues as it is a common belief that Harvard students are intelligent
while knowing that a person is a Westlake University student provides little information
about the person’s intelligence because Westlake University does not actually exist.
Thus, knowing that a person is a student at “Westlake University” cannot provide
distinctive information about a person’s intelligence beyond knowing that the person in
question is a university student. Other than these variations, experimental stimuli were
identical.

Participants were then informed that they would be basing their impression of the
participant upon the target’s limited F acebook proﬁle, and they should take as long as

necessary to form an impression of the target. In order to enhance the believability of the

23

task, participants were instructed that they are to evaluate several students’ F acebook
proﬁles, and after viewing each proﬁle they will be asked a series of questions. In
actuality, only the ﬁrst proﬁle will contain the induction of interest. Further, in order to
encourage the careful evaluation of each proﬁle, participants were told that the person
who completed the task by making the best evaluation of each target would earn a $50
dollar gift certiﬁcate. (In actuality this gift certiﬁcate was randomly awarded to a
participant.) After participants viewed the experimental stimuli, they completed several
dependent measures.
Participants and Design

Participants (N = 79) were students in undergraduate communication courses at a
large public university in the Midwestern U.S.A. Participants were randomly assigned to
one of four experimental conditions using a javascript program (Burton & Walther,
2001). Experimental conditions reﬂected differences in the distinctiveness of group-
based intelligence cues and the relevance of intelligence to the experimental task.
Speciﬁcally, the study design was a 2 (distinctiveness of group-based intelligence cues:
high vs. low) x 2 (relevance of intelligence judgments to the task: relevant vs. irrelevant)
between-subj ects design. All aspects of the experiment were administered online via
Internet web pages and forms. Participants completed the experiment in locations of
their own choosing, ostensibly in an environment where they routinely use the Internet.
Dependent Measures

Perceptions of the target’s intelligence were measured using ﬁve original Likert-
type items including “This person is intelligent,” “This person knows lots of facts,” and

“This person is very smart” (see Appendix B). The response set for this measure was a

24

seven- point scale anchored by “Strongly Agree” and “Strongly Disagree” with
“Neutral” anchoring the mid-point of the scale. In an attempt to hold participant fatigue
constant among conditions and dependent measures, this scale contained an additional
seven ﬁller items that measured various, similar aspects of the target’s personality (e. g.,
the target’s task attractiveness; see McCroskey & McCain, 1974).

In order to assess whether the intelligence scale was unidimensional, a
conﬁrmatory factor analysis was conducted using the internal consistency theorem
(Hunter & Gerbing, 1982). Additionally, a 12-item (ostensibly unidimensional) scale
measuring participants’ perceptions of the target’s extraversion was collected in order to
assess parallelism. These analyses indicated that the ﬁrst intelligence item should be
eliminated. With this item eliminated, the data suggested that the intelligence scale was
unidimensional. Errors of internal consistency were small for the intelligence scale (e <
.07). Only one error of parallelism was substantial (e = .21); other errors of parallelism
were small (all other e < .13). Moreover, on average, errors were small (RMSE = .08).
Because there was no evidence that the item that produced a large error of parallelism
and lacked content validity and because the average error was small, this item was
retained in the ﬁnal factor solution (see Table l for items and factor loadings). Finally,
the alpha reliability estimate based on standardized items was acceptable (or = .79).
Induction Check

In order to determine that the intelligence cue distinctiveness induction was
effective, a manipulation check was conducted. An independent sample of participants
(N = 87) viewed the single piece of information about the target that was manipulated in

the actual experimental stimulus. That is, participants received a single statement about a

25

target, which informed them that the target was “a student at (Harvard / Westlake)
University.” After viewing this information, participants reported their degree of
conﬁdence in their impression of the target’s intelligence. Participants’ conﬁdence in
their intelligence impression judgment was measured using ﬁve Likert-type items such
as “I feel very conﬁdent about my impression of the person’s intelligence,” and “It is
difficult to make a judgment about this person’s intellectual abilities” (reverse-scored).
(See Appendix C for a sample stimulus and a complete listing of scale items). This scale
displayed acceptable inter-item reliability, or = .88. An independent samples t-test was
used to evaluate whether greater conﬁdence in intelligence judgments was aroused by
the distinctive one than the equivocal cue. Participants who saw the distinctive group cue
(i. e., this person is a Harvard student) to the target’s intelligence reported substantially
greater conﬁdence in their intelligence impression judgment (n = 42, M = 5.00, SD =
1.02) than did participants who were exposed to the equivocal group cue (i. e., this
person is a Westlake student) to the target’s intelligence (n = 45, M = 3.18, SD = 0.97), t
(85) = 8.48, p < .01 , n2 = .46. The experimental induction of greater or lesser cue
distinctiveness was successful. The data did not suggest that the group cue induction had
an effect on participants’ judgmental conﬁdence of the target’s extraversion; distinctive
group cues (n = 42, M = 5.08, SD = 1.40) did not produce signiﬁcantly different
extraversion impression conﬁdence judgments than did equivocal group cues (n = 45, M
= 4.82, SD = 1.42), t (85) = 0.86, p = .60.

An examination of the experimental task was conducted to insure that this task
manipulation successfully induced different degrees of situational task relevance.

Another independent sample of participants (N = 64) viewed the experimental task

26

presented in the study and rated the relevance of intelligence judgments to the task on a
6-item Likert-type scale (see Appendix D) including items such as “In order to complete
this task I need information about the person’s intelligence,” and “It is not vital to this
task to form an intelligence impression” (reverse-scored). A Cronbach’s alpha reliability
estimate showed that the scale displayed acceptable reliability (or = .94). As expected,
the participants who viewed the intelligence-relevant (game-show) experimental task
responded that they viewed intelligence to be more relevant (n = 32, M = 5.87, SD =
0.84) than those who viewed the intelligence-irrelevant (overnight host) experimental
task (n == 32, M = 3.54, so --—- 1.18), t (62) = 9.06, p < .01, n2 = .57. Thus, the
experimental induction of greater or lesser relevance of an intelligence impression was

successful.

27

Chapter 3

STUDY 1: RESULTS AND DISCUSSION

The ﬁrst study predicted that when group membership cues provide distinctive
information about a F acebook proﬁle owner’s intelligence for a task in which
intelligence judgments are highly relevant, participants would rate the target to be most
intelligent. However, when group membership cues were either equivocal, irrelevant, or
both equivocal and irrelevant, participants would rate the target to be less intelligent. In
order to test this hypothesis, a contrast analysis evaluated whether the data were
consistent with the a priori contrast weights (see Table 2). The means of the study
conditions (see Table 3 for means and standard deviations) were not consistent with
contrast analysis predictions, t (75) = 0.81, p = .21 (one-tailed), n2 < .01. Thus, the data
were not consistent with the hypothesized interaction.

In order to determine whether either group cue relevance or distinctiveness alone
inﬂuenced intelligence judgments, a post-hoe two-way ANOVA was conducted. The
data did not suggest that whether a cue provided information that was more or less
relevant to the experimental task did not affect intelligence judgments, F (1, 75) = 0.01,
p = 0.921. However, cue distinctiveness appeared to inﬂuence intelligence judgments
such that distinctive group cues to intelligence were associated with judgments of
greater intelligence than were equivocal group cues to intelligence, F (l, 75) = 4.87, p =
0.03, n2 = .06 (see Table 3 for means and standard deviations). In this case, cue
distinctiveness had a signiﬁcant effect on impression judgments such that a highly

distinctive intelligence cue caused participants’ judgments of a target’s intelligence to be

28

greater than a less distinctive intelligence cue. However, the situational relevance of an
intelligence cue did not affect intelligence judgments of a target.
Discussion

This study examined whether a special class of group cue, a sticky group cue,
held great impression bearing value when participants were presented with a distinctive
cue that was also task-relevant. The data suggested that the combination of both
distinctive cues that were task-relevant were not required for group impression cues to
be inﬂuential. Further, while the distinctiveness of the one made that one especially
informative to participants making impression judgments, the task-relevance of an
intelligence impression did not signiﬁcantly affect impression judgments.

These ﬁndings may be attributable to several factors. First, it is possible that
group based (categorical) impressions occur regardless of what cues are task-relevant.
As F iske and Neuberg (1990) argue, categorical impressions may be the ﬁrst type of
impression to form. If perceivers are unmotivated to evaluate a target beyond this initial
categorical level, perceivers may instead rely on the best available cue to form an
intelligence impression on the basis of categorical information alone (Gigerenzer &
Goldstein, 1996; Gigerenzer & Kurz, 2001). In this study, the only cue which varied was
the categorical cue participants saw was the university afﬁliation of the target (i. e.,
Westlake or Harvard University). Because this was the only cue that varied between
subjects and participants saw only one of these cues, they may have judged this cue
about the target to be the best possible cue to the target’s intelligence and may have
based their judgment about the target on that cue alone regardless of whether the task

rendered an intelligence judgment task-relevant or not.

29

Another related explanation of these ﬁndings is that it is possible that group
based impression cues—because of their primacy—have different requirements than
interpersonal impression cues for what makes them sticky or not. If, as Fiske and
Neuberg (1990) suggest, categorical impressions form ﬁrst, and, as Gigerenzer and
Goldstein (1996) and Gigerenzer and Kurz (2001) suggest, we form impressions on the
basis of the one best piece of available information, one might expect that an observer
who is seeking to form an impression of another person might ﬁnd a distinctive
categorical impression cue (as opposed to the task one is to be performing) to be most
informative and form a group-based impression with little concern about whether that
impression is task-relevant or not. Task relevance may matter only when the categorical
impression formed by the perceiver is judged to be insufﬁcient to accomplish his or her
goals. Although the only variance in the stimuli reﬂected differences in the group
afﬁliation of the proﬁle owner, it is possible that, in both the intelligence-judgment
relevant and irrelevant condition, participants felt that they had adequate information
upon which to make their impression judgments. Thus, participants may not have been
motivated to move beyond their initial categorical impressions of the target.

In Hancock and Dunham’s (2001) study, task-relevant personality characteristics
emerged after interactions with a partner. These interactions may have allowed increased
exposure to interpersonal impression cues. Because Hancock and Dunham’s research
utilized interactive dyads there may have been several types of impression-bearing cues
available to perceivers. For instance, it is possible that participants made disclosures
about their personality characteristics in their interactions. Such disclosures may have

provided additional information about a target’s personality that allowed a perceiver to

30

better judge the target’s personality. Disclosure cues such as these may be particularly
distinctive cues to a person’s identity, and when these cues reﬂect on a characteristic
that helps a perceiver accomplish his or her task goals they may be particularly potent
pieces of impression-bearing information. Because the ﬁrst study was focused primarily
on the distinctiveness and relevance of group cues, disclosure cues were held constant
and intentionally did not reﬂect on relevant impression judgments. However, these cues
are of particular interest given that previous research has found that disclosures are more
frequently used in CMC than in face-to-face communication (Tidwell & Walther, 2002)
and in CMC a message sender’s explicit statements (disclosures) of positive affection
are strongly related to a receiver’s feelings immediacy and affection (Walther, Loh, &
Granka, 2005). In short, in CMC, communicators make disclosures and these disclosures
allow a person to communicate immediacy and affection. In order to test whether both
the task-relevance and distinctiveness of individual disclosure impression cues affected

interpersonal impression formation in CMC, a second study was conducted.

31

Chapter 4

STUDY 2

A second study used the extraversion / introversion dimension as the impression
judgment of interest. Extraversion judgments have been one of the most commonly
studied impression judgments in communication technology research. Previous studies
of impression formation in CMC have utilized extraversion as a judgmental criterion of
interest (e. g., Hancock & Dunham, 2001; Tong, Van Der Heide, Langwell, & Walther,
2008; Walther et al., 2009.) Research in this line has been guided by the fact that
extraversion has been associated with a number of positive outcomes (for review see
Eysenck & Eysenck, 1985). Thus, knowing an individual’s level of extraversion is a
helpful and commonly made judgment. Additionally, extraversion is a personality
characteristic about which both group and interpersonal cues may have some bearing.
Finally, this research seeks to test the critical proposition that the relevance and
distinctiveness of a cue impacts the degree to which the cue is utilized when a perceiver
forms an impression of a target. As such, extraversion is a favorable impression
judgment because previous research has suggested the warranting value of extraversion
cues is less important to the judgmental process than other factors (Walther et al., 2009).
For example, Walther et al. (2009) found that a target’s direct self-disclosure of
extraversion signiﬁcantly affected an observer’s judgment of a target’s extraversion.
This study requires a judgmental criterion that allows the sticky cue hypothesis to be
tested most clearly. In order for this hypothesis to be most clearly tested, the judgmental
criterion of the study needs to be able to be systematically varied while reducing the

other characteristics that might cause an observer’s judgments to vary. In this case,

32

because the warranting value of an extraversion disclosure is not vital to an observer’s
extraversion judgments, extraversion is a good candidate for a judgmental criterion for
the present study as it allows for judgments to occur on the basis of cue characteristics
other than the warranting value of the cue.

The second study examines the effects of sticky cues to extraversion impression
judgments. This research predicts that perceivers judge targets to be most extraverted
when extraversion cues are both distinctive and relevant to the task. Additionally, the
second study added two control conditions that were predicted to produce the lowest
extraversion judgments because the disclosure cues gave no information about the
target’s extraversion. These conditions were added to determine whether an equivocal
cue allowed observers to make some judgment about a target’s extraversion (resulting in
higher extraversion judgments than in the control conditions) but still lower than the
condition containing a task-relevant, distinctive cue. Thus, the second study predicted
that extraversion judgments would be greatest when the stimulus contained a distinctive
cue that was task relevant, extraversion judgments would be lowest when stimuli
contained non-distinctive (control) cues, and extraversion judgments would be moderate
when stimuli contained an equivocal cue (regardless of the task) or when the task was
relevant and the extraversion cue was distinctive.

Method
Procedure and Materials

As in study 1, all aspects of the experiment were administered online via Internet

web pages and forms. Participants completed the experiment in locations of their own

choosing, ostensibly in an environment where they routinely use the Internet. Study 2

33

was conducted at the same time as study 1. Participants were randomly assigned to one
of the two studies by a javascript program (Burton & Wather, 2001).

As in the ﬁrst study, participants were instructed that their task for the session
was to evaluate of one of several students. In order to induce different levels of the
relevance of extraversion judgments to the experimental task, a javascript program
(Burton & Walther, 2001) randomly directed half of participants to a task for which
extraversion judgments are highly relevant: helping to select possible student hosts for a
program developed by the Admissions Ofﬁce. Participants were informed that the ideal
candidate for this program is someone who is very social and extraverted. Thus, forming
an interpersonal impression of the target’s extraversion was task relevant. Alternatively,
half of participants were directed to a task for which extraversion judgments are
irrelevant: helping to select candidates as individual competitors on a “college-week”
version of a television quiz show (see Appendix A). Participants were informed that the
ideal candidate is someone who is highly intelligent. Participants based their impression
of the proﬁle owner upon the target’s limited F acebook page. Thus, the same tasks from
study I were utilized in study 2; however, for study 2 the task-relevant impression was
reversed. That is, for study 2 the quiz show did not require an extraversion judgment
(the focal dependent variable) to successfully complete the task, while the student-
hosting task did require an extraversion judgment to successfully complete the task.
Thus for the student hosting task extraversion was a relevant impression, but for the
quiz-show task an extraversion impression was not relevant. As in study 1, in order to
enhance task-believability, participants were instructed to evaluate several students’

Facebook proﬁles, and after viewing each they were asked a series of questions. In

34

actuality, only the ﬁrst proﬁle contained the induction of interest. Also, as in the ﬁrst
study, participants were instructed that the person who made the best judgment would be
awarded a $50 dollar gift certiﬁcate.

In the distinct cue to extraversion condition, the stimulus (see Figure 2) included
a statement in the “about me” portion of the proﬁle that included a direct claim of
extraversion—such direct claims have been successful at inducing judgments of greater
extraversion in previous research (Walther et al., 2009). This section of the proﬁle stated
that, “I am a sociable people-person...until you get to know me, then I’m EXTREMELY
extraverted and outgoing!!!” In the equivocal cue to extraversion condition the stimulus
included a statement in the “about me” portion of the proﬁle that indirectly indicated—
through greater verbosity (Eysenck & Eysenck, 1985; Rutter, Morley, & Graham,
l972)—-that the target was extraverted. In this condition the “about me” section of the
proﬁle stated “I am a conscientious person. I try to do a good job with the things that
come my way. I enjoy American Idol at the moment, (I’m a Simon Cowell fan). I enjoy
college sports. . .speciﬁcally I love basketball (March is my favorite time of year). . .and
I’m also a foodie. I speciﬁcally enjoy Italian and Middle Eastern food. There are a lot of
great restaurants around that serve these types of foods and I enjoy going to them.”
Finally, a control condition contained the brief ﬁller statement, “I really like U2” in the
“about me” section of the proﬁle. Other than these manipulations experimental stimuli
were identical.
Participants and Design

Participants (N = 108) were students in undergraduate communication courses at

a large public university in the Midwestern U.S.A. Participants were randomly assigned

35

to one of six experimental conditions. Experimental conditions reﬂected differences in
the distinctiveness of interpersonal extraversion cues and the relevance of extraversion
cues to the experimental task. Speciﬁcally, the study design was a 3 (distinctiveness of
interpersonal extraversion cues: high vs. low vs. control) x 2 (relevance of extraversion
judgments to the task: relevant vs. irrelevant) between-subjects design.

Dependent Measures

Participants rated the target on extraversion according to the extraversion
subscale of the NEO-FFI (Costa & McCrae, 1991; see Appendix E) that was adapted for
observers (Hancock & Dunham, 2001; Walther et al., 2009). This scale consisted of 12
Likert-type items including “This person likes to have a lot of people around,” and “This
person really enjoys talking to people.” The response set for this measure was a seven-
point scale anchored by “Strongly Agree” and “Strongly Disagree” with “Neutral”
anchoring the mid-point of the scale.

In order to assess whether this scale was unidimensional, a conﬁrmatory factor
analysis was conducted using the internal consistency theorem (Hunter & Gerbing,
1982). Additionally, a ﬁve-item (ostensibly unidimensional) scale measuring
participants’ perceptions of the target’s intelligence was collected in order to assess
parallelism. These analyses indicated that the ﬁrst intelligence item and items 3, 5, and 7
ﬁom the extraversion scale be eliminated. With these items eliminated, the data
suggested that the extraversion scale was unidimensional. All errors of internal
consistency were small (all e ’s < .09). Only one error of parallelism was substantial (e =
.20); other errors of parallelism were small (all other e < .15). Moreover, on average

errors were small (RMSE = .08). Because there was no evidence that the item that

36

 

produced an error of parallelism lacked content validity and because the average error
was small, this item was retained in the ﬁnal factor solution (see Table 1 for items and
factor loadings). Finally, alpha reliability estimates based on standardized items were
acceptable for (or = .87).
Induction Check

In order to determine that the extraversion cue distinctiveness induction was
effective, a manipulation check was conducted. An independent sample (N = 84) viewed
one of the three disclosures (distinctive, equivocal, or control) the target made about
him/herself as described above. Each participant viewed only the single disclosure about
the target that was manipulated in the actual experimental stimulus (whether the target
claimed that: “I am a sociable people-person...until you get to know me, then I’m
EXTREMELY extraverted and outgoing!!!” [distinctive condition], “I am a
conscientious person. I try to do a good job with the things that come my way. I enjoy
American Idol at the moment, (I’m a Simon Cowell fan). I enjoy college
sports. . .speciﬁcally I love basketball (March is my favorite time of year). . .and I’m also
a foodie. I speciﬁcally enjoy Italian and Middle Eastern food. There are a lot of great
restaurants around that serve these types of foods and I enjoy going to them”,
[equivocal condition], or “I really like U2” [control condition]. After viewing this
information, participants answered a number of questions about the target including ﬁve
Likert-type items measuring participants’ degree of conﬁdence in their impression of the
target’s extraversion. This scale included items such as “I feel very conﬁdent about my

impression of the person’s extraversion,” and “It is difﬁcult to make a judgment about

37

this person’s sociability” (reverse-scored). This scale displayed acceptable inter-item
reliability, or = .86.

A contrast analysis evaluated whether participants reported greatest conﬁdence
in their extraversion judgments after viewing the disclosure from the distinctive
extraversion condition, than they did after viewing the equivocal disclosure cue, while
the lowest levels of conﬁdence in participants’ extraversion impressions were found
after viewing the disclosure from the control condition. Participants who saw the
distinctive extraversion one reported greater conﬁdence in their extraversion impression
judgment (n = 28, M = 5.42, SD = 1.19) than did participants who were exposed to the
equivocal extraversion cue (n = 28, M = 4.61, SD = 1.06), while participants who were
exposed to the disclosure in the control condition reported being least conﬁdent of their
impression of the target’s extraversion (n = 28, M = 3.82, SD = 1.56), t (81) = 4.64, p <
.01 , n2 = .21. An examination of the residual explained variance showed that after the
expected effect was accounted for the residual explained variance was relatively small
and did not achieve statistical signiﬁcance, F (1, 81) = 0.01, p > .99. Finally, a least
significant difference analysis of each of the groups with one another showed that after
viewing the stimuli was signiﬁcantly different from one another in expected directions
(all p’s < .05). Thus, the experimental induction of greater or lesser extraversion cue
distinctiveness was judged to be successful. The data did not suggest that the
extraversion cue induction affected participants’ conﬁdence in their judgments of
intelligence. A one-way ANOVA did not ﬁnd any signiﬁcant differences in conﬁdence

in intelligence judgments among the distinctive cue condition (n = 28, M = 4.35, SD =

38

1.57), the equivocal cue condition (n = 28, M = 4.22, SD = 1.33), and the control
condition (n = 28, M = 4.02, SD = 1.29), F (2, 81) = 0.39,p = .68.

The experimental tasks were also checked to determine that they induced higher
or lower degrees of task relevance as intended. Another independent sample (N = 64)
viewed the experimental task presented in the study and rated the relevance of
extraversion judgments to the task on a 6-item Likert-type scale (see Appendix F)
including items such as: “Knowing whether a person is a people-person is relevant to
this task,” and “It is not vital to this task to form an extraversion impression” (reverse-
scored). A Cronbach’s alpha reliability estimate showed that the scale displayed
acceptable reliability (or = .94). As expected the participants who viewed the
extraversion-relevant task responded that they viewed extraversion to be more relevant
(M = 5.81, SD = 0.86) than those who viewed the extraversion-irrelevant task (M = 3.97,
SD = 1.37), t (62) = 6.52, p < .01, n2 = .41. Thus, the experimental induction of greater

or lesser relevance of an extraversion impression was successful.

39

Chapter 5

STUDY 2: RESULTS AND DISCUSSION

This study predicted that participants would rate the target to be most extraverted
when cues to the target’s extraversion were both relevant and distinctive, participants
would rate the target to be less extraverted when cues to the target’s extraversion were
either relevant or distinctive, or neither relevant nor distinctive, and participants would
rate the target to be least extraverted when they saw the extraversion cue in the control
condition (see Table 2 for contrast weights). In order to test this hypothesis, a contrast
analysis evaluated whether means were consistent with this predicted pattern (means and
standard deviations are presented in Table 3). The observed pattern of means was
consistent with the pattern of means predicted by the contrast analysis, t (102) = 2.56, p
= .01 (one-tailed), n2 = .06. In order to determine whether the pattern of results was
likely to be consistent with the predicted contrast but not other possible contrast patterns,
the residual explained variance was analyzed (see Keppel & Wickens, 2004). After the
variance predicted by the model was accounted for, there was little residual explained
variance, F (4, 103) = 0.18, p = .95. Thus, the data were consistent with the sticky cues
hypothesis for interpersonal impression judgments.

Discussion

Consistent with the sticky-cues hypothesis, this study suggests that distinctive
interpersonal impression cues that are also task-relevant are especially potent inﬂuencers
of extraversion impression judgments. In other words, the data were consistent with the
trend suggesting that distinctive and relevant interpersonal cues to higher extraversion

caused judgments of greater extraversion more together than they do in isolation of one

40

another and that both equivocal and distinctive cues caused target ratings to be larger
than control cues. The size of the sticky-cue effect was not particularly large. However,
it is comparable to other effect sizes reported for other studies examining the effects of
different types of cues in CMC and communication modalities on impression formation.
For example, Walther et al., (2009) reported an effect size of approximately n2 = .02 for
the effect of the warranting value of a cue on physical attractiveness impressions.
Additionally, Hancock and Dunham (2001) reported an ordinal interaction consistent
with the effect of task-relevance on impression formation with an effect size of
approximately n2 = .04. Further, a study examining the effect of system-generated cues
(i. e., number of friends) in a Facebook environment found similar effect sizes (r12 = .02)
for the number of friends presented on a target’s proﬁle and participants’ judgments of
the target’s social attractiveness (Tong et al., 2008), and another study found an effect
size of 112 = .04 for the effect of credibility cues present in a text-based message on

credibility impressions (Van Der Heide & Walther, 2009).

41

Chapter 6

GENERAL DISCUSSION

The two experiments reported here examined how the relevance of an impression
to the task and the distinctiveness of an impression cue affect impression formation in
CMC. The data suggest that when individuals are exposed to more distinctive
impression-bearing cues they form more extreme judgments of the characteristic upon
which the distinctive cue provides information. In some situations, cues can potently
inﬂuence of impressions when a task requires an individual to form an impression of
some particular characteristic and there is a distinctive cue present that reﬂects upon that
characteristic (i. e., an individual is seeking information about a target’s extraversion
and a distinctive cue to the target’s extraversion is present). However, in other situations
(i. e., when a distinctive categorical cue about group membership which, distinctively,
reﬂects on a target’s intelligence), the relevance of a speciﬁc impression to the task was
not an important determinant of impression formation. Only the distinctiveness of a cue
strongly informed impressions.

One of the contributions of this research is to extrapolate from attribution theory
(Kelly, 1967, 1973) and PMM theory (Gigerenzer & Goldstein, 1996; Gigerenzer &
Kurz, 2001) to inform a perspective on impression formation in CMC. This research
argued that because perceivers are motivated to be efﬁcient processors of information,
perceivers do not consider all relevant social information when they form impressions of
a target. Instead perceivers focus on the cues that are most relevant to the impression
they need to fulﬁll their social goals. Further, this perspective suggested that distinctive

cues allow a perceiver to form an extreme impression of a target on the basis of a single

42

exposure to the target. Taken together, this suggested that in CMC, where impression
cues are more limited in short-term encounters, a perceiver should seek an impression
cue that is relevant to his or her task, and should base his or her judgments of a target
most strongly on a distinctive cue that allows a perceiver to best accomplish his or her
goals.

An analogy helps to describe this perspective. One can imagine the impression
formation process as building a tower of blocks. One can imagine each block as an
impression cue providing an observer with information about a characteristic of the
target. First, this perspective suggests that people try to build speciﬁc types of towers (i.
e., try to form speciﬁc impressions) in order to fulﬁll their goals. Second, this
perspective suggests that not all building blocks are the same size. Some blocks (i. e.,
equivocal cues) add only a little to the impression formation tower, other blocks (i. e.,
distinctive cues) are signiﬁcantly larger and add more to the impression formation
tower. Finally, perceivers may focus upon some already-large building blocks
(distinctive cues) when a perceiver speciﬁcally seeks them. Thus perceivers are even
more likely to use these blocks to provide a particularly large amount of impression-
bearing information about a target.

These data offered general support for the sticky-cues perspective. Speciﬁcally,
interpersonal disclosure cues were sticky when they provided information that was task-
relevant and distinctively reﬂected upon a target’s extraversion. The data also suggested
that in some cases even when certain impressions are not speciﬁcally task-relevant,
perceivers utilize distinctive cues to form well-developed impressions of a target’s

characteristics. Speciﬁcally, the ﬁrst study found that a distinctive cue reﬂecting

43

positively on a target’s intelligence produced more positive intelligence judgments than
did an equivocal cue reﬂecting on the target’s intelligence regardless of whether an
intelligence impression helped a perceiver to accomplish his or her goals. As previously
suggested, these ﬁndings may have occurred because participants were not motivated to
form intelligence impressions beyond what could be inferred by the categorical
university afﬁliation cues present in the stimuli. Generally, this ﬁnding is supportive of
the SIDE model (Reicher et al., 1995). In a single, short-term exposure to a target,
participants formed an impression of the target’s intelligence on the basis of salient
group identity cues.

The only caveat to this support is that the SIDE model (Reicher et al., 1995)
requires a target to be visually anonymous in order for such group impression
attributions to occur. In this study, a photograph of the target was visible on stimuli. This
suggests that in some situations SIDE’s requirement of visual anonymity may not be
required for salient group cues to prime a group-cue-based impression of a target. Future
research is required to understand the speciﬁc conditions under which group impressions
(a) have primacy over individuated impressions and (b) are able to form even when
visually identifying photographs of a target are present. Although the second study did
show that individuating self-disclosures have a potent effect on extraversion
impressions, it is not possible to draw conclusions about the primacy of group
impressions because in the second study all cues to the group membership of the target
were held constant. Future research should vary interpersonally individuating cues,

categorical cues, and the impression relevance of a task to determine if categorical

44

impressions truly do form ﬁrst and, if they do, if interpersonal cues may change that
group impression when they are task relevant.

A related explanation of these ﬁndings may be that, in the short-term, zero
history impression formation, the value of categorical impressions does not affect the
impression that a perceiver forms of a target. This research assumed that categorical
impression values varied among different situations and proposed that this variance
should affect impression formation. It is possible that categorical impression values only
affect impression formation after some period of interaction between a perceiver and a
target. In this case, when it is useful to form a categorical impression of another person,
cues to a person’s group identity and the characteristics associated with that group
identity might be especially salient and greatly shape impressions. However, when it is
not useful to form a categorical impression of another person these group cues may have
less impact on impressions. Finally, when a perceiver does not interact with a target (or
perceive that future interaction with a target is likely), the perceiver may simply form an
impression of a target that is comprised by the best available categorical cues about the
target. In this study, no actual interaction between perceiver and target occurred, nor did
the perceiver anticipate the possibility of interacting with the target. Because of this,
perceivers may not have been motivated by the experimental task to move beyond the
salient group impression brought about by the group cue. Future research that allows for
greater interactivity between a target and perceiver is necessary to determine if the effect
of categorical impression value on impression formation changes as interaction occurs

between a target and a perceiver.

45

Another explanation for the ﬁndings in the ﬁrst study is that participants may
have focused on group-based impressions because they did not anticipate interacting
with the person who ostensibly owned the Facebook proﬁle presented in the stimulus.
Walther (1994) found that when people anticipated future interaction with others in
CMC, they generally show more affect, depth, and trust for their partners than when
people did not anticipate future interaction with their partners. Additionally, the
anticipation of future interaction has been shown to motivate people to attend to cues to

a target’s characteristics (Kellermann & Reynolds, 1990; Ramirez, 2007). Because the

All}, .11 Li .1"!

focus of this research was to examine the effect of the impression relevance of a

 

"F‘lﬂ- ’5 ‘

situation and the distinctiveness of a cue on initial impression formation, the ﬁrst study
is unable to conﬁrm this possibility. In order to determine if the anticipation of future
interaction causes people to be more attentive to task demands while they form
impressions of others, future research should vary the amount of anticipation of future
interaction a perceiver has with a target. Such investigations may reveal that the
anticipation of future interaction heightens not only people’s attention to the personal
characteristics of others, but also it may motivate a perceiver to pay more attention to the
speciﬁc task-demands of a situation, which may, in turn, deepen impressions that helps a
person accomplish their goals in CMC.

Another area for future work revolves around the clariﬁcation of the concepts of
distinctiveness and relevance. This research conceptualized distinctiveness as a
perceptual judgment that a perceiver made about a particular cue. That is, a cue was
distinctive when a perceiver determined that the cue clearly provided information about

a particular characteristic of the target. While the distinctiveness of a cue clearly must be

46

perceived by an observer in order for the cue to be sticky, there may also be objective
characteristics that assign greater distinctiveness to cues. Future research should
examine what these objective cue differences are in order to better understand the
concept of cue distinctiveness.

Similarly, the present study examined relevance as a perceptive characteristic.
That is, cues were task relevant when they provided information that the observer
deemed relevant to the task. In the present research, this was accomplished by giving
participants a task which varied in terms of how relevant participants should view a
speciﬁc personal characteristic to completing the task, and this induction was
instantiated by directly informing participants about the characteristic that they should
ﬁnd most relevant to the task. In the relatively controlled setting of the experiments
presented here, this induction of impression task-relevance was successful. However, in
the unregulated wilds of the Internet relevance may change according to context, and
even different observers will likely bring individual differences that could affect how
relevant people ﬁnd certain personal characteristics to completing a task. In some cases,
this may affect which cues observers select as particularly relevant. The dynamics of
task-relevance on impression formation deserve future research attention.

Boundaries, Implications, and Future Directions

The present ﬁndings may be bounded by several conditions. Just because a cue is
distinctive and offers a perceiver relevant information about a target may not always
mean that the cue is utilized by a perceiver and affects impression judgments. If a
perceiver suspects that a target may be presenting a sticky cue to receive some social

beneﬁt from that cue, it seems likely that the perceiver would not weight such a cue as

47

heavily as if it could be substantiated (Donath, 1999, 2007; Walther et al., 2009). Donath
(2007) suggested that a great deal of suspicion surrounds self-descriptions on the
Internet and thus easily manipulable cues are not as reliable and less trusted by
perceivers. Walther et al.’s ﬁndings suggest that for some judgments, from which a
person is expected to receive social beneﬁt (i. e., physical attractiveness), suspicion may
cause a perceiver to rely on cues with greater warranting value. However, for other
judgments (i. e., extraversion) self-disclosures may be trusted. Taken together, these
ﬁndings suggest that there is likely some variance in terms of the amount of suspicion
about a person’s disclosures of various characteristics. It seems possible that when
suspicion is low that a target is inﬂating his or her self-presentation, task-relevant and
distinctive cues may be enough to form an impression of a target. However, when
suspicion of inﬂated self-presentation is high, task-relevant and distinctive cues may
also need to have great warranting value in order to be sticky. Future research should
address the possibility that the degree of suspicion a perceiver has about a target’s
claims may make the warranting value of cues a necessary factor of the stickiness of
those cues.

This research examined impression formation in CMC after a single, non-
interactive exposure to a target. This represents a departure from some previous research
on impression formation in CMC that explored how impressions formed among people
who were previously unacquainted over an extended period of interaction (Ramirez,
2007; Walther, 1993). Although this research did not examine the same interactive
process of impression formation in CMC, sticky cues may also inﬂuence our

understanding of the way impression formation occurs. First, the zero-history, text-only

48

 

interactive discussions may be informed by the presence of a social networking proﬁle.
That is, interactive discussions in CMC that start off at ground zero may be relatively
rare. Because information about a discussion partner is only a Google or Facebook
search away (Ramirez, Walther, Burgoon, & Sunnafrank, 2002), it may be possible that
prior to interaction people have already formed a well-developed impression of one
another. Further, recent technological developments have allowed people to make higher
bandwidth self-presentations. For example, one is no longer limited to a text-based self-
description on the Internet, it is now possible to learn about others through posted
photographs and videos, through one’s social network—even by what a person’s ﬁ’iends
say about them. Thus, preformed impressions may inﬂuence the interaction individuals
have and subsequently the impressions they form of one another.

The present model may also inform how impressions form within the bounds of
interactive CMC. Walther’s (1992) SIPT suggests that people in CMC form impressions
of others over time, through interactive discussions wherein cues about a partner’s
identity are encoded, transmitted, and decoded through text-based interaction. The
ﬁndings presented by this research suggest that not all of these cues are equal. To
employ the building-block analogy, some cues may have the ability, even in text-based
interactive CMC, to be large blocks that have a great impact on the impressions people
form of one another, while other blocks may only add a little to a person’s impression of
another. Walther’s (1993) examination of global impression development during an
extended period of interaction in CMC, which gave participants multiple tasks over the
course of their interaction, found that global impressions developed at slower rates than

in face-to-face interaction. However, in the end impression development in CMC

49

 

approached levels similar to face-to-face interaction. It is possible that the use of focused
measures of speciﬁc impressions that are task relevant may have shown a similar pattern
of narrow impression formation as was seen in Hancock and Dunham’s (2001) work. It
is possible that if people are given a speciﬁc task (the successful completion of which
requires a focused impression of some of a partner’s characteristics) these focused
impressions may develop rather quickly as people are using a variety of information
seeking techniques to extract the cues that help them to accomplish their goals. This may
especially be the case if these cues are particularly distinctive. Future research should
address these possibilities.
Limitations

One limitation of the ﬁrst study may have been that a different dependent
variable would be more sensitive to the actual effect of sticky cues on intelligence
impression formation. In this study, participants judged targets that were ostensibly
members of the Harvard University network to be more intelligent than targets
ostensibly from the ﬁctional Westlake University network. While Harvard may have
been a reliable indicator of intelligence no matter the task, participants who were
seeking information about the success of targets on a popular quiz show may have
attended speciﬁcally to information that helped them to complete their task of predicting
the best performer on the quiz show. As such, a dependent variable such as participants’
predictions of the success of the target’s success on a quiz show may have been more
sensitive to the predicted sticky cue effects. Future research should address this

possibility.

50

Another related limitation also concerns participants’ motivation to process the
stimulus proﬁles. In order to heighten motivation to process the proﬁles accurately,
participants were told that the best proﬁle evaluator would receive a 50-dollar reward. It
is possible that this motivator too successfully caused participants to attend to stimulus
proﬁles. This may explain why, in the ﬁrst study, participants generally found targets
from the Harvard University network to be more intelligent than those from the P
Westlake University network. However, in the second study where inductions were l
more subtle manipulations of statements of self-disclosure, participants’ motivation may

have been at a more optimal level. This suggests that motivation to form an impression

 

may be an important variable that may affect the role of sticky cues in the impression
formation process. This possibility deserves future research attention.
Conclusion

This research makes an important contribution to the literature regarding how
people weight social information as they form impressions of others in CMC. Beyond
understanding how people may form impressions of one another on social network sites,
this work has a number of other practical applications. For example, understanding the
dynamics of cue distinctiveness and relevance may help researchers to better understand
what cues make a user of an online commerce system a trustworthy seller, or what
signals might make an online health information diffuser appear to be most trustworthy.
In addition to the theoretical research directions outlined above, research into the
practical applications of how sticky cues may help both users to navigate social
processes on the Internet and designers of CMC systems to build systems that intuitively

facilitate these social processes.

51

Appendix A
Participant Instructions

Thank you for taking part in this study of participant evaluation. The Department of
Communication has agreed to conduct this research on behalf of the [Admissions Ofﬁce
/ Jeopardy College-Week Talent Committee]. Today, your role will be to help us by
judging several of your fellow students’ Facebook proﬁles.

The [Admissions Ofﬁce / Jeopardy College-Week Talent Committee] is implementing a
new program for [high school seniors who are unfamiliar with MSU that lets them spend
a night in the dorms socializing with a host student / MSU students to demonstrate their
mathematical abilities on a team selected to compete in a mathematical problem solving
contest]. Your role will be to evaluate how good a [host the students you’ve been
assigned / Jeopardy contestant] these other MSU students would be. The ideal candidate
for this program would be someone who is [extremely extraverted and sociable /
extremely intelligent].

So that you have some information to base your judgment on, you will be directed to a
Facebook proﬁle for each of the students you’ve been assigned to. There should be
plenty of information on these proﬁles for you to make a judgment.

Aﬁer you feel that you have formed an adequate impression of the candidate selected for
you to evaluate, you may click the link at the bottom of the webpage with questions
about the candidate.

Please take as much time as you need to form an impression of the candidate.

52

 

 

 

Appendix B
Intelligence Impression Measure

Personality questions: Please read the following statements carefully and click the circle
to mark the choice that best describes the extent to which you agree or disagree with the
statement about the person whose information you just looked at. Use the following
choices to indicate whether your agreement with the statement.

Describe the other person honestly and as accurately as possible. Even if you did not get
much information that would deﬁnitely tell you how to answer the question, indicate
your most likely impression. fl

 

1. This person is intelligent.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7 i
2. This person knows lots of facts.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7
3. This person is very smart.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7
4. This person enjoys intellectual challenges.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7
5. I could trust this person’s answer to almost any question to be correct.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

53

 

Appendix C
Group Perception Study

In this research study you will be asked to look at some facts about several individuals
and answer some questions about those individuals. Your participation should take no
longer than ten minutes.

You will only be given a small amount of information about the individuals in the study,
please do your best to answer the questions even though you have only a small amount
of information about the person. a

PERSON #1
Fact: This person is a student at Harvard University.

 

1. I feel very conﬁdent about my impression of the person’s intelligence.
Strongly Neutral Strongly g
Disagree Agree .21.
l 2 3 4 5 6 7

2. It is difficult to make a judgment of this person’s intellectual abilities.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

3. Knowing what I know about the person I can judge his/her intelligence

accurately.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

4. I am uncertain whether or not this person is smart.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

5. I am conﬁdent in my judgment about whether he/she knows a lot of facts.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

54

Appendix D

Intelligence Impression Relevance Scale

1. In order to complete this task I need information about the person’s
intelligence.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7
2. It is not vital to this task to form an intelligence impression.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7
3. Paying attention to information about whether a person is very smart or not is
very important to this task.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7
4. To complete this task it is NOT important to know anything about a person’s
intelligence.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7
5. To complete this task it would be good to know whether a person is bright.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7
6. Knowing whether a person is brilliant is relevant to this task.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

55

Appendix E
Extraversion Impression Measure

Personality questions: Please read the following statements carefully and click the circle
to mark the choice that best describes the extent to which you agree or disagree with the
statement about the person whose information you just looked at. Use the following
choices to indicate whether your agreement with the statement.

Describe the other person as honestly and as accurately as possible. Even if you did not
get much information that would deﬁnitely tell you how to answer the question, indicate _
your most likely impression. F

 

1. This person likes to have a lot of people around.
Strongly Neutral Strongly
Disagree Agree ..
1 2 3 4 5 6 7 t
2. This person laughs easily.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7
3. This person would rather go his/her own way than be a leader of others.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7
4. This person really enjoys talking to people.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7
5. This person usually prefers to do things alone.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

56

10.

ll.

12.

This person is a very active person.

Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

This person doesn't consider him/herself especially "light hearted."

Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

This person likes to be where the action is.

Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

This person often feels as if he/she is bursting with energy.

Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

This person is a cheerful, high-spirited person.

Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

This person is not a cheerful Optimist.

Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

This person’s life is fast-paced.

Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

57

Appendix F
Extraversion Impression Relevance Scale

In order to complete this task I need information about the person’s

extraversion.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7 E

It is not vital to this task to form an extraversion impression.

Strongly Neutral - Strongly
Disagree Agree
1 2 3 4 5 6 7

 

Paying attention to information about whether a person is a people-person is
very important to this task.

Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

To complete this task it is NOT important to know anything about a person’s
sociability.

Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7
To complete this task it would be good to know whether a person is
outgoing.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

Knowing whether a person is a people-person is relevant to this task.

Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

58

Appendix G
Person Perception Study

In this research study you will be asked to look at some facts about several individuals
and answer some questions about those individuals. Your participation should take no
longer than ten minutes.

You will only be given a small amount of information about the individuals in the study,
please do your best to answer the questions even though you have only a small amount
of information about the person.

PERSON #1
Fact: This person makes the following claim about him/herself:
“I am a sociable people-person...until you get to know me, then
I’m EXTREMELY extraverted and outgoing!!!”

1. I feel very conﬁdent about my impression of the person’s extraversion.
Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

2. It is difﬁcult to make a judgment of this person’s sociability.

Strongly Neutral Strongly
Disagree Agree

1 2 3 4 5 6 7

3. Knowing what I know about the person I can judge his/her outgoingness
accurately.

Strongly Neutral Strongly
Disagree Agree

1 2 3 4 5 6 7

4. I am uncertain whether or not this person is a people-person.

Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

59

.'—-’.e...t!‘1 --__. -"

 

5.

I am conﬁdent in my judgment about whether he/she is sociable.

Strongly Neutral Strongly
Disagree Agree
1 2 3 4 5 6 7

60

Appendix H
Table 1.

Item - Factor Loadings

 

 

 

 

Factors

Study 1 Study 2
Items 1 2 1 2
1. This person knows lots of facts. .77* .13 .84* .13
2. This person is very smart. .72* .09 .89* .20
3. This person enjoys intellectual challenges. .90* .15 .84* .13
4. I could trust this person’s answer to almost .43* -.05 .63* .04
any question to be correct.
5. This person likes to have a lot of people .21 .43* .08 .77*
around.
6. This person would rather go his/her own .16 .83* .21 .54*
way than be a leader of others.
7. This person really enjoys talking to -.23 .66* .18 .75*
people.
8. This person is a very active person. .11 .67"‘ .22 .67*
9. This person likes to be where the action is. .03 .82* .06 .74*
10. This person often feels as if he/she is --** --** -.01 .69*
bursting with energy.
11. This person is a cheerful, high-spirited —.01 .56* .15 .76*
person.
12. This person is not a cheerful optimist. .08 .79* .03 .36*
13. This person’s life is fast-paced. .21 .66* .01 .59*

 

Notes. Factor loadings were estimated using the centroid method. * indicates factor hypothesized to drive
item. Factors 1 and 2 are intelligence and extraversion, respectively. ** indicates item was excluded from
analysis.

61

 

 

 

 

 

 

Appendix I
Table 2.
Contrast Weights for Predictions
High Cue Low Cue Control
Distinctiveness Distinctiveness
High Low High Low High Low
Cue Cue Cue Cue Cue Cue
Releva Releva Releva Releva Releva Releva
nce nce nce nce nce nce
Study 1:
Intellig
ence +3 -1 -1 -1
Judgrne
nts
Study 2:
Extrave
rsion +3 +1 +1 +1 -3 -3
Judgrne
nts

 

62

 

 

Table 3.

Means (and Standard Deviations)

Appendix J

 

 

 

 

 

 

High Cue Low Cue Control
Distinctiveness Distinctiveness
High Low High Low High Low
Cue Cue Cue Cue Cue Cue
Releva Releva Releva Releva Releva Releva
nce nce nce nce nce nce
Study 1:
Intellig 4.92 5.08 4.59 4.47
ence (0.63) (1.19) (0.88) (0.94)
Judgm n=18 n=l9 n=23 n=l9
ents
Study 2:
Extrav 4.79 4.53 4.50 4.50 4.13 4.27
ersion (0.98) (0.66) (0.89) (0.76) (0.74) (0.87)
Judgm n=18 n=18 n=18 n=18 n=18 n=18
ents

 

63

 

”I

 

1171*, .
q .

 

 

    

View More; of (.MS (6?)
Suggest friends for Clans
Sent: Chi-1‘s a Message
Poke Ch.- is

Maturation
.q. ,,_ 3

Harvard University

7‘ . _ . ..
P.“ .1 .‘V '1 'a- ‘r-I'.‘

Single

May 26. 1987

Friends

 

253 friends See All
I .
lcf' Our: e Run
08mm Mathew/1:2 ammo.

Eru‘ Quinn Lad

Quincy TuPasue Ebert:

Mona: who. .

Chris Smith
Harvard University

 

 

Wall Info Photos

Basic Information

Personal Information

3.

Conmt Information

Education

Harvard Unwencty

May 26. 1987
Single

Hangar; out wit found». matting. Mir-ﬁg, other; to rr-y
3.?191’25' (37mg; rn rl‘r: unmet. snort-1

Lost. tic-use. Bent-3. T't'w‘Gt. BJf‘v the 'Ja‘rrcnre Slaver. CSI
I'm a really bog U2 tan.

Detail: lI

Horus: lJ" vcz's'tv ‘39
Cream. Frencr

DillJIU Hglr Sci 0c? ‘05

YMCA

Volunteer Sup-Aer! rostr-

Septembe" 2007 . Present

I'm a volunteer kids basketball coach.

Study 1 Sample Stimulus Material

64

 

I:
w,

' I-!'l‘
‘

 

 

 

   

View Flt-3195 at {Trim (6 7)
Suggest Fuenos for Cnris
Set“: Chris .1 Message

Poke this

Information
I‘. f ff. .’ r
Michigan State University

..~~r~.\

3-‘4

Single
May 26. 1937

Friends
ZS] ‘r ends

[rvrl

wa‘KY

Go on
Tepastte

G
‘3} ,'
»V:“.,.-.‘r
4‘ ‘

LC'I
[barge

 

Chris Smith

Figure 2

 

Michigan State University

Wall Info

Basic Information

o.-

Personal Infovmatlon

Contact Information

~

Education

Phoros

Michigan State University

May 26. 1987
Single

Hangar; 01.! a {1" then-'15.. read “.1, ﬂ'll rig. gong to my
parents cottage. in the 53l“"“’€(. sows

Last, Home. (ﬁnes, l'r 'lgc. Bu‘lv the trawl)": Slayer. C51

I am a sociable people person ..until 701.: get to know me. then
I'm EXTREMELY extraverted and outgoangw.

Dumb. l‘

M (higan Starc- Us' versity '99
("gram I’¢”<I'

Denali; ng'l Sweet ‘05

YMCA

VOI'JI‘ICQ' Sup-xv? Coach

September 2007 - Present

I'm a volunteer kids basketball coach.

Study 2 Sample Stimulus Material

65

 

 

REFERENCES

Albright, L., Kenny, D. A., & Malloy, T. E. (1988). Consensus in personality judgments
at zero acquaintance. Journal of Personality and Social Psychology, 55, 387-3 95.

Ambady, N., & Rosenthal, R. (1993). Half a minute: Predicting teacher evaluations from
thin slices of behavior and physical attractiveness. Journal of Personality and
Social Psychology, 64, 431-441 .

boyd, d., & Ellison, N. B. (2007). Social network sites: History, deﬁnition, and
scholarship. Journal of Computer-Mediated Communication, 13, 210-230.

Brunswik, E. (1956). Perception and the representative design of psychological
experiments. Berkeley, CA: University of California Press.

Burton, M. C., & Walther, J. B. (2001). The value of web log data in use-based web
design and testing. Journal of Computer-Mediated Communication, 6 (3).
Retrieved October 29, 2008 from
http://jcmc.indiana.edu/vol6/issue3/burton.html.

Costa, P. T., & McCrae, R. R. (1991). NEO-FFI Form S. Odessa, FL: Psychological
Assessment Resources.

Donath, J. (1999). Identity and deception in the virtual community. In M. A. Smith & P.
Kollock (Eds), Communities in cyberspace (pp. 29-59). New York: Routledge.

Donath, J. (2007). Signals in social supemets. Journal of C omputer-Mediated
Communication, 13, 23 1-251 .

Epley, N., & Kruger, J. (2005) When what you type isn’t what they read: The
perseverance of stereotypes and expectancies over e-mail. Journal of
Experimental Social Psychology, 41, 414-422.

Eysenck, H. J ., & Eysenck, M. W. (1985). Personality and individual diﬁkrences: A
natural science approach. New York: Plenum.

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. In M. P. Zanna (Ed.), Advances in
experimental social psychology (Vol. 23, pp. 1-74). San Diego, CA: Academic
Press.

66

Gigerenzer, G. (1993). The bounded rationality of probabilistic mental models. In K. I.
Manktelow & D. E. Over (Eds.), Rationality: Psychological and philosophical
perspectives (pp. 284-313). London: Routledge.

Gigerenzer, G., & Goldstein, D. (1996). Reasoning the fast and frugal way: Models of
bounded rationality. Psychological Review, 103, 650-669.

Gigerenzer, G., Hoffrage, U., & Kleinbtilting, H. (1991). Probabilistic mental models: A
Brunswikian theory of conﬁdence. Psychological Review, 98, 506-528.

Gigerenzer, G., & Kurz, E. M. (2001). Vicarious functioning reconsidered: A fast and
frugal lens model. In K. R. Hammond & T. R. Stewart (Eds.), The essential
Brunswik: Beginnings, explications, applications (pp. 342-348). Oxford: Oxford
University Press.

Hancock, J. T., & Dunham, P. T. (2001). Impression formation in computer-mediated
communication revisited: An analysis of the breadth and intensity of impressions.
Communication Research, 28, 325-347.

Herring, S. C., & Martinson, A. (2004). Assessing gender authenticity in computer-
mediated language use. Jounal of Language and Social Psychology, 23, 424-446.

Hunter,J. E., & Gerbing, D. W. (1982). Unidimensional measurement, second order

factor analysis, and causal models. Research in Organizational Behavior, 4, 267-
320.

Jacobson, D. (1999). Impression formation in cyberspace: Online expectations and
ofﬂine experiences in text-based virtual communities. Journal of Computer-
Mediated Communication, 5 (l ). Retrieved February 28, 2008 from
http://www.ascusc.org/jcmc/volS/issue1/jacobson.html.

Kellermann, K., & Reynolds, R. (1990). When ignorance is bliss: The role of motivation
to reduce uncertainty in uncertainty reduction theory. Human Communication
Research, I 7, 5-75.

Kelley, H. H. (1967). Attribution theory in social psychology. In D. Levine (Ed.), Nebraska
symposium on motivation (V 01. 15, pp. 192-23 8). Lincoln, NE: University of
Nebraska Press.

Kelley, H. H. (1973). The process of causal attribution. American Psychologist, 28, 107-
128.

Keppel, G., & Wickens, T. D. (2004). Design and analysis: A researcher ’s handbook
Upper Saddle River, NJ: Pearson Prentice Hall.

67

Kiesler, S., Siegel, J. & McGuire, T.W. (1984). Social psychological aspects of computer--
mediated communication. American Psychologist, 39, 1123-1134.

Klein, 0., Spears, R., & Reicher, S. (2007). Social identity performance: Extending the
strategic side of the SIDE model. Personality and Social Psychology Review, 11,
28-45.

Krosnick, J. (1991). Response strategies for dealing with the cognitive demands of
attitude measures in surveys. Applied Cognitive Psychology, 5, 213-236.

Maheswaran, D., & Chaiken, S. (1991). Promoting systematic processing in low-
motivation settings: Effect of incongruent information on processing and
judgment. Journal of Personality and Social Psychology, 61, 13-25.

Marcus, B., Machilek, F., & Schlitz, A. (2006). Personality in cyberspace: Personal web
sites as media for personality expressions and impressions. Journal of
Personality and Social Psychology, 90, 1014-1031.

McCroskey, J., & McCain, T. (1974). The measurement of interpersonal attraction.
Speech Monographs, 41, 261-266.

Miller, G. R., & Steinberg, M. (1975). Between people: A new analysis of interpersonal
communication. Chicago, IL: Science Research Associates, Inc.

Oberlander, J. and Gill. A. J. (2006). Language with character: A corpus-based study of
individual differences in e-mail communication. Discourse Processes, 42, 239-
270.

O’Keefe, D. J. (1990). Persuasion: Theory and research. Newbury Park, CA: Sage.

Orvis, B. R., Cunningham, J. D. & Kelley, H. H. (1975). A closer examination of causal
inference: The roles of consensus, distinctiveness, and consistency information.
Journal of Personality and Social Psychology, 32, 605-616.

Peﬁa, J ., & Hancock, J. T. (2006). An analysis of socioemotional and task-oriented

communication in an online multiplayer video game. Communication Research,
33, 92-109.

Petty, R. E., Cacioppo, J. T., & Goldman, R. (1981). Personal involvement as a
determinant of argument-based persuasion. Journal of Personality and Social
Psychology, 41, 847-855.

Postmes, T., & Baym, N. (2005). Intergroup dimensions of the Internet. In J. Harwood
& H. Giles (Eds.) Intergroup communication: Multiple perspectives (pp. 213-
238). New York: Peter Lang.

68

 

 

Postmes, T., Spears, R., Sakhel, K., & de Groot, D. (2001). Social inﬂuence in
computer-mediated groups: The effects of anonymity on social behavior.
Personality and Social Psychology Bulletin, 27, 1243-1254.

Ramirez, Jr., A. (2007). The effect of anticipated future interaction and initial impression
valence on relational communication in computer-mediated interaction.
Communication Studies, 58, 53-70.

Ramirez, Jr., A., Walther, J. B., Burgoon, J. K., & Sunnafrank, M. (2002). Information
seeking strategies, uncertainty, and computer-mediated communication: Toward
a conceptual model. Human Communication Research, 28, 213-228.

Reicher, S. D., Spears, R., & Postmes, T. (1995). A social identity model of deindividuation
phenomena. European Review of Social Psychology, 6, 161-198.

Rogers, P., & Lea, M. (2004). Cohesion in online groups. In K. Morgan, C. A. Brebbia,
J. Sanchez, & A. Voiskounsky (Eds.), Human perspectives in the Internet
society: Culture, psychology and gender (pp. 115-124). Southampton, UK: WIT
Press.

Rutter, D. R., Morley, I. E., & Graham, J. C. (1972). Visual interaction in a group of
introverts and extraverts. European Journal of Social Psychology, 2, 371-3 84.

Sassenberg, K. (2002). Common bond and common identity groups on the Internet:
Attachment and normative behavior in on-topic and off-topic chats. Group
Dynamics: Theory, Research, Practice, 6, 27-37.

Sassenberg, K., & Boos, M. (2003). Attitude change in computer-mediated
communication: Effects of anonymity and category norms. Group Processes &
Intergroup Relations, 6, 405-422.

Sassenberg, K., & Postmes, T. (2002). Cognitive and strategic processes in small
groups: Effects of anonymity of the self and anonymity of the group on social
inﬂuence. British Journal of Social Psychology, 41, 463-480.

Simon, H. A. (1945). Administrative behavior: A stuay of decision-making proceses in
administrative organization. New York: Free Press.

Simon, H. A. (1956). Rational choice and the structure of the environment.
Psychological Review, 63, 129-138.

Simon, H. A. (1982). Models of bounded rationality. Cambridge, MA: MIT Press.
Tidwell, L. C., & Walther, J. B. (2002). Computer-mediated communication effects on

disclosure, impressions, and interpersonal evaluations: Getting to know one another
a bit at a time. Human Communication Research, 28, 317-348.

69

Todorov, A. (2000). The accessibility and applicability of knowledge: Predicting the
context effects in national surveys. Public Opinion Quarterly, 64, 429-451.

Tong, S. T., Van Der Heide, B., Langwell, L., & Walther, J. B. (2008). Too much of a good
thing? The relationship between number of friends and interpersonal impressions on
Facebook. Journal of Computer-Mediated Communication, 13, 531-549.

Turner, J. C. (1982). Towards a cognitive redeﬁnition of the group. In H. Tajfel (Ed.),
Social identity and intergroup relations (pp. 15-40). Cambridge: Cambridge
University Press.

Turner, J. C. (1985). Social categorization and the self-concept: A social cognitive
theory of group behavior. In E. J. Lawler (Ed.), Advances in group processes:
Theory and research (Vol. 2, pp. 77-122). Greenwich, CT: JAI Press.

Utz, S. (2000). Social information processing in MUDs: The development of friendships
in virtual worlds. Journal of Online Behavior, 1(1). Retrieved October 25, 2008
from http://www.behavior.net/JOB/v1nl/utz.html

Van Der Heide, B., & Walther, J. B. (2009, May). Computer-mediated communication
and persuasion: Testing a social information processing/heuristic-systematic
synthesis. Paper presented at the annual meeting of the International
Communication Association, Chicago.

Vazire, S., & Gosling, S. D. (2004). e-Perceptions: Personality impressions based on
personal websites. Journal of Personality and Social Psychology, 87, 123- 132.

Walther, J. B. (1992). Interpersonal effects in computer-mediated interaction: A relational
perspective. Communication Research, 19, 52-90.

Walther, J. B. (1993). Impression development in computer-mediated interaction. Western
Journal of Communication, 5 7, 381-398.

Walther, J. B. (1994). Anticipated ongoing interaction versus channel effects on relational

communication in computer-mediated interaction. Human Communication
Research, 20, 473-501.

Walther, J. B. (1996). Computer-mediated communication: Impersonal, interpersonal, and
hyperpersonal interaction. Communication Research, 23, 3-43.

Walther, J. B., Loh, T., & Granka, L. (2005). Let me count the ways: The interchange of
verbal and nonverbal cues in computer-mediated and face-to-face afﬁnity. Journal
of Language and Social Psychology, 24, 36-65.

70

Walther, J. B., & Parks, M. R. (2002). Cues ﬁltered out, cues ﬁltered in: Computer-
mediated communication and relationships. In M. L. Knapp & J. A. Daly (Eds),

Handbook of interpersonal communication (3rd ed., pp. 529-563). Thousand
Oaks, CA: Sage.

Walther, J. B., Van Der Heide, B., Hamel, L., & Shulman, H. (in press). Self-generated
versus other-generated statements and impressions in computer-mediated

communication: A test of warranting theory using Facebook. Communication
Research, 36, 229-253.

Walther, J. B., Van Der Heide, B., Kim, S., Westerman, D., & Tong, S. T. (2008). The
role of friends’ behavior on evaluations of individuals’ F acebook proﬁles: Are

we known by the company we keep? Human Communication Research, 34, 28-
49.

Weisband, S., & Atwater, L. (1999). Evaluating self and others in electronic and face-to-
face groups. Journal of Applied Psychology, 84, 632-639.

Yee, N. (2007). Motivations of play in online games. Journal of CyberPsychology and
Behavior, 9, 772-775.

71

 

YLIBR

 

llllHillHilllllllllllllllllllllllﬂll llllllllss

1B