This is to certify that the
dissertation entitled

On the Measurement of Intraindividual Personality Variability

presented by

Brendan Michael Baird

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Psycholgcly

 

 

77’5’k

 

Major Professor’s Signature
3/5; cl.

Date

 

MSU is an Affirmative Action/Equal Opportunity Institution

 

 

 

LIBRARY
Michigan State
University

 

 

 

o
.
.
u
.
a
u
o
u
a
.
v
.
u
u

 

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/C|RCIDateDue.indd-p.1

 

 

 

 

 

ON THE MEASUREMENT OF INTRAINDIVIDUAL PERSONALITY VARIABILITY
By

Brendan Michael Baird

A DISSERTATION
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Psychology

2006

ABSTRACT
ON THE MEASUREMENT OF INTRAINDIVIDUAL PERSONALITY VARIABILITY
By
Brendan Michael Baird
Intraindividual personality variability, which is the extent that a person’s behavior
changes over time or across situations, is a centerpiece of modern research in personality
psychology. Previous research has shown that individual differences in variability are
stable and may even have important implications for psychological health. However,
when certain self-reports methods are used to assess this trait, response styles may
contribute reliable, yet irrelevant variance to indexes of intraindividual variability. In
two studies, I examined the validity of various self-report indexes of variability and used
these measures to test predictions about broad outcomes for psychological well-being.
Findings indicate that individuals who provide self-reports of personality that are
inconsistent across roles or over time also tend to report greater variability in judgments

of conceptually unrelated phenomena.

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................ iii
LIST OF FIGURES .............................................................................................. iv
INTRODUCTION ................................................................................................ 1
THEORETICAL BACKGROUND ...................................................................... 2
On the Nature of Intraindividual Variability ............................................ 4
Implications of Intraindividual Variability ............................................... 6
ASSESSING INTRAINDIVIDUAL VARIABILITY ......................................... 10
Top-Down Approaches ............................................................................ 10
Bottom-Up Approaches ........................................................................... 12
Response Styles in Variability Measures ................................................. 15
Extreme Responding .................................................................... l6

OVERVIEW OF STUDIES ................................................................................. 19
STUDY 1 ............................................................................................................. 20
Method ..................................................................................................... 20
Participants .................................................................................. 20

Materials ...................................................................................... 20

Analytic technique ....................................................................... 22

Results ..................................................................................................... 23
Discussion ................................................................................................ 28
STUDY 2 ............................................................................................................. 30
Method ..................................................................................................... 32
Participants .................................................................................. 32

Materials ...................................................................................... 33

Global reports .................................................................. 33

Daily reports .................................................................... 34

Procedure ..................................................................................... 34

Analytic technique ....................................................................... 35

Results ..................................................................................................... 36
Discussion ................................................................................................ 40
GENERAL DISCUSSION .................................................................................. 42
Summary of Findings .............................................................................. 42
Implications ............................................................................................ 43
Limitations and Future Directions .......................................................... 47

iii

CONCLUSION .................................................................................................... 49

REFERENCES .................................................................................................... 50
FOOTNOTES ...................................................................................................... 57
APPENDICES ..................................................................................................... 74

iv

LIST OF TABLES

Table l. Bivariate correlations among indexes of personality variability, response style,
and relationship satisfaction (Study 1).

Table 2. Means and standard deviations of well-being measures (Study 2).

Table 3. Bivariate correlations among indexes of personality variability and response
style (Study 2).

LIST OF FIGURES

Figure 1. Latent-trait model of variability (Study 1).

Figure 2. Path diagram of relationships among latent variability factors and self—concept
clarity (Study 1).

Figure 3. Path diagram of relationships among latent variability factors and relationship
satisfaction (Study 1).

Figure 4. Latent-trait model of variability and response style (Study 1).

Figure 5. Path diagram of relationships among latent variability factors, self-concept
clarity, and response style (Study 1).

Figure 6. Path diagram of relationships among latent variability factors, relationship
satisfaction, and response style (Study 1).

Figure 7. Latent-trait model of bottom-up variability from global and daily reports
(Study 2).

Figure 8. Path diagram of relationships among latent factors of bottom-up variability and
self-consistency (Study 2).

Figure 9. Path diagram of relationships among latent factors of bottom-up variability and
well—being (Study 2).

Figure 10. Latent-trait model of variability and response style (Study 2).

Figure 11. Path diagram of relationships among latent factors of bottom-up variability
and response style (Study 2).

Figure 12. Path diagram of relationships among latent factors of bottom-up variability,
self-consistency, and response style (Study 2).

Figure 13. Path diagram of relationships among latent factors of bottom-up variability,
well-being, and response style (Study 2).

vi

INTRODUCTION

Monday morning, 9:00am. A woman sits in a board meeting, listening attentively
as her fellow executives give their weekly reports. On a notepad in front of her is a list
of questions that she prepared in advance and a set of detailed notes that grow longer
with each presentation. Every once in a while, the woman raises her hand to ask pointed
questions or to interject her views on a topic. When it is time, she promptly adjourns the
meeting in order to be on time for her next appointment.

Thursday evening, 9:00pm. A woman deals cards to a group of friends during
their weekly poker game. As other people share stories about their vacations and tell
dirty jokes, the woman quietly keeps score and watches bubbles form in her soft drink.
Each time it is her turn to wager, she cautiously considers her chances of winning and
often folds without placing a bet. As the game wears on, she yawns widely and looks
forward to a quiet weekend.

The personalities depicted in these scenarios are obviously quite different from
one another. The woman in the first scenario is extraverted and conscientiousness,
whereas the woman in the second scenario is introverted and perhaps a little neurotic.
But what if both descriptions were actually about the same person? How might we
describe her personality, given the differences in her behavior from one situation to the
next? Perhaps it would be useful to say that the woman is inconsistent, variable, or
unstable. We might also want to understand why the woman behaves so differently at
work than she does around her friends. Perhaps there is something unique about

someone whose behavior changes a great deal from situation to situation.

THEORETICAL BACKGROUND

Personality psychologists are increasingly interested in the kinds of people that
exhibit a lot of variability in their behavior, as well as in the processes that might produce
this individual difference (Cervone, 2004; Cloninger, 2003; Fleeson, 2004; Mischel &
Shoda, 1995; Pervin, 2001; Roberts & Pomerantz, 2004). In order to make inferences
about these processes, researchers have come up with a variety of self-report methods to
study how people change. However, in many instances, it is unclear whether these
methods are able to capture constructs that are psychologically meaningful. Therefore, to
it is important to rigorously examine the psychometric properties of intraindividual
variability measures and to identify potential limitations of the information they provide.

Most psychologists acknowledge that environmental factors have the potential to
exert powerfiil influences on behavior, yet they often disagree about the meaning of
changes that occur in people’s behavior across situations. For example, it would be hard
to argue against the notion that people tend to behave differently at work than they do
when playing card games. However, there are different views about why these changes
occur. On one hand, differences in people’s behaviors across situations may simply be
due to contextual factors that have a similar impact on everyone. If so, then the ways
people change across situations may have little to do with personality. On the other
hand, it may also be the case that some people are more sensitive to situational influences
than are other people and, as a result, change their behavior more dramatically (Snyder,
1974). In other words, there may be stable personality processes that moderate the
impact of situational factors on behavior, leading some people to change their behavior a

great deal across situations and other people to behave the same way all the time

(Mischel & Shoda, 1995). If intraindividual variability is an individual difference, then
research should focus on identifying the processes and outcomes related to it.

Although most people may tend to be outgoing and talkative around their friends,
and quiet and contemplative in school, there are certain kinds of people who behave the
same way in each situation and others whose behavior varies widely (Bem & Allen,
1974; Fleeson, 2001; Lanning, 1988; Paulhus & Martin, 1988). In one study,
Schneiderman (1980) asked participants to act out the role of a teacher in each of three
different ways: introverted, extraverted, and natural. Across these three conditions,
extraverted behaviors were most common in the extraverted condition and least common
in the introverted condition, demonstrating that most people were able to act
appropriately in each condition. However, people who described themselves as being
highly variable showed the most change in observable behavior across the conditions. In
another study, Fleeson (2001) found that extraversion varied as a fiinction of the time of
day and the number of people present at a given moment, but that a great deal of
intraindividual variability in extraversion could also be attributed to participants’
personalities.

In addition to observable behavior, there may also be meaningful individual
differences in the variability of other aspects of personality. For instance, there is no
question that affect changes as a function of the situations that people encounter, yet
research has shown that some people exhibit more variability in their moods across
situations than other people (Diener & Larsen, 1984; Penner, Shiffman, Paty, &
Fritzsche, 1994). Similarly, although cycles of physical activity and inactivity can

account for some of the changes in moods that occur throughout the course of a day or

across different days of the week, there are substantial individual differences in the extent
to which people’s moods fluctuate over time (Larsen, 1987; Larsen & Kasimatis, 1990;
Watson, 2000). Furthermore, intraindividual variability in affect has been linked to
broad personality traits, such as extraversion and neuroticism (Eid & Diener, 1999;
McConville & Cooper, 1999). Therefore, although some situations tend to evoke
positive feelings and other situations tend to evoke negative feelings, certain kinds of
people have moods that change a great deal from situation to situation whereas others
have moods that rarely change at all.

Like many other aspects of personality, intraindividual variability appears to be
quite stable over time (Baumeister, 1991). For instance, F leeson (2001) reported three-
week stability estimates for variability scores on the Big Five that were often above .75,
and a recent study of longer-term stability found that variability scores correlated roughly
.70 across six to nine months (Baird, Le, & Lucas, 2006). Considering that mean-levels
of the Big Five traits are only slightly more stable than variability scores, these findings
are quite remarkable and suggest that variability may be an important component of
personality. However, psychologist have a great deal yet to learn about the nature of
personality variability.

0n the Nature of Intraindividual Variability

Changes in a person’s behavior can occur for a number of reasons (F iske & Rice,
1955). For instance, physiological processes, such as arousal in the cardiovascular or
central nervous systems, can produce changes in observable behavior. A person might
also behave differently around certain people than he or she does around other people.

Finally, a person may behave differently from one time to another because of events that

occur in between observations. Although each of these processes are interesting in its
own right, they all add up to individual differences in intraindividual variability that can
be studied empirically.

When studying individual differences in intraindividual variability, it is important
to consider the variety of ways that change can be conceptualized (Nesselroade, 1992;
Tellegen, 1988). So far I have focused only on short-term variability that occurs over
brief intervals of time or across situations. However, psychologists are often interested in
how pe0ple behave at different points throughout the lifespan and seek to understand
changes that occur over the course of several years. In contrast to the spontaneous and
temporary fluctuations in behavior across situations that are the focus of this paper,
longer-term changes constitute shifts in baseline behaviors that occur gradually. As such,
they may be theoretically independent of shorter-term variability. More importantly, the
processes underlying developmental shifts may be different from those that produce
changes in personality across situations.

Another issue that should be addressed involves determining the appropriate unit
of analysis. Some theorists have argued that variability is multi-faceted and that people
can differ from one another in both the number of traits and the kinds of traits upon
which they vary (Baumeister & Tice, 1988; F leeson & Jolley, in press; Mischel & Shoda,
1999). Furthermore, the processes that govern variability in one trait may be different
fiom those that govern variability in another trait. In other words, the extent to which a
person is variably extraverted may be independent of his or her tendency to be variably
neurotic. According to this perspective, it is important to assess the variability in each

trait separately.

In contrast, other researchers have argued that different types of variability,
although conceptually distinct, are actually linked to the same underlying mechanism.
For instance, variable people may behave differently from one situation to another
because they are particularly sensitive to social influences (Snyder, 1974). When
situational cues indicate that a person’s typical ways of behaving might be inappropriate
at a given moment, he or she may vary in extraversion and neuroticism simultaneously.
In support of this perspective, much of the research that has examined variability in
multiple traits has shown that trait-specific indexes of variability tend to be strongly
correlated with one another and that these relationships can be explained by a single
factor (Baird et al., 2006; Bid & Diener, 1999; Penner et al., 1994). This suggests that
intraindividual variability may be a broad trait that is distinct from other dimensions of
personality. Therefore, it is important to consider whether this trait is associated with
outcomes in other areas of psychological functioning.

Implications of Intraindividual Variability

Perhaps the most important outcomes that have been linked to intraindividual
variability are those that relate to psychological health. Although a great deal of research
has been done on the topic, theorists have tended to disagree about whether variability is
associated with well-being. Some theorists believe that variability reflects an ability to
adapt to the environment (Bern & Lewis, 1975; Paulhus & Martin, 1988; Snyder, 1974).
For example, Bern and Lewis (1975) argued that individuals who can behave in ways that
are both feminine and masculine are likely to behave appropriately across different kinds
of situations. Similarly, individuals who can engage in a wide range of behaviors may be

better able to act in ways that fit a variety of situations than are rigid people (Paulhus &

Martin, 1988). Therefore, if a person acts differently around friends than he or she acts
around family, then this may be an indication that the person is adapting appropriately to
each situation. According to this perspective, intraindividual variability should be
associated with positive outcomes.

In contrast, other theorists have argued that variability is a sign of psychological
immaturity (Jourard, 1963; Maslow, 1968; Rogers, 1961). For instance, Maslow (1968)
argued that healthy people are able to transcend the environment and express their true
selves by resisting social pressures. When people succumb to situational influences, they
must expend a great deal of psychological energy to repress their inner selves, and
ultimately will experience lower levels of well-being as a result. Similarly, people who
are variable may lack a sense of self-direction (Rogers, 1961), and may be more worried
about what other people expect of them rather than about their own values (J ourard,
1963). In other words, intraindividual variability may result from conflicts between
one’s dispositional tendencies and external pressures to behave in ways that are socially
appropriate, but which may not come naturally to the person. According to this view,
intraindividual variability should be associated with lower levels of psychological well-
being.

Finally, if personality variability is the result of heightened sensitivity to social
influences, then it may also have an impact on the quality of interpersonal relationships.
Of all the external pressures that people encounter, those that are imposed by family and
friends may be among the strongest and most influential. In fact, over time, these
relationships have the potential to alter one’s sense of self and to shape important aspects

of one’s identity (Aron, Aron, Tudor, & Nelson, 1991). To the extent that an individual’s

family and friends share similar expectations for his or her behavior, the person should
experience few conflicts about how to behave. Furthermore, according to self-
determination theory (Ryan & Deci, 2000), social relationships that promote a sense of
autonomy and self-direction should lead to positive outcomes for health and well-being.
In other words, relationships in which people are able to determine their own actions tend
to be more satisfying than those that encourage conformity or dependency. Therefore,
the quality of one’s relationships with family and friends may be a predictor of
intraindividual variability.

Alternatively, personality variability may have an adverse impact on the
development of social relationships. For instance, because variable people continually
change their behavior to please other people, they may feel that other people don’t
understand them or appreciate them for who they really are. In fact, recent empirical
evidence suggests that they might be right. Compared to people with consistent
personalities, variable people tend to be harder for others to judge (Colvin, 1993), and
tend describe their own personalities differently than other people describe them (Baird
et al., 2006; Biesanz & West, 2000). In other words, people tend to find it difficult to
understand what another person’s personality is like when that person’s behavior changes
from situation to situation. As a result, variable people may have a hard time establishing
close bonds with others and may feel dissatisfied with their social relationships.

If intraindividual variability has implications for psychological well-being, as
many theorists believe it does, then it is important that research be done to examine these
links directly. In the studies described below, I tested whether or not personality

variability was associated with self-reports of relationship satisfaction (Study 1) and

psychological health (Study 2). To provide the best possible test for competing theories
of variability, careful attention had to be given to the measurement process. Therefore, in
the next section, I outline several important details about existing variability indexes and

consider the potential strengths and weaknesses of each.

ASSESSING INTRAINDIVIDUAL VARIABILITY

The psychometric assessment of intraindividual variability can be quite
challenging and ofien requires the use of complex statistical analyses (Nesselroade, 1992;
Tellegen, 1988). Perhaps the greatest challenge, however, is simply gathering useful
information about the ways people change. In many instances, it is nearly impossible to
observe people directly as they move from one situation to the next, or to create a variety
of situations in a laboratory that can mimic the complexity of real-life situations.
Therefore, researchers often rely on self-report measures to assess variability instead.

In general, self-reports can be used to study variability in one of two ways
(Cervone, 1999). First, a researcher can simply ask people direct questions about how
variable they are. This approach, referred to as top-down measurement, has been used in
previous research on self-esteem stability (Rosenberg, 1985) and self-concept pluralism
(McReynolds, Altrocchi, & House, 2000). Alternatively, researchers can also assess
variability by examining the extent to which a person’s response varies each time he or
she is asked about a certain trait. Examples of this approach, referred to as bottom-up
measurement, have been used in research on role variability (Donahue, Robins, Roberts,
&, John, 1993) and trait scalability (Lanning, 1988; Reise & Waller, 1993). The
distinctions between top-down and bottom-up indexes are important because each
method may actually provide unique information about personality.

T op-down approaches

Certain self-report measures are designed to assess people’s perceptions of

variability in their own behavior. For example, the Self-Pluralism Scale (McReynolds et

al., 2000) includes items such as “I act and feel essentially the same whether at home, at

10

work, or with friends” and “My personality is always the same regardless of whom I’m
with or the situation I’m in.” Presumably, people who strongly endorse these statements
should exhibit little variability in their behavior. Other top—down measures focus
explicitly on the underlying processes that might lead to observable variability in
behavior. For example, the Self-Concept Clarity scale (Campbell, Trapnell, Heine, Katz,
Lavallee, & Lehman, 1996), includes items such as “My beliefs about myself often
conflict with one another” and “Sometimes I feel that I am not really the person that I
appear to be.” Similarly, Rosenberg (1989) developed a 5-item scale to capture
variability in self-esteem that includes items such as “Does your opinion of yourself tend
to change a good deal, or does it always continue to remain the same?” and “Do you ever
find that on one day you have one opinion of yourself and on another day you have a
different opinion?” These instruments may provide an ideal way to test competing
theories about the mechanisms underlying variability in observable behavior.

The ways people respond to top-down measures may also help researchers learn
about the underlying structure of variability. For instance, if variability is multi-
dimensional, then a person who gets a high score on a measure of self-concept pluralism
wouldn’t necessarily get a high score on a measure of self-esteem stability. Furthermore,
if different top—down measures capture distinct phenomena, then they should predict
different patterns of change in personality. However, for the most part, these measures
tend to be highly correlated with one another, and often predict many of the same
outcomes. This suggests that different top—down measures might actually tap into the
same global tendency to be variable.

On the other hand, strong correlations among various top-down indexes might

ll

also be due to a certain degree of item overlap across different measures. For instance,
items that assess behavioral consistency ofien appear in the same questionnaire with
items that assess affective consistency. The self-pluralism measure (McReynolds et al.,
2000) even includes an item about naming different parts of oneself and another item
about not being able to remember one’s own actions. As a result, it is unclear whether
this kind of measure captures a general tendency to vary or simply combines multiple
traits of variability into a single index. Therefore, research should be done to test
whether or not top-down indexes are associated with changes in observable behavior.

To date, the studies that have directly compared subjective judgments of
variability with measures of how much people actually change are somewhat
inconclusive. In one study, Robins, Noftle, Trzesniewski, and Roberts (2005) found that
participants’ retrospective judgments of how much their personalities changed during
college were moderately correlated with actual changes in traits that were assessed across
four years. However, in another study (Baird et al., 2006), scores on a top-down measure
were only weakly correlated with an experience-sampling measure of variability in
behavior across moments. Therefore, it is important to understand how top-down
measures differ fiom alternative indexes of personality variability.

Bottom-up approaches

In addition to capturing global perceptions of variability, researchers can use self-
reports to estimate variability from people’s responses to questions about specific aspects
of behavior. In other words, variability can also be measured from the bottom-up. One
way this can be done is by computing the internal consistency of a person’s responses

across items that measure the same trait (Bern & Allen, 1974; Lanning, 1988). For

12

example, if a measure contains multiple items that assess extraversion, then each item in
the scale can be treated as a separate indicator of the underlying trait. A person who is
consistently extraverted should endorse the items “I am talkative” and “I am outgoing”
equally, whereas a person who is variably extraverted should endorse one item but not
the other (Reise & Waller, 1993). In addition, if variability is a global trait, then the
person who responds consistently across extraversion items should also respond
consistently across items that measure conscientiousness.

Although researchers have demonstrated that response consistency may be linked
to some important individual differences (Berg & Collier, 1953; Goldberg, 1978;
McFarland & Sparks, 1985; Siegrist, 1996), this approach to measuring personality
variability suffers from two important limitations. First, indexes of cross-item variability
can be confounded with mean scores across the same items. Specifically, in order for a
person to get a high mean score on an extraversion scale, he or she must give responses
that are consistently high across all of the items (Paunonen & Jackson, 1985). In
contrast, a person can get a moderate score either by giving consistently moderate
responses or by giving a combination of high and low responses. Therefore, individual
differences in mean levels must be carefully separated from indexes of internal
consistency. The second limitation concerns the potential impact of scale reliability on
response consistency. Specifically, a person may be more likely to give variable
responses across items from an unreliable measure than from a reliable measure,
regardless of how variable his or her personality is. Therefore, researchers have turned to
alternative ways of estimating bottom-up consistency from self-reports.

Another way to compute bottom-up indexes is by comparing a person’s responses

13

across repeated measurements. In other words, if an individual is asked to answer the
same set of questions multiple times, then the extent to which his or her responses change
across occasions can be used as an index of personality variability (Block, 1961;
Donahue et al., 1993; Bid & Diener, 1999; Fleeson, 2004; Kernis & Goldman, 2003;
Penner et al., 1994). For example, researchers interested in the ways someone’s
extraversion changes across social roles might ask the person to describe what his or her
personality is like around friends and around family members. Someone who is variably
extraverted should say that he or she is highly talkative around friends but not at all
talkative around family members, whereas someone who is consistently extraverted
should report being equally talkative in both kinds of situations.

An advantage of this approach over measures of intemal-consistency is that it can
be used to capture variability in specific behaviors, as opposed to variability across
behaviors that are related to one another, but are nonetheless distinct. Furthermore,
repeated-measures indexes may also be less sensitive to scale reliability than indexes of
intemal-consistency because responses are compared within each item, instead of across
items. However, repeated-measures indexes still suffer from some important limitations.
First, estimates of consistency can become confounded with mean scores in ways that
alter the interpretation of findings (Baird et al., 2006; Bid & Diener, 1999; Locke, 2003).
In order for a person to have an average that is extremely high or extremely low across
occasions, he or she must give the same answer each time. Alternatively, people with
moderate scores could get those scores by being consistently moderate or by giving a
combination of high and low responses. Therefore, just as with other bottom-up indexes,

individual differences in mean scores should be separated from variability.

14

A second limitation of the repeated-measures approach is that it may be sensitive
to unique response biases that are not typically found in global questionnaires. For
instance, when people are asked to answer the same questions several times, norms that
govern the exchange of information in everyday conversations may have an influence on
the ways they respond (Schwarz, 1999). Baird & Lucas (2006) found that differences in
self-reported personality across contexts were significantly larger if participants were
asked the same questions multiple times within the same questionnaire than if they were
asked each question only once. This means that a certain amount of inconsistency may
be due to the measurement process, rather than to individual differences in personality.

I Response Styles in Variability Measures

For a long time, researchers have acknowledged that self-report measures are
sensitive to certain biases that may threaten their validity (Cronbach, 1950; Jackson &
Messick, 1958; Rundquist, 1950; see also Schirnmack, Bookenholt, & Reisenzein, 2002).
For instance, in order to make a favorable impression on others, some people may
describe their personalities in ways that exaggerate their positive qualities (Edwards,
1957). To solve the problem of response styles, researchers have developed ways of
aggregating self-reports across repeated measurements (Epstein, 1979; Stone, Shiffman,
& DeVries, 1999). Instead of asking people to describe what they are generally like,
researchers can ask them to describe what they are like at specific moments in time, and
then combine these multiple reports into a single index. Presumably, these aggregated
measures of personality traits are less sensitive to response styles than global reports.

However, new evidence suggests that aggregated measures of personality might

also be sensitive to response styles. Watson and Tellegen (2002) argued that the

15

tendency to blindly agree or disagree with items can create problems when multiple
responses are combined into a single index. Specifically, compared to global ratings of
affect, aggregated measures of momentary affect yield stronger positive correlations
among affect scales of similar valence (e. g. joy and happiness), and weaker negative
correlations among scales of opposite valence (e. g. joy and sadness). Watson and
Tellegen called this the “aggregation paradox” (p.596), and showed that it is directly
related to acquiescent responding. In other words, although aggregated measures are less
sensitive to socially desirable responding than are global reports, they may still be
susceptible to other response biases.

If response styles have an impact on the correlations among mean scores on an
aggregated measure, then they may also have an impact on indexes of intraindividual
variability across repeated measurements. In other words, some people may use response
scales in ways that inflate the variability in their responses, whereas other people may
respond in ways that are more consistent. This would mean that bottom-up indexes of
intraindividual variability are contaminated by stable individual differences in response
style. Of particular concern may be the tendency of some participants to use extreme
numbers on a scale more than others.

Extreme responding

In Berg’s (1953) study of extreme response style, participants were asked to
report their reactions to several neutral words and abstract shapes. For each word or
shape, participants indicated their initial impression of the stimulus on a 4-point scale.
The number of times participants expressed an extreme response (“like much” or “dislike

much”) to neutral words was correlated with the number times they expressed an extreme

l6

response to abstract shapes. Subsequent research has shown that the tendency to give
extreme responses is also quite stable (Austin et al., 1998; Greenleaf, 1992). This is
important for research on intraindividual personality variability because bottom-up
indexes are likely to be inflated by extreme responding.

To date, research has yet to confirm that extreme responding has a direct impact
on measures of personality variability. For instance, Penner et a1. (1994) used standard
deviations in responses on global measures of optimism and self-consciousness to index
extreme responding. These indexes were correlated with each other, but were not
correlated with the standard deviations in momentary affect ratings. Similarly, Diener &
Larsen (1982) included two artifact checks to test whether response extremity could
account for differential consistency in affect ratings. First, respondents provided detailed
descriptions of what each number meant to them. Then, independent raters coded those
descriptions for conservativeness of ratings. Second, respondents indicated where they
thought each scale number would fall on a line. Although these artifact checks were
correlated with one another, they did not account for individual differences in affect
consistency.

Given that so little research has been done on response styles in the measurement
of intraindividual variability, it is somewhat unclear how to study them. Perhaps the best
test of extreme responding would be to compare standard deviations from completely
unrelated scales and see how strongly they relate to one another (Greenleaf, 1992). For
instance, variability in a person’s repeated responses on a bottom-up measure of
personality across contexts should not be associated with variability in his or her

descriptions of daily weather conditions. Similarly, variability in a person’s descriptions

17

of other people should not be associated with variability in his or her moods across days
of the week. Finally, variability in a person’s descriptions of neutral objects should not
be associated with variability in his or her daily reports of personality. However,
correlations among these measures would provide particularly strong evidence that

estimates of intraindividual personality variability are capturing irrelevant information.

18

OVERVIEW OF STUDIES

So far, I have raised a number of important issues related to the study of
intraindividual personality variability. First, individuals clearly differ from one another
in the extent to which their behaviors and moods change across situations, but the reasons
for this individual difference remain unclear. Second, empirical evidence indicates that
intraindividual variability may be linked to a single underlying mechanism. Third, some
theorists have argued that variability in observable behavior results from underlying
conflicts in the self-concept and has implications for interpersonal relationships. Fourth,
intraindividual variability can be measured with a variety of self-report methods,
including top-down and bottom-up assessments. Finally, bottom-up measures may be
sensitive to response styles that inflate the associations among trait—specific indexes of
variability.

The studies presented in this paper were designed with several of these issues in
mind. In order to accurately assess intraindividual personality variability, it is important
to establish the validity of bottom-up indexes. Therefore, I compared a variety of
variability measures with reports of conceptually irrelevant phenomena. If measures of
personality variability are correlated with variability in responses on unrelated measures,
then the interpretation of bottom-up indexes must be reevaluated. For instance, response
style may have an impact on the underlying factor structure of trait-specific measures or
on the associations between bottom-up and top-down indexes. Finally, it is also
important to determine whether bottom-up indexes can be used to test hypotheses about

the relationship between intraindividual variability and psychological well-being.

19

STUDY 1
Bottom-up indexes of intraindividual personality variability are supposed to
capture important individual differences and predict important outcomes related to
psychological health. However, these indexes may instead capture systematic variance
that is due to habitual styles of responding to self-report measures. In this study, I
examined two different bottom-up indexes of personality variability. The first is based
on the internal consistency of an individual’s responses to multiple items within the same
scale. The second is based on the standard deviation of a person’s responses to the same
items across different hypothetical contexts. Along with these measures, I also computed
indexes of response style from self-reports of phenomena that are conceptually unrelated
to personality variability. The first is based on the internal consistency of an individual’s
ratings of inanimate objects. The second is based on the standard deviation of a person’s
ratings of different cartoon characters. If bottom-up indexes are valid, then they should
be positively correlated with top-down measures and should not be correlated with
indexes that are derived from self-reports of unrelated phenomena.
Method
Participants
One hundred forty-nine university students were recruited from undergraduate
psychology courses during the summer semester. A researcher visited each classroom
and asked volunteers to complete a brief, anonymous survey about personality and well-
being. In exchange for completing the survey, participants received course credit.
Materials

Participants completed the 50-item, International Personality Item Pool measure

20

of the Big Five traits (Goldberg, 1999; see Appendix). This measure consists of several
short statements that are each rated on a Likert scale from 1 (“very inaccurate”) to 5
(“very accurate”). Negatively worded items were reverse keyed so that each trait would
be assessed with 10 items of the same valence. Participants then completed the Self-
Concept Clarity scale (Campbell et al., 1996; see Appendix). This measure consists of
12 items, each rated on a Likert scale from 1 (“strongly disagree”) to 5 (“strongly
agree”), that measure top-down perceptions of variability (a = .90). Scale scores were
computed from the average across the items (M = 3.32, SD = .92). Next, participants
were asked to complete a measure of personality across contexts (see Appendix). For
each of 4 different social roles, participants indicated how well 15 different words
describe their behavior on a Likert scale from 1 (“does not describe me”) to 5 (“describes
me very well”). From these ratings, item-specific variability scores can be combined into
bottom-up indexes for each of the Big Five traits and for overall variability (Baird et al.,
2006; Fleeson, 2001). Next, participants completed a measure of satisfaction with
friendships and kinships (see Appendix). This measure was derived from a measure of
perceived autonomy support (Deci & Ryan, 2006) and consists of 18 items (a= .93) that
were each rated on a 5-point scale. Scale scores were computed by averaging together
each participant’s responses on all the items (M = 4.05, SD = .75).

Finally, participants completed the two measures of response style. First, they
were asked to indicate, on a 5-point scale, how well each of ten adjectives describes four
different characters from the television show, The Simpsons (see Appendix). Recent
research found a high level of consensus among people’s descriptions of these cartoon

characters, even among raters with varying familiarity with the show (Kenny & Kenny,

21

2006), suggesting that there are recognizable differences among these characters’
personalities. For example, most people describe Bart Simpson as being more
extraverted than Lisa Simpson, and Marge Simpson as being more conscientious than
Homer Simpson. However, for the purposes of this study, I was interested in the extent
to which each participant rated these characters differently. Because there are real
differences among the Simpsons, each person should report a similar amount of cross-
character variability. Furthermore, any individual difference in variability across
characters should be uncorrelated with variability in self-reported personality.

The second response style measure consisted of a survey of attitudes toward
neutral objects that has been used in previous research to assess positive response biases
(Judge & Bretz, 1993). Participants reported their feelings toward each of 25 objects
using a Likert scale that ranged from 1 (“very dissatisfied”) to 5 (“very satisfied”). With
this measure, an estimate of response style variability can be derived from the
consistency of each participant’s ratings across the objects. Because most of these
objects are relatively benign, each person should report a similar amount of cross-object
variability, and individual differences in variability across objects should not be
associated with cross-character variability in ratings of the Simpsons. Furthermore,
variability in ratings of neutral objects should not correlate with cross-role or cross-item
variability in self-reported personality.

Analytic technique

In this study, I used structural equations modeling (Kline, 2005) to examine the

relationships among trait-specific variability indexes and to test the associations between

latent variability factors and measures of self-consistency and relationship satisfaction.

22

In the first set of analyses, I tested the fit of a measurement model that included a single
variability factor and then added directional paths from the latent variability factors to
indicators of the outcome variables in structural models. Zero-order correlations from
equivalent models are presented in parentheses next to the regression coefficients. In the
second set of analyses, 1 tested the associations between response style variability and the
latent personality variability factors. Then I examined the impact of response style by
comparing the directional coefficients from the first set of analyses to those that emerged
when response style variability was added to the models. If bottom-up indexes are
contaminated by response style, then controlling for variability in ratings of the Simpsons
and neutral objects may change the nature of associations among measures of theoretical
interest. For all models, parameter estimates were computed using maximum likelihood
estimation in AMOS 5 (Arbuckle, 2003), and model fit was assessed from the
comparative fit index (CPI) and root mean square error of approximation (RMSEA). As
a rule of thumb, if a model fits the data well, the CPI will be .95 or higher and the
RMSEA will be .05 or lower.
Results

A major challenge in the analysis of intraindividual variability is to use statistical
indexes that are able to capture the constructs of interest. Several variability indexes are
available (Tellegen, 1988), but some of them are inappropriate because they have the
potential to confound additional personality traits with intraindividual variability (Baird
et al., 2006). In this study, I used linear regression analyses to control for the quadratic
relationship between means and standard deviations during the computation of bottom-up

indexes. Specifically, each bottom-up standard deviation was regressed onto the mean

23

score and the squared mean score from the same items. The residuals from these
equations represent the extent to which a person with a given mean score has a high
amount of variability in his or her responses relative to other individuals with the same
mean. When variability scores are computed in this way, the result is a normally
distributed index with a mean of zero.

The first index of intraindividual variability that I examined was derived from
responses on the IPIP. This measure is typically used to assess trait levels of the Big
Five personality factors and, in this sample, provides reliable estimates of Extraversion
(a =.90), Agreeableness (a=.83), Conscientiousness ((3 =85), Neuroticism (a=.88), and
Openness (a=.82).l However, for the purposes of this study, I computed the standard
deviation of each participant’s responses across the 10 items within each of the five
subscales. Then, to correct for trait levels on these scales, each trait-specific standard
deviation was regressed onto the corresponding mean scores and squared mean scores
from the same scale. The bivariate correlations among these corrected indexes ranged
from .22 to .42 (all ps < .05; mean r = .30).

The second index of personality variability was derived from responses to the
measure of personality across contexts. Again, this measure consists of 15 items that
were administered repeatedly across 4 social roles (Friend, Family, Romantic Partner,
and Student). The cross-role standard deviation of each item was regressed onto the
corresponding cross-role mean and squared mean from the same item. Then, the residual
standard deviations from items that measured the same trait were averaged together to
yield separate variability scores for each of the Big Five factors. The bivariate

correlations among these trait-specific indexes ranged from .35 to .55 (all ps < .05; mean

24

r = .47). In addition, the cross-role variability indexes were significantly (p < .05)
correlated with the cross-item variability indexes for the corresponding trait for
Extraversion (r = .22), Agreeableness (r = .30), Conscientiousness (r = .23), Neuroticism
(r = .24), and Openness (r = .26). In other words, people who responded variably across
extraversion items on the IPIP also reported varying in their levels of extraversion across
roles. When these indexes were examined in a latent factor model (12 = 55.83, df= 34,
CF] = .94; RMSEA = .07), the correlation between Role variability and IPIP variability
was .71 (see Figure 1). Taken at face value, this seems to provide initial support for
theories that link intraindividual variability to a single underlying mechanism.

To test whether variability in behavior is associated with top-down perceptions of
variability, I compared bottom-up indexes from the IPIP and Role measures with scores
on the self-concept clarity scale. The only trait-specific, bottom-up index that was
significantly correlated with self-concept clarity was variability in extraversion across
contexts (r = -.16; p<.05). When the associations among these indexes were tested in the
latent-factor model (12 = 62.30, df= 42, CF! = .95; RMSEA = .06), neither the IPIP
variability factor (r = -.09; n.s.) nor the Role variability factor (r = -.05; n.s.) predicted
self-concept clarity (see Figure 2). In other words, bottom-up indexes of variability were
not associated with top-down perceptions of self-consistency.

To test whether personality variability is associated with relationship quality, I
compared the bottom-up indexes with reports of relationship satisfaction. None of the
trait-specific indexes were correlated with relationship satisfaction. Furthermore, in a
model that included directional paths from the latent variability factors to the outcome

measure (12 = 66.20, df= 42, CF] = .94; RMSEA = .06), neither bottom-up index was

25

associated with relationship satisfaction (see Figure 3). Taken together, these findings
suggest that bottom-up indexes may not capture the constructs of interest.

To test the presence of response—style variance in the bottom-up measures, I
computed variability indexes from ratings of satisfaction with neutral objects and from
ratings of the personality traits of the Simpsons characters. First, the standard deviation
of each participant’s ratings across the 25 neutral objects was regressed onto the mean
and squared mean of those ratings. Second, for each of the 10 items in the Simpsons
measure, the cross-character standard deviation was regressed onto the cross-character
mean and squared mean from the same item. Then these ten, corrected standard
deviations were averaged together. These two indexes capture sources response
variability that are conceptually distinct from one another and from intraindividual
personality variability. However, variability in responses on the neutral objects measure
was significantly correlated with variability in ratings of Simpsons characters (r = .47, p<
.05), suggesting that these measures capture the same tendency to give variable
responses.

Next, I tested the associations between response styles and self-reports of
personality variability and relationship satisfaction. Bivariate correlations among the
response style measures, cross-item and cross-role variability, self-concept clarity, and
relationship satisfaction are presented in Table 1. Both of the bottom-up indexes were
significantly correlated with variability across the Simpsons and with variability across
objects. In other words, participants whose self-reported personality varied across items
or across roles tended to rate the personalities of Simpsons characters, as well as their

own satisfaction with neutral objects, in ways that were more variable than participants

26

who gave consistent self-reports of personality. Finally, bottom-up variability was not
correlated with self-concept clarity or with relationship satisfaction. However, self-
concept clarity was positively correlated with relationship satisfaction (r = .45, p< .05).

In order to assess the impact of response style on the psychometric properties of
bottom-up indexes, the Simpsons and neutral objects indexes were added to the latent
factor model (Figure 4). In this model (1’ = 77.63, df= 52, CF] = .95; RMSEA = .06),
response style variability explained a significant amount of variance in both of the latent
variability factors. Perhaps more importantly, as can be seen in the models depicted in
Figure 5 (1’ = 82.76, df= 60, CF1= .95; RMSEA = .05) and Figure 6 (1’ = 86.54, df=
60, CF! = .94; RMSEA = .06), accounting for response style did not change the patterns
of associations between the bottom-up indexes and the measures of self-concept clarity
and relationship satisfaction. In other words, intraindividual variability was not
associated with top-down consistency or with psychological well-being.

An alternative explanation for these findings is that the top-down measure is
really capturing something entirely different from variability in behavior. In fact, many
of the items in the self-concept clarity scale assess people’s overall evaluations of their
identities or the extent to which they feel that they know themselves (e.g., items 3, 7, &
11), instead of their perceptions of how much their personality traits actually vary. In
other words, this scale may be an inappropriate criterion for testing the validity of
bottom-up variability, and the lack of convergence with the bottom-up indexes should not
come as a surprise. A more appropriate test would be to use items that are directly
related to bottom-up variability. Fortunately, some of the self-concept clarity items are

more face-valid that the others, including “On one day I might have one opinion of

27

myself and on another day I might have a different opinion” (item 2), “My beliefs about
myself seem to change very frequently” (item 8), and “If I were asked to describe my
personality, my description might end up being different from one day to another day”
(item 9). Therefore, I examined how strongly each item correlated with the two
aggregated bottom-up indexes. Item-specific correlations with IPIP variability ranged
from -.21 (item 11) to .06 (item 3). However, the correlations for items two (r = .00),
eight (r = -. 10), and nine (r = -.01) were non-significant and near-zero. Instead, only
items 11 (“In general, I have a clear sense of who I am and what I am.”) and 12 (“It is
often hard for me to make up my mind about things because I don’t really know what I
want”) were significantly correlated with IPIP variability. Item-specific correlations
with cross-role variability ranged from -. 12 (item 8) to .06 (item 10). None of these
correlations were significant, and even the most face valid items were not associated with
bottom-up variability.
Discussion

In Study 1, trait-specific indexes of cross-role variability were significantly
correlated with one another, indicating that people who reported varying in extraversion
across roles also reported varying in the other Big Five traits across those same roles.
Similarly, trait-specific indexes of cross-item variability were significantly correlated
with one another, indicating that people who responded inconsistently across items that
measured extraversion also responded inconsistently across items that measured other
Big Five traits. Furthermore, latent factors of cross-role variability and cross-item
variability were strongly correlated with one another. However, these bottom-up indexes

of variability were not associated with top-down judgments of self-concept clarity or

28

with ratings of relationship satisfaction. Instead, bottom-up variability was associated
with a tendency to describe neutral objects and cartoon characters inconsistently. When
measures of response style were added to the analyses, there were no changes the overall
fit of the models and no differences in the relationships between the bottom-up indexes
and the substantive measures of subjective variability and well-being.

Clearly, bottom-up indexes of variability that are derived from global self-reports
of personality are contaminated by response styles, and this has an impact on the
psychometric properties of these measures. However, it is possible that this phenomenon
is limited to survey methods of assessing variability. In other words, something about
the questionnaire itself may have elicited greater variability in people’s responses (Baird
& Lucas, 2006; Schwarz, 1999). Perhaps because all of the measures were administered
at the same time, participants adopted temporary styles of responding in order to
complete the survey. If so, then response style may be less of a problem when bottom-up
indexes are computed from repeated measurements that are administered at different
points in time. Therefore, in Study 2, I examined the associations between global

measures of variability and an index that is derived from daily reports of personality.

29

STUDY 2

In recent years, researchers interested in intraindividual variability have
developed ways of assessing changes in personality that occur across actual experiences
and situations (Eid & Diener, 1999; Fleeson, 2001; Kernis, 2005). Although these
methods provide an ideal opportunity to test theories that link behavioral inconsistency to
broad individual differences, only a few studies have directly compared global indexes of
personality variability with changes in ongoing behavior. In research that utilized an
experience-sampling design, Baird et al. (2006) tested the validity of several global
indexes that are similar to the ones used in Study 1. They found that a top—down measure
of self-consistency was not associated with variability across reports of moment-to-
moment behavior. In other words, people who reported having inconsistent feelings and
beliefs were no more likely to behave inconsistently than people who held consistent
perceptions of themselves. In contrast, a bottom-up index of cross-role variability
predicted variability in behavior across moments, as well as changes in personality over
time and self-other agreement.

While these findings may provide initial support for the validity of bottom-up
indexes, they may also indicate that daily-diary measures are sensitive to the same
response style that impacts global measures. In fact, although the global and momentary
indexes were correlated in the Baird et al. study, both were only modestly correlated with
the top-down measure and neither was correlated with subjective well-being. In other
words, people who give variable responses may do so consistently, regardless of item
content, and this may lead to spurious correlations between global and momentary

measures of variability. Therefore, in Study 2, I examined associations among indexes of

30

variability from survey and daily-diary methods. If global measures of intraindividual
variability are valid, then they should predict the extent to which a person’s behavior
varies from day to day, above and beyond any effect of shared response style.

In this study, I assessed response style variability with measures of three different
phenomena that are unrelated to personality variability. First, as in Study 1, participants
were asked to rate their satisfaction with each of several neutral objects. This measure
provides an estimate of how consistently a person responds to different items within a
scale. Second, each day for two weeks, participants were asked to answer questions
about current weather conditions. Although there are objective changes in the weather,
individual differences in the variability of weather reports should be relatively small and
should not be correlated with variability in ratings of neutral objects. Finally, each
participant was asked to describe the ways a close friend behaves in different situations.
Unlike on the Simpsons measure, where everyone was asked to rate the same set of
targets and each target was described once, on this measure, participants were free to
choose the person that they wanted to rate and then described that person’s behavior
multiple times. This measure is also similar to the self-report of personality across
contexts, except that friends were rated within contexts that were different than those
within which participants rated themselves. Although variable people may have friends
that are also variable, if variability in the descriptions of a friend correlate with
variability in satisfaction with neutral objects or in reports of the weather, then this
measure may capture response style.

Presumably, variability in reports of satisfaction with neutral objects, variability

in descriptions of a friend’s behavior across contexts, and variability in reports of the

31

weather should not be associated with variability in one’s own personality across items,
in different contexts, or over time. On the other hand, if these measures are correlated
with one another, then this would suggest that bottom—up indexes of personality
variability capture a response style.

Finally, the results of Study I raised some doubts about the appropriateness of
self-concept clarity as a criterion for the bottom-up indexes. It is possible that this
measure, and others like it, are better suited for studying variability in evaluative
components of personality, such as affect and self-esteem. In other words, because
measures of self-concept clarity and self-esteem stability focus explicitly on consistency
in people’s beliefs and opinions about themselves, they should be more strongly
associated with variability in relevant traits than with variability in behavior. In other
words, if top-down indexes are capturing constructs that are different from those that are
measured with bottom-up indexes, then they should be differentially associated with
these alternative measures. Therefore, in Study 2, I examined compared measures of
day-to-day variation in positive and negative affect and self-esteem with top-down and
bottom-up indexes of variability in behavior.

Method
Participants

Two-hundred fifty-two undergraduate students signed up to participate in a three-
week study of personality and emotions. The entire study took place on the internet and
correspondence between researchers and participants was conducted via electronic mail.
In exchange for completing the study, participants were given course credit. Individuals

who dropped out before the end of the study received partial credit.

32

Materials

Global reports. Global personality variability was assessed using with the same
two bottom-up measures that were used in Study 1. Responses on the 50-item IPIP and
the 15-item x 4-role contextual measures were used to compute indicators of trait-
specific variability . Subjective variability was assessed using the Self-Concept Clarity
scale (a= .89) and the Stability of Self Scale (Rosenberg, 1989), which contains 5 items
that capture consistency in one’s opinions and ideas about himself or herself (a' = .86).
To compute top-down indicators from these measures, responses on each item were
converted into z-scores and then averaged together.

Participants also completed several global measures of psychological well-being.
The Intensity and Time Affect Survey (ITAS; Diener, Smith, & Fujita, 1995) was used to
assess trait levels of positive affect (a = .82) and negative affect (a = .90). On this
measure, participants indicated, on a 5-point Likert scale from 1 (“very slightly or not at
all”) to 5 (“extremely”), the extent to which they experience each of several emotions.
Global self-esteem was measured using ten items from Rosenberg’s (1989) measure (a =
.91), which were rated on a 5-point Likert scale ranging from 1 (“does not describe me at
all”) to 5 (“describes me very well”). Participants also completed Spielberger’s (1983)
20-item Trait Anxiety Scale (a = .92), the lO-item short form (a = .81) of the Center for
Epidemiological Studies Depression scale (Cole, Rabin, Smith, & Kaufman, 2004), and a
10-item measure of dispositional optimism (a = .82).

Finally, two of the response style measures were included among the global
assessments. Variability in satisfaction with neutral objects was assessed with the same

measure that was used in Study 1. A second index was derived from participants’ ratings

33

of a close friend’s personality across three contexts (“ in general”, “at work”, & “with
strangers”). This measure is similar to the contextual self-report, except that friends were
rated on five adjectives and across three contexts that were all different from those on the
self-report (see Appendix). Presumably, the extent to which a participant’s friend is
different around strangers than at work should not predict how much the participant
varies from situations with romantic partners or family members.

Daily reports. To assess personality variability across ongoing experiences,
participants were asked to give daily reports of their behavior and affect (see
Appendices). On this measure, each of the Big Five traits was assessed with two
adjectives that also appeared on global self-report of variability across contexts.
Respondents were also asked to report their self-esteem each day by indicating how well
the statement “Today I feel that I am a person of worth, at least on an equal basis with
others” described them. Finally, to measure response style, participants were asked to
report how sunny, windy, and cold it was each day, on a 1 (“not at all”) to 5 (“very”)
scale. Because all the participants were living in the same city at the time of the study,
and because everyone completed the daily reports during the same two-week period, they
were exposed to the same weather conditions. Therefore, any individual differences in
the variability of weather reports are assumed to reflect response style.

Procedure

To sign up for the study, participants logged on to the psychology department’s
website and completed an online consent form. During this initial session, participants
completed the 50-item IPIP, the 60-item measure of role variability, the IT AS, the neutral

objects questionnaire, and the self-concept clarity measure. Then, for the remainder of

34

the study, participants were instructed to log on to a password-protected website, where
they could complete the daily reports and a follow-up questionnaire.

Participants completed the daily reports each evening during the same two-week
period. However, participants were informed that the website could only be accessed
between 5:00pm and midnight each day, and that they would not allowed to make up for
any days that were missed. On average, the daily report sessions lasted about two
minutes. Participants who participated in the daily-diary portion of the study (N = 239)
completed an average of 10.68 daily reports, and roughly 90% completed at least seven.

Finally, at the end of the two weeks, 229 participants were asked to complete a
follow up questionnaire on the same website. That questionnaire included the 50-item
IPIP, the stability of self measure, and the global measures of self-esteem, optimism,
depression, and anxiety.

Analytic technique

The analyses for this study were conducted using modeling strategies that were
similar to those in Study 1. First, trait-specific indicators of variability from the three
self-report measures were examined in a measurement model that included latent factors
of IPIP variability, Role variability, and Daily variability. Then, in a series of structural
models, directional paths from the latent variability factors to the outcome variables were
added. Zero-order correlations from equivalent models are presented in parentheses next
to the regression coefficients. Finally, the three measures of response style variability
were added to these models in order to examine the association between bottom-up
variability and response style, as well as the impact on the covariances and directional

coefficients from the first set of analyses.

35

Results

Bottom-up indexes of personality variability were computed from the global
measures using the regression procedure described in Study 1. First, standard deviations
across the 10 items in each of the IPIP subscales were regressed onto the means and
squared means from the same scales. The residuals from these equations represent the
extent that a person’s responses varied relative to other people with the same mean
scores. Next, cross-role standard deviations for each of the 15 items in the contextual
measure were regressed onto the cross-role means and squared means from the same
items. Then, the residual standard deviations from items that measured the same trait
were averaged together to create five, trait-specific indexes. An index of daily behavior
variability was computed for each participant who completed at least seven daily reports
(N = 217). The cross-day standard deviations on each of the 10 items were regressed
onto the cross.day means and squared means from the same item. Then, residuals from
items that measured the same trait were combined to create trait-specific indexes of daily
variability for each of the Big Five. Variability scores for daily positive and negative
affect and daily self-esteem were computed in the same way as the daily behavior
variability indexes.

The relationships among indexes from the three bottom-up measures were
examined in a latent-factor model (Figure 7). In this model (12 = 149.20, df= 87, CF] =
.96; RMSEA = .05), variability factors from the two global measures were significantly
correlated with one another, and both were moderately correlated with daily variability.
This means that people who gave variable responses across IPIP items also rated

themselves differently across contexts, and reported behaving in variable ways across

36

ongoing experiences. In other words, each measure appears to capture a general trait of
personality variability. However, as can be seen in Figure 8, these measures were
differentially associated with the top-down measures (12 = 177.82, df= 113, CFI = .96;
RMSEA = .05). Specifically, variability on the Role measure was associated with lower
self—concept consistency, whereas variability on the IPIP measure was associated with
higher self-concept consistency. Variability in the daily reports was not associated with
scores on the top-down measures.

Finally, the association between intraindividual variability and psychological
well-being was examined in the model found in Figure 9 (1" = 321.40, df= 183, CFI =
.94; RMSEA = .05). Specifically, each of the three latent variability traits was regressed
onto a well-being factor that included positive and negative affect, self-esteem,
depression, optimism, and anxiety (Table 2 contains descriptive statistics for well-being
measures). In this model, role variability was associated with significantly lower well-
being, whereas IPIP variability was associated with significantly higher well-being.
Daily variability did not predict psychological well-being. Taken together, these findings
provide contradictory evidence for the implications of intraindividual variability.
Although the three latent factors are positively correlated with one another, each is
differentially associated with subjective variability and with psychological well-being.
Therefore, it is important to determine whether the bottom-up indexes are capturing valid
individual differences or response styles.

Indexes of response style variability were computed from responses on the
measures of phenomena that are conceptually unrelated to personality variability. First,

the standard deviations in participants’ responses on the neutral objects measure were

37

regressed onto the mean and squared mean of those ratings. Second, the cross-role
standard deviation on each item in the friend measure was regressed onto the mean and
squared mean from the same item. Finally, the standard deviations in reports of daily
weather conditions were regressed onto the mean and squared mean of each item.
Bivariate correlations among these three indexes, as well as their associations with
personality variability, can be found in Table 3. All of response style indicators were
significantly correlated with the three measures of bottom-up variability. However, the
correlation between variability in weather reports and variability in friend ratings was not
significant. It is possible that the relatively weak relationships between the weather
index and the other response style indexes are due to objective stability in weather
conditions during the study.2 In other words, because the weather did not change a great
deal, this may not be an ideal way to measure response styles. On the other hand, the fact
that individual differences in weather variability were significantly correlated with the
bottom-up indexes may be particularly strong evidence of a response style. Removing
this index from the analyses did not alter the overall pattern of results. Therefore, in
subsequent analyses, the relationships among the three response style indicators were
modeled as a latent trait.

To examine the impact of response style on the psychometric properties of
variability measures, I added the three response style measures to the models presented
above. In the model depicted in Figure 10 (1’ = 216.56, df= 132, CFI = .95; RMSEA =
.05), a significant amount of variance in each of the three bottom-up indexes was
explained by the latent response style factor. This suggests that the correlations among

variability indexes that were found in previous analyses may overestimate the true

38

relationship between global variability and daily variability. In fact, in a model depicted
in Figure 11 (1’ = 212.71, df= 129, CFI = .95; RMSEA = .05), neither global index
predicted daily variability when response style was controlled. In other words, people
who responded variably on the IPIP, or who described themselves variably across roles,
also tended to describe unrelated phenomena in variable ways, and this explains why they
also reported differences in their behavior across ongoing experiences.

So how might response style impact the associations between intraindividual
variability and substantive outcomes like self-consistency and psychological well-being?
This question can be answered by including the response style variables in the models
that were tested above.3 Recall that in the previous models both indexes were
significantly correlated with these outcomes, but in opposite directions. As can be seen
in Figure 12 (1’ = 144.22, df= 84, CF] = .95; RMSEA = .05) and Figure 13 (1’ = 276.56,
df= 146, GP] = .93; RMSEA = .06), neither of the global indexes predicted self-
consistency or psychological well-being when response style was partialled from the
equations.

A final issue is to determine the meaning of the top-down measures. Specifically,
if self-concept clarity and self-esteem stability are unrelated to variability in behavior,
then what might explain individual differences on these measures? Perhaps these
measures capture perceptions of variability in personality traits other than the big five.
Therefore, I examined the correlations between each of the top-down indexes and daily
variability in affect and self-esteem. Scores on the self-concept clarity scale were
significantly correlated with variability in positive affect (r = -.29, p<.05) but were

unrelated to variability in negative affect (r = —.04, ns) and self-esteem (r = -.10, ns).

39

Similarly, scores on the self-esteem stability scale were significantly correlated with
variability in positive affect (r = -.29, p<.05) but were unrelated to variability in negative
affect (r = -.l 1, ns) and self-esteem (r = -.13, ns). In other words, people who
experienced consistent levels of positive affect were more likely to report having stable
beliefs and opinions about themselves than people who experienced unstable levels of
positive affect.

In contrast, each of the bottom-up indexes of variability in behavior were
associated with daily variability in affect and self-esteem. Specifically, cross-role
variability was significantly correlated with variability in positive affect (r = .41, p<.05),
negative affect (r = .30, p<.05), and self-esteem (r = .30, p<.05). IPIP variability was
significantly correlated with variability in positive affect (r = .26, p<.05), negative affect
(r = .28, p<.05), and self-esteem (r = .27, p<.05). Finally, daily variability was
significantly correlated with variability in positive affect (r = .65, p<.05), negative affect
(r = .55, p<.05), and self-esteem (r = .55, p<.05). In other words, people who behaved in
variable ways were also likely to experience variable levels of affect and have unstable
levels of self-esteem. This suggests that bottom-up indexes of behavior variability are
capturing meaningful individual differences, but that they are unrelated to top-down
perceptions of self-consistency.

Discussion

In Study 2, people who described themselves differently across contexts also
tended to give inconsistent responses on the IPIP. Although the correlation was
somewhat weaker than in Study 1, bottom-up indexes of variability from these two

measures appear to capture the same underlying construct. Furthermore, replicating

4o

findings from Baird et al. (2006), global measures of personality variability were
significantly correlated with changes in reports of ongoing behavior in the daily diary
assessment. In other words, people who said that they tend to behave differently in
different contexts also reported differences in their ongoing behavior across days.

On the surface, these findings seem to support the validity of bottom-up methods
of measuring personality variability. However, each of these measures was also
correlated with variability in reports of irrelevant phenomena. That is, people who gave
variable reports of their own personalities also rated neutral objects in variable ways,
described their friends differently across contexts, and gave inconsistent reports of the
weather. When this response style was statistically controlled, the two global measures
no longer predicted variability in daily reports and neither index was associated with

substantive measures of self-consistency and well-being.

41

GENERAL DISCUSSION
The studies reported in this dissertation were designed to address four issues.
First, personality variability may be a global trait that permeates multiple aspects of
behavior. Therefore, I examined the relationships among trait-specific measures of
variability to see whether these indexes could be explained by a single underlying factor.
Second, the extent to which a person’s personality varies across situations may be
associated with his or her perceptions of variability. Therefore, I examined the
relationships between bottom-up indexes and top-down measures of self-concept clarity
and self-esteem stability. Third, intraindividual variability may have implications for
outcomes related to psychosocial adjustment. Therefore, I tested the associations among
various measures of personality variability and self-reports of relationship satisfaction
and psychological health. Finally, bottom-up indexes of variability may be sensitive to
unique response biases that arise when respondents are asked provide repeated self-
reports of personality. Therefore, I compared scores on various bottom-up measures with
variability in ratings of several unrelated phenomena.
Summary of Findings
In both studies, bottom-up indexes of trait-specific variability were significantly
correlated with one another. In other words, people who reported varying in their levels
of one trait also reported a great deal of variability in other traits. Furthermore, latent
traits of bottom-up variability from two global measures were strongly related to one
another. In Study 2, both of these global measures were significantly correlated with
variability in daily reports of behavior. Taken at face value, these findings support

theories that link intraindividual variability to a single underlying mechanism.

42

However, on the whole, bottom-up indexes were not associated with top-down
ratings of self-consistency. In Study 1, neither of the global measures was correlated
with self-concept clarity. In Study 2, bottom-up indexes from global measures were
correlated with self-concept clarity and self-consistency, but in opposite directions, and
daily variability was uncorrelated with the top—down measures. Finally, bottom-up
measures were not reliably associated with well-being. In Study 1, neither of the global
indexes was correlated with relationship satisfaction, whereas in Study 2, IPIP variability
and Role variability were both correlated with psychological health, but in opposite
directions. In other words, although trait-specific variability appears to be linked to a
single underlying factor, this factor is not associated with inconsistency in the self-
concept or with lower psychological health.

If bottom-up indexes are not correlated with measures of theoretical interest, then
why are they so strongly related to one another? In Study 1, people who gave variable
responses on self-report measures of personality also reported inconsistent levels of
satisfaction with neutral objects and rated the personalities of cartoon characters in
variable ways. In Study 2, people who described their own personalities in variable ways
also described their friends’ personalities in variable ways and gave variable reports of
objective weather conditions. When the correlations with these irrelevant indexes were
controlled, the relationships among bottom-up measures of personality variability were
substantially reduced.

Implications
A number of theorists have argued that intraindividual variability is an important

component of personality and some have even developed complex models to account for

43

the processes underlying it. Many of these models make clear predictions about
relationships among multiple indicators of consistency. For instance, theories that link
intraindividual variability to identity conflicts or to unstable self-concepts predict that
multiple aspects of a person’s behavior should be similarly unstable. Furthermore, these
theories predict that the internal conflicts that lead people to behave inconsistently can be
assessed by asking variable people direct questions about their self-conceptions. Finally,
conflict theories also predict that personality variability should be associated with lower
levels of psychological well-being.

In general, the results of these studies indicate that variability is a global trait, but
this trait does not appear to be linked to underlying conflicts. Bottom-up indexes were
not associated with self-concept clarity or with self-esteem stability, nor did they predict
outcomes related to psychological health. In contrast, top-down measures were
significantly correlated with both relationship satisfaction and well-being. A simple
interpretation of these findings might be that variability in the self-concept is not related
to variability in behavior, and that top-down measures tap into processes that are
different than those captured by bottom-up measures. However, an alternative
explanation is that the single factor that links bottom-up indexes is a response style,
rather than a meaningful psychological process.

If bottom-up indexes are contaminated by response styles, then these methods of
assessing personality variability may be inappropriate for tests of substantive hypotheses.
Therefore, in order to understand the theoretical implications of these findings, it is
important to consider what this response style represents. For instance, there may be

meaningful individual differences in styles of information processing that lead to

44

differences in response consistency (Robinson, Goetz, Wilkowski, & Hoffman, 2006). In
other words, the tendency to give variable responses may be related to rigidity or
irnpulsivity in decision making. If so, then researchers may be able to use bottom-up
indexes as indirect or implicit measures of personality (see also McGee, 1962).

However, response consistency may also be tied to methodological factors that inflate the
differences in a person’s responses across items or on different measurement occasions.
For instance, certain conversational norms of information sharing may lead people to
respond in variable ways in order to avoid giving redundant information (Baird & Lucas,
2006). Therefore, future research must examine the factors that can account for
individual differences in response consistency.

In their classic review, F iske and Rice (1955) described different types of
variability that may be important for psychological research. The first type, called
spontaneous variability, occurs when a person behaves differently over time because of
organic processes within the individual, such as fluctuations in hormone levels, blood
pressure, or brain chemistry. The second type of variability, referred to as systematic
variability, occurs when a person responds differently to the same stimulus over time
because of prior repetitions of exposure-and-reaction to that stimulus. Finally, the third
type of variability occurs because of changes in the stimulus itself or in the environment
within which the stimulus is encountered. Although Fiske and Rice did not give this type
of variability a name, it might be helpful to call it reactive variability because it reflects
the influence of external factors on behavior. This framework provides a good
foundation for variability research because it clearly differentiates among various aspects

of personality variability that are theoretically meaningfirl. However, in practice, it may

45

be difficult to separate one type of variability from another.

To illustrate the differences among these types of variability, consider an example
of a hypothetical participant fi'om Study 2. On Monday, a person reports a high level of
talkativeness and a low level of irritability, but then on Thursday he or she reports a low
level of talkativeness and a high level of irritability. Why did the person’s responses
change? If the person was healthy on Monday, but then was sick on Thursday, the
change in physical health might have contributed to differences in his or her behavior.
This would be an example of spontaneous variability. If the person interacted with a
friend on Monday and a family member on Thursday, then the differences in behavior
could be attributed to differences in the social context. This would be an example of
reactive variability. Finally, perhaps the person reported behaving differently on
Thursday simply because he or she was asked the same set of questions on Monday. In
other words, the process of responding to the same questions over and over again may
have led the person to give different answers over time. This would be an example of
systematic variability. As this example demonstrates, intraindividual variability may
reflect a variety of processes, some of which are more relevant to personality theories
than others.

For the most part, models of personality variability are based on the assumption
that changes in self-reports reflect either spontaneous or reactive mechanisms. On one
hand, psychophysiological models of temperament (Geen, 1997), which link behavior to
biological processes, can be used to explain instances of spontaneous variability. For
example, Larsen (1985) argued that circadian rhythms may contribute to changes in a

person’s personality over the course of a day. On the other hand, social-cognitive models

46

of personality, which link behavior to characteristic thoughts and beliefs about the
environment, can be used to explain reactive variability. For example, in Mischel and
Shoda’s (1995) CAPS model, variability in behavior results from idiosyncratic ways of
processing information about social situations. However, because both sets of theories
are often tested using self-report methods, it is impossible to rule out the possibility that
changes in reports of personality are the result of systematic variability.

Because of the increasing interest in intraindividual variability, response styles
may have practical implications for other areas of research as well. For instance,
researchers interested in self-esteem stability often ask people to give repeated reports of
their self-esteem in order to see how much each person changes (Kernis, 2005). If
bottom-up indexes of self-esteem stability are affected by response styles, then these
measures may not correlate with top-down judgements. In addition, researchers
interested in individual differences in the complexity of self-reported affect should be
aware of the potential for response styles to impact indexes of emotional variability
(Tugade, Fredrickson, & Barrett, 2004). In particular, people who tend to give variable
answers may be likely to report large differences in their emotions, and this may lead to
inaccurate estimates of variability.

Limitations and Future Directions
An important limitation of these studies is that they only include self-report
methods of assessing personality variability. Although a great deal of previous research
on personality variability has relied exclusively on similar measures, these findings may
not generalize to alternative methods. For instance, it is possible that a different structure

of trait-specific variability would emerge if personality was measured with observational

47

methods. Therefore, more studies should be done to compare variability in self-reports
with systematic observations of behavioral variability (Schneiderman, 1980). This would
help rule out response styles as an explanation for associations among indexes of
variability in different traits.

In addition to observational methods, it is also important to compare self-reports
of variability with variability in informant ratings of behavior. For instance, if a person
reports behaving differently across roles, then people who know that person in those
contexts should describe him or her in different ways. This may also allow for a more
direct test of relationship outcomes. Perhaps disagreement between relationship partners
about what a person’s personality is like produces feelings of variability and lower
relationship satisfaction without leading to variable behavior.

Finally, these studies are also limited to self-reports that are made over relatively
short periods of time. In other words, it is important to investigate how response styles
might impact measures of variability that are based on longer measurement intervals.
For instance, it is possible that the response styles that impact short-term variability
estimates can help explain certain effects of instrumentation that have been documented
in longitudinal panel studies. Therefore, future research should incorporate measures of
short-term variability with measures of developmental changes in personality that occur

across the life span.

48

CONCLUSION
Theories that can account for intraindividual variability hold great promise for
personality psychology (Fleeson & J olley, 2006), but in order to capitalize on this
potential, researchers must continue to refine their methods and consider alternative
explanations for their findings. In other words, individual differences in response
consistency are likely to reflect a combination of factors, and some of these mechanisms
may be irrelevant to personality. Therefore, perhaps the greatest challenge is to separate

meaningful response variability from measurement unreliability.

49

REFERENCES

Arbuckle, IL (2003). Amos 5 [Computer software]. Chicago: SmallWaters.

Aron, A., Aron, E.N., Tudor, M., & Nelson, G. (1991). Close relationships as including
other in the self. Journal of Personality and Social Psychology, 60, 241-253.

Austin, E.J., Deary, I.J., Gibson, G.J., McGregor, M.J., & Dent, J .B. (1998). Individual
response spread in self-report scales: Personality correlations and consequences.
Personality and Individual Differences, 24, 421-438.

Baird, B.M., Le, K., & Lucas, RE. (2006). On the nature of intraindividual personality
variability: Reliability, validity, and associations with well-being. Journal of
Personality and Social Psychology, 90.

Baird, B.M., & Lucas, RE. (2006). “...and how about now?” Inflation of response
variability in contextual reports. Unpublished manuscript.

Baumeister, RF. (1991). On the stability of variability: Retest reliability of metatraits.
Personality and Social Psychologl Bulletin, 1 7, 633-639.

Baumeister, R.F. & Tice, D.M. (1988). Metatraits. Journal of Personality, 56, 571-598.

Bern, D.J., & Allen, A. (1974). On predicting some of the people some of the time: The
search for cross-situational consistencies in behavior. Psychological Review, 81,
506-520.

Bem, S.L., & Lewis, SA. (1975). Sex role adaptability: One consequence of
psychological androgyny. Journal of Personality and Social Psychology, 31,
634-643.

Berg, LA. (1953). The reliability of extreme position response sets in two tests. The
Journal of Psychology, 36, 3-9.

Berg, I.A., & Collier, J .S. (1953). Personality and group differences in extreme response
sets. Educational and Psychological Measurement, 13, 164-169.

Biesanz, J .C., & West, S.G. (2000). Personality coherence: Moderating self-other
agreement and profile consensus. Journal of Personality and Social Psychology,
79, 425-437.

Block, J. (1961). Ego identity, role variability, and adjustment. Journal of Consulting
Psychology, 25, 392-397.

50

Campbell, J .D., Assanand, S., & DiPaula, A. (2003). The structure of the self-concept
and its relation to psychological adjustment. Journal of Personality, 71, 115-140.

Campbell, J.D., Trapnell, P.D., Heine, S.J., Katz, I.M., Lavallee, L.F., & Lehman, DR.
(1996). Self-concept clarity: Measurement, personality correlates, and cultural
boundaries. Journal of Personality and Social Psychology, 70, 141-156.

Cervone (1999). Bottom-up explanation in personality psychology: The case of cross-
situational coherence. In D. Cervone & Y. Shoda (Eds.), The coherence of
personality: Social-cognitive bases of consistency, variability, and organization
(pp. 303-341). New York: Guilford.

Cervone, D. (2004). The architecture of personality. Psychological Review, 111, 183-
204.

Cloninger, CR. (2003). Completing the psychobiological architecture of human
personality development: Temperament, character, and coherence. In U.M
Staudinger & U. Lindenberger (Eds.), Understanding human development:
Dialogues with lifespan psychology (pp. 159-181). Norwell, MA: Kluwer.

Cole, J .C., Rabin, A.S., Smith, T.L., & Kaufman, AS. (2004). Development and
validation of a Rasch—derived CES-D short form. Psychological Assessment, 1 6,
360-372.

Colvin, CR. (1993). “Judgable” people: Personality, behavior, and competing
explanations. Journal of Personality and Social Psychology, 64, 861-873.

Cronbach, LJ. (1950). Further evidence on response sets and test design. Educational
and Psychological Measurement, 10, 3-31.

Deci, E.L., & Ryan, RM. (2006). The health care climate questionnaire. Retrieved from
the Self-Determination Theory web site: http://www.psych.rochester.edu/SDT/
measures/auton_hlth.html.

Diener, E., & Larsen, R]. (1984). Temporal stability and cross-situational consistency of
affective, behavioral, and cognitive responses. Journal of Personality and Social
Psychology, 47, 871-883.

Diener, E., Smith, H., & Fujita, F. (1995). The personality structure of affect. Journal of
Personality and Social Psychology, 69, 130-141.

Donahue, B.M., Robins, R.W., Roberts, B.W., & John, OF. (1993). The divided self:
Concurrent and longitudinal effects of psychological adjustment and social roles
on self-concept differentiation. Journal of Personality and Social Psychology,
64, 834-846.

51

Edwards, AL. (1957). The social desirability variable in personality assessment and
research. New York: Dryden.

Eid, M., & Diener, E. (1999). Intraindividual variability in affect: Reliability, validity,
and personality correlates. Journal of Personality and Social Psychology, 76,
662-676.

Epstein, S. (1979). The stability of behavior I: On predicting most of the people much of
the time. Journal of Personality and Social Psychology, 3 7, 1097-1126.

Fiske, D.W., & Rice, L. (1955). Intra-individual response variability. Psychological
Bulletin, 52, 217-250.

Fleeson, W. (2001). Toward a structure- and process-integrated view of personality:
Traits as density distributions of states. Journal of Personality and Social
Psychology, 80, 1011-1027.

Fleeson, W. (2004). Moving personality beyond the person-situation debate: The
challenge and the opportunity of within-person variability. Current Directions in
Psychological Science, 13, 83-87.

Fleeson, W., & Jolley, S. (2006). The challenge and the opportunity of intraindividual
variability: A theory of human flexibility. In D. Mroczek & T.D. Little (Eds.),
Handbook of personality development (pp. 41-59). Mahwah, NJ: Erlbaum.

Geen, R.G. (1997). Psychophysiological approaches to personality. In R. Hogan, J.
Johnson, & S. Briggs (Eds.), Handbook of personality psychology (pp.387-414).
San Diego, CA: Academic Press.

Goldberg, L.R. (1978). The reliability of reliability: The generality and correlates of
intra-individual consistency in responses to structured personality inventories.
Applied Psychological Measurement, 2, 269-291.

Goldberg, L.R. (1999). A broad-bandwidth, public domain, personality inventory
measuring the lower-level facets of several five-factor models. In I. Mervielde, I.
Deary, F. DeFruyt, & F. Ostendorf (Eds.), Personality Psychology in Europe
(V 01. 7, pp. 7-28). Tilburg, the Netherlands: Tilburg University Press.

Greenleaf, EA. (1992). Measuring extreme response style. The Public Opinion
Quarterly, 56, 328-351.

Jackson, D.N., & Messick, S. (1958). Content and style in personality assessment.
Psychological Bulletin, 55, 243-252.

Jourard, SM. (1963). Personality adjustment: An approach through the study of healthy
personality. New York: Macmillan.

52

Judge, T.A., & Bretz, RD. (1993). Report on an alternative measure of affective
disposition. Educational and Psychological Measurement, 53, 1095-1104.

Kenny, D.A., & Kenny, D.T. (2006). The personalities of the Simpsons. In A. Brown &
C. Logan (Eds.), The psychology of the Simpsons (pp.187-200). Dallas, TX:
BenBella Books.

Kemis, M.H. (2005). Measuring self-esteem in context: The importance of stability of
self-esteem in psychological functioning. Journal of Personality, 73, 1569-1605.

Kernis, M.H., & Goldman, BM. (2003). Stability and variability in self-concept and
self-esteem. In M.R. Leary & J .P. Tangney (Eds.), Handbook of self and identity
(pp. 106-127). New York: Guilford.

Kline, RB. (2005). Principles and practice of structural equation modeling. New York:
Guilford.

Lanning, K. (1988). Individual differences in scalability: An alternative conception of
consistency for personality theory and measurement. Journal of Personality and
Social Psychology, 55, 142-148.

Larsen, R.J. (1985). Individual differences in circadian activity rhythm and personality.
Personality and Individual Differences, 6, 305-31 1.

Larsen, R.J. (1987). The stability of mood variability: A spectral analytic approach to
daily mood assessments. Journal of Personality and Social Psychology, 52,
1195-1204.

Larsen, R.J., & Kasirnatis, M. (1990). Individual differences in the entrainment of mood
to the weekly calendar. Journal of Personality and Social Psychology, 58, 164-
171.

Locke, K.D. (2003). H as a measure of complexity of social information processing.
Personality and Social Psychology Review, 7, 268-280.

Maslow, AH. (1968). Toward a psychology of being. Princeton, NJ: Van Nostrand.

McConville, C., & Cooper, C. (1999). Personality correlates of variable moods.
Personality and Individual Differences, 26, 65-78.

McFarland, S.G., Sparks, CM. (1985). Age, education, and the internal consistency of
personality scales. Journal of Personality and Social Psychology, 49, 1692-1702.

McGee, R.K. (1962). Response style as a personality variable: By what criterion?
Psychological Bulletin, 59, 284-295.

53

McReynolds, P., Altrocchi, J ., & House, C. (2000). Self-pluralism: Assessment and
relations to adjustment, life changes, and age. Journal of Personality, 68, 347-
381.

Mischel, W., & Shoda, Y. (1995). A cognitive-affective system theory of personality:
Reconceptualizing situations, dispositions, dynamics, and invariance in
personality structure. Psychological Review, 102, 246-268.

Mischel, W., & Shoda, Y. (1999). Integrating dispositions and processing dynamics
within a unified theory of personality: The cognitive-affective personality system.
In LA. Pervin & O.P. John (Eds.), Handbook of personality: Theory and
research (pp. 197-218). New York: Guilford.

Nesselroade, J.R. (1992). Interindividual differences in intraindividual change. In L.M.
Collins and J .L. Horn (Eds.), Best methods for the analysis of change: Recent
advances, unanswered questions, future directions (pp.92-105). Washington,
DC: American Psychological Association.

Nesselroade, J.R., (2004). Intraindividual variability and short-term change.
Gerontology, 50, 44-47.

Paulhus, D.L., & Martin, CL. (1988). Functional flexibility: A new conception of
interpersonal flexibility. Journal of Personality and Social Psychology, 55 , 88-
101.

Paunonen, S.V., & Jackson, D.N. (1985). Idiographic measurement strategies for
personality and prediction: Some unredeemed promissory notes. Psychological
Review, 92, 486-511.

Penner, L.A., Shiffman, S., Paty, J .A., & Fritzsche, BA. (1994). Individual differences
in intraperson variability in mood. Journal of Personality and Social Psychology,
66, 712-721.

Pervin, LA. (2001). A dynamic systems approach to personality. European
Psychologist, 6, 172-176.

Reise, S.P., & Waller, N.G. (1993). Traitedness and the assessment of response pattem
scalability. Journal of Personality and Social Psychology, 65, 143-151.

Roberts, B.W., & Pomerantz, EM. (2004). On traits, situations, and their integration: A

developmental perspective. Personality and Social Psychology Review, 8, 402-
416.

Robins, R.W., Noftle, E.E., Trzesniewski, K.H., & Roberts, B.W. (2005). Do people

know how their personality has changed? Correlates of perceived and actual
personality change in young adulthood. Journal of Personality, 73, 489-521.

54

Robinson, M.D., Goetz, M.C., Wilkowski, B.M., & Hoffman, SJ. (2006). Driven to
tears or to joy: Response dominance and trait-based predictions. Personality and
Social Psychology Bulletin, 32, 629-640.

Rogers, CR. (1961). On becoming a person. Oxford, England: Houghton Mifflin.

Rosenberg, M. (1989). Society and the adolescent self-image. Middletown, CT:
Wesleyan University Press.

Rundquist, BA. (1950). Response sets: A note on consistency in taking extreme
positions. Educational and Psychological Measurement, 10, 97-99.

Ryan, R.M., & Deci, EL. (2000). Self-detennination theory and the facilitation of
intrinsic motivation, social development, and well-being. American Psychologist,
55 , 68-78.

Schimmack, U., Bbckenholt, U., & Reisenzein, R. (2002). Response styles in affect
ratings: Making a mountain out of a molehill. Journal of Personality Assessment,
78, 461-483.

Schneiderman, W. (1980). A personality dimension of consistency versus stability,
without the use of self-reports or ratings. Journal of Personality and Social
Psychology, 39, 158-164.

Schwarz, N. (1999). Self—reports: How the questions shape the answers. American
Psychologist, 54, 93-105.

Siegrist, M. (1996). The influence of self-consciousness on the internal consistency of
different scales. Personality and Individual Differences, 20, 115-117.

Snyder, M. (1974). Self-monitoring of expressive behavior. Journal of Personality and
Social Psychology, 30, 526-537.

Spielberger, CD. (1983). Manual for the State- Trait Anxiety Inventory. Palo Alto, CA:
Consulting Psychologists Press.

Stone, A.A., Shiffman, S.S., & DeVries, M.W. (1999). Ecological momentary
assessment. In D. Kahneman, E. Diener, & N. Schwarz (Eds.), Well-being: The
foundations of hedonic psychology (pp. 26-39). New York: Russell Sage.

Tellegen, A. (1988). The analysis of consistency in personality assessment. Journal of
Personality, 56, 621-663.

Tugade, M.M., Fredrickson, B.L., & Barrett, LP. (2004). Psychological resilience and

positive emotional granularity: Examining the benefits of positive emotions on
coping and health. Journal of Personality, 72, 1161-1190.

55

Watson, D. (2000). Mood and temperament. New York: Guilford.

Watson, D., & Tellegen, A. (2002). Aggregation, acquiescence, and the assessment of
trait affectivity. Journal of Research in Personality, 3 6, 589-597.

56

FOOTNOTES
1. Although the items on these scales exhibit high levels of internal consistency across
individuals, it is still possible for individuals to differ from one another in the extent that
their responses change across items (McFarland & Sparks, 1985).
2. According to meteorological records downloaded from Accuweather.com, average
daily precipitation was less than a tenth of an inch during the two week period, and no
rainfall was recorded on 10 of the days. High temperatures ranged from 38 to 81 degrees
(M = 63, SD = 12.48), but only five days had high temperatures that were beyond one
standard deviation of the mean. Finally, average daily wind speeds ranged from 4.50 to
15.50 mph (M = 9.85, SD = 3.44), but winds on nine days were within one standard
deviation of the mean.
3. Because variability in daily reports was not associated with self-consistency or well-
being in previous models, this index was removed from these analyses. However,
models that included daily variability indicated that response style had no impact on

these associations, and findings for the global indexes were nearly identical.

57

Table l. Bivariate correlations among indexes of personality variability, response style,

and relationship satisfaction (Study 1).

 

 

Mm 1 2 3 4 5 6
1. IPIP .--

2. Contextual .54* .--

3. Simpsons .43* .45* .--

4. Neutral objects .40* .47* .47* .--

5. Self-consistency -.01 -.11 .00 .01 .--

6. Satisfaction .04 .01 .16* .17* .45* .--

 

Note. IPIP = average of corrected trait-specific variability indexes across IPIP items;
Contextual = average of corrected trait-specific variability indexes across roles;
Simpsons = corrected cross-character standard deviation; Neutral objects = corrected
cross-item variability from ratings of neutral objects; Self-consistency = mean score on
the self-concept clarity scale; Satisfaction = means score on the measure of satisfaction

with relationships. *p<.05; N=149.

58

Table 2. Means and standard deviations of well-being measures (Study 2).

 

 

Measure M SD
Positive Affect 3.69 .60
Negative Affect 2.22 .63
Self-esteem 3.88 .72
Depression 1 .79 .48
Optimism 3.61 .60
Anxiety 2.08 .5 1

 

59

Table 3. Bivariate correlations among indexes of personality variability and response

 

 

style (Study 2).

Mm 1 2 3 4 5 6 7 8
1. IPIP --

2. Contextual .47* --

3. Daily .44* .57* ~-

4. Neutral objects .46* .40* .34* --

5. Friend ratings .35“ .59“ .47* .33* --

6. Weather reports .16* .15* .23* 11 .17* --

7. Self-consistency .09 -.15* -.10 05 -.02 04 --

8. Self-stability .03 -.13 -.15* .06 -.06 .02 .64* .--

 

Note. IPIP = average of corrected trait-specific variability indexes across IPIP items;

Contextual = average of corrected trait-specific variability indexes across roles; Daily =

average of corrected trait-specific variability indexes across days; Neutral Objects =

corrected cross-item variability from ratings of neutral objects; Friend ratings = average

of corrected cross-context standard deviations from descriptions of a friend; Weather =

corrected cross-day standard deviation in reports of weather conditions; Self-consistency

= mean score on the self-concept clarity scale; Self-stability = mean score on the self-

esteem stability scale. *p<.05; N=203.

60

 

 

 

 

 

 

 

 

 

 

 

   
 

55%; 29m

42..

61

   
 

.ezsac; .5:

 

 

 

 

 

 

 

 

 

 

 

 

.: see genesis 383 Essence ._ 2&5

 

 

 

 

 

 

 

 

 

 

 

_

h.

5232.; 23.

       

  
 

ham—Batuxr Emu

: 7V 2.-

    
 

 

 

33.80
Eom

 

 

a: S;

 

 

 

 

 

 

 

 

 

 

 

 

.Q «98%» 5.983 33:96.1?» wee Echoexbfiaetg 383 3825 «AexgetSmxxw .5295» Sam .N arm

62

 

 

 

 

 

 

 

 

 

 

 

 

   
 

55%; 23.

:3 8.-

 

_. :2 eozoemtum

 

 

 

     
 

Geo ac.

ham—Bats» m:—

ch.

 

 

 

 

 

 

 

 

 

 

 

 

.Q 33%.» echoekfiueu AEESBNE nee Eewuekbmmeete; E83 33:3 ”SeaeeheNeLXe SENSE Set .m panama

63

 

 

389:5

 

 

 

 

3830
.830 Z

 

 

 

 

 

 

 

 

 

 

 

 

25m omeonoM

   
 
      
 

£38; 23—

5:32.; be

 

 

 

 

 

 

 

 

 

 

 

.3 been.» mace. uncommon 38 bficetefxe 3V9: “Shane‘s .v 2&5

64

 

 

 

 

 

 

 

 

 

 

 

 

 

as. 3.53.; 2cm

 

 
 

 

 

 

    

 

 

 

 

 

 

 

889:5
2.3m bacon—mom V Sawflmmo
cacao
35:02

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.3 beam.» 35.... oedema: 38 8.23.9 “moonofbem .ESeeKbEaete: 3&3 Megs ”Aimeehefixxo Engage «3% .m PBwE

65

 

 

 

 

 

 

 

 

 

 

 

 

  

5:33.; use

 
   

 

 

 

 

 

 

 

 

 
  

 

 

 

uVB.
2.89:5
25m uaonmom A MNQ hm. V eocoémtam
*
a _ b.
28.30
.3282
.2. 3:36; ea

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

.Q ESQ e3». museums; nee .eozeQfiBa SN‘ESBNR .Eeuoetxbficete; SSS Medea uAfiQoESmLRe SENSE Sam .9 can—mi

66

 

 

 

 

 

 

 

 

 

 

 

 

 

   
 

.2...
eases; ace

 

 

 

 

.3. 3:33.; baa . e

 

 

   
 

 

 

~3=n£3> m7:

 

unh.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.«N beam.» atom? 3.83 .33 EQSM EQKAESSLSE Effieaenxe EVE: adhereeefl .5 Human

67

 

 

 

 

     
 

esssa> 23.

 

   
   

 

 

 

 

    
   
 

. Aeom.uv .Lm...
808mb e _ 5
flow
aaaamaooaam A $1 2.- Esea> ban
30280
1.20m e ~m.

 

 

83 .3.

e=Baa> .9:

 

 

 

 

 

 

 

 

 

 

.«N beam» Ae§§§8$eq wee hamburg $75839? EoBeKEeNS Meese «Senechefixfi Seaweae Sum .w 0.5mm...“

ens.

 

emu.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

68

 

 

 

 

 

 

 

 

 

 

 

 

 

beta}- . 8.-

 

  

 

        

 

 

easae; 29m

 

AGO awn-

 

 

?o_--v .Nmr

 

 

 

 

   
 

 

 

 

 

 
   

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

.QOQ u—wf
.28: e »
—m0mwO—O£0%wm Ar Amfirv No.1 ~—-DM§> —~QQ cem- U
mm chm. chm.
2: v cNm umw.— < —
5h.-
<z .. ESE; are
. m
on.
E .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.«N beam.» 533 Neomwfiefikum use AEBetg Q=-Ee-on\e uncuoeKEeBN arose m&-~§e.~ue~mx\e Seams-S «BK .m Eswmm

69

 

 

3.53%
5533

 

 

chm.

 

 

bEnatm> emo-

Bea

 

 

chm-

 

 

$00.30
35:02

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

25m canon—mom

 

 

 

 

 

 

   

 

 

 

 

 

new.
0
East; 29:
2
as. v Esme; ban . o
<
Essa; m7:
«mu. m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.«N been» 3b.,- e-SeRac-a nee ASSSE>KQ 33E huh-Emae-N .3 oSME

70

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mtomo-m
358.3

 

 

 

 

game;
22;

 

 

 

 

 

sumac
35:0 2

 

 

.«N beam.» can.» become-ex more 3:33.:er &=-Ee-ee\e ESUQESS MeeEe aA336u3~eLKe EEMSE Set .2 2:me

  

East; 23.

013m omeoamom

 

tax—Betaxr an:

 

 

 

 

 

 

 

 

 

 

 

 

och.

  

umb-

 

 

 

 

 

 

 

 

 

 

 

 

 

 

71

 

 

 

 

 

 

 

 

 

 

 

 

 

         
 

3.800%
.053 3

 

 

5:525, 2cm

 

G

 

  

 

 

 

 

 

 

 

 

 

 

 

 

  

8.- .3. 883
S. 8. cam
5:232; _. 25m omeoqmom V
@025 «900000
atom
chm.
38.30 0:38; 3.:
35:02

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.«N AESN» 038. 0.30%»; ES» .«OzSSe-eooSen .SmchS» «Na-50:30 SSS-\ESS Neefie nASmeSuSeu-xe SENSE Eek .2 BBNE

72

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mtogom
.853 >9

 

 

 

 

§=fi§>
v: at m

 

 

 

 

 

38.30
.8302

 

 

Lac

     
 

§=§s> 23.

*—w..

 
      
 

5.3:

mad
23w omuoamom V Rama—053m

92..-
ban—$3.? mama

 

 

 

 

 

 

 

 

 

 

 

 

 

5355.

 

 

.Eo

 

 

dun

 

 

mm

 

 

<2

 

 

 

<m

 

 

.«N Rush.» 3%“. 9283.2 38 .m=.~3.=m§ 6.:Butg Q=-E§3\a 3835283 wags «$520352? 523% Sam .2 uBmE

73

APPENDICES

74

Appendix A. Self-Concept Clarity Scale.

Please indicate the degree to which you agree or disagree with the following statements
using the scale below.

Scale:

1 — Strongly disagree

2 — Slightly disagree

3 — Neither disagree nor agree
4 - Slightly agree

5 — Strongly agree

1. My beliefs about myself ofien conflict with one another.
2. On one day I might have one opinion of myself and on another day I might have a
different opinion.
3. I spend a lot of time wondering about what kind of person I really am.
4. Sometimes I feel that I am not really the person that I appear to be.
5. When I think about the kind of person I have been in the past, I’m not sure what I
was really like.
6. I seldom experience conflict between the different aspects of my personality.
7. Sometimes I think I know other people better than I know myself.
8. My beliefs about myself seem to change very frequently.
9. If I were asked to describe my personality, my description might end up being
different from one day to another day.
10. Even if I wanted to, I don’t think I could tell someone what I’m really like.
11. In general, I have a clear sense of who I am and what I am.
12. It is often hard for me to make up my mind about things because I don’t really know
what I want.

75

Appendix B. Relationship satisfaction measure.

1 — Strongly disagree

2 — Somewhat disagree
3 — Neutral

4 - Somewhat agree

5 - Strongly agree

339°89‘99pr

I feel understood by my friends.

I am able to be open with my friends.

My friends convey confidence in me.

I feel that my fi'iends accept me.

I feel a lot of trust in my friends.

I feel that my friends care about me as a person.

I don’t feel very good about the way my friends talk to me.
I feel able to share my feelings with my friends.

I am satisfied with my relationships with friends.

I feel understood by my family.

. I am able to be open with my family.

. My family convey confidence in me.

. I feel that my family accept me.

. I feel a lot of trust in my family.

. I feel that my family care about me as a person.

. I don’t feel very good about the way my family talk to me.
. I feel able to share my feelings with my family.

. I am satisfied with my relationships with family.

76

 

Appendix C. Stability of Self Scale.

1. Does your opinion of yourself tend to change a good deal, or does it always
continue to remain the same?

1) Changes a great deal.

2) Changes somewhat

3) Changes very little

4) Does not change at all

2. Do you ever find that on one day you have one opinion of yourself and on
another day you have a different opinion?

1) Yes, this happens often

2) Yes, this happens sometimes

3) Yes, this rarely happens

4) No, this never happens

3. I have noticed that my ideas about myself seem to change very quickly.
1) Agree
2) Disagree

4. Some days I have a very good opinion of myself; other days I have a very poor
opinion of myself.

1) Agree

2) Disagree

5. I feel that nothing, or almost nothing, can change the opinion I currently hold
about myself.

1) Agree

2) Disagree

77

Appendix D. Self-concept Differentiation Scale.

Below is a list of adjectives that can be used to describe people. We would like you to
think about how you see yourself when you are with friends. For each of the following

adjectives, using the scale below, please indicate how each describes you when you are
BEING A FRIEND.

SCALE:

1 — Does not describe me

2 — Describes me a little

3 — Describes me moderately
4 — Describes me well

5 — Describes me very well

1. Talkative 11. Adventurous
2. Cooperative 12. Caring

3. Organized l3. Responsible
4. Irritable l4. Insecure

5. Intelligent 15. Creative

6. Assertive

7. Rude

8. Hardworking

9. Optimistic

10. Inquisitive

Next we would like you to think about how you see yourself when you are with a
romantic partner. For each of the following adjectives, using the scale below, please
indicate how each describes you when you are BEING A ROMANTIC PARTNER.

SCALE;

1 — Does not describe me

2 — Describes me a little

3 - Describes me moderately
4 — Describes me well

5 — Describes me very well

1. Talkative 11. Adventurous
2. Cooperative l2. Caring

3. Organized 13. Responsible
4. Irritable 14. Insecure

5. Intelligent 15. Creative

6. Assertive

7. Rude

8. Hardworking

9. Optimistic

10. Inquisitive

78

Next we would like you to think about how you see yourself when you are with family
members. For each of the following adjectives, using the scale below, please indicate
how each describes you when you are BEING A FAMILY MEMBER.

SCALE:

1 — Does not describe me

2 - Describes me a little

3 — Describes me moderately
4 — Describes me well

5 — Describes me very well

1. Talkative 11. Adventurous
2. Cooperative 12. Caring

3. Organized l3. Responsible
4. Irritable l4. Insecure

5. Intelligent 15. Creative

6. Assertive

7. Rude

8. Hardworking

9. Optimistic

10. Inquisitive

Next we would like you to think about how you see yourself when you are being a
student. For each of the following adjectives, using the scale below, please indicate how
each describes you when you are BEING A STUDENT.

SCALE:

1 — Does not describe me

2 — Describes me a little

3 — Describes me moderately
4 — Describes me well

5 — Describes me very well

1. Talkative ll. Adventurous
2. Cooperative 12. Caring

3. Organized 13. Responsible
4. Irritable l4. Insecure

5. Intelligent 15. Creative

6. Assertive

7. Rude

8. Hardworking

9. Optimistic

10. Inquisitive

79

Appendix E. Daily Questionnaire.

1. How sunny is it today?
1.) Not sunny at all
2.) A little sunny
3.) Moderately sunny
4.) Mostly sunny
5.) Very sunny

2. How cold is it today?
1.) Not cold at all
2.) A little cold
3.) Moderately cold
4.) Fairly cold
5.) Very cold

3. How windy is it today?

1.) Not windy at all
2.) A little windy

3.) Moderately windy
4.) Fairly windy

5.) Very windy

For the following adjectives, please indicate how well each describes how you have been

today.

1 - Does not describe me

2 - Describes me a little

3 — Describes me moderately
4 — Describes me well

5 — Describes me very well

1. Talkative
2. Cooperative
3. Organized
4. Irritable

5. Energetic

6.
7.
8.
9.
10.

Insecure
Rude
Inquisitive
Responsible
Creative

Please indicate how much you felt each of the following emotions today.

1 — None at all

2 - Less than usual

3 — No more than usual

4 — More than usual

5 - A lot more than usual

Affection
Joy
Anger
Shame
Sadness
Love
Worry

999.4%»!9?

80

. Guilt

. Loneliness
10.
ll.
12.
13.
14.

Contentment
Disgust
Nervousness
Amusement
Interest

Appendix F. 5 0-Item IPIP Scale.

On this page, there are phrases describing people's behaviors. Please use the rating scale
below to describe how accurately each statement describes you. Describe yourself as you
generally are now, not as you wish to be in the future. Describe yourself as you honestly
see yourself in relation to other people you know of the same sex as you are, and roughly
your same age. Please read each statement carefully, and then fill in the bubble on the
scantron that corresponds to the number on the scale.

SCALE;

1 — Very Inaccurate

2 — Moderately Inaccurate

3 — Neither Inaccurate nor Accurate
4 — Moderately Accurate

5 — Very Accurate

I.Am the life of the party.
2. F eel little concern for others.
3. Am always prepared.
4. Get stressed out easily.
5. Have a rich vocabulary.
6. Don’t talk a lot.
7. Am interested in people.
8. Leave my belongings around.
9. Am relaxed most of the time.
10. Have difficulty understanding
abstract ideas.
11.Feel comfortable around people.
12. Insult people.
13. Pay attention to details.
14. Worry about things.
15. Have a vivid imagination.
16. Keep in the background.
17. Sympathize with others’ feelings.
18. Make a mess of things.
19. Seldom feel blue.
20. Am not interested in abstract ideas.
21. Start conversations.
22. Am not interested in other people’s
problems.
23. Get chores done right away.
24. Am easily disturbed.
25. Have excellent ideas.
26. Have little to say.
27. Have a soft heart.

81

28. Often forget to put things back in
their proper place.

29. Get upset easily.

30. Do not have a good imagination.

3 1. Talk to a lot of different people at
parties.

32. Am not really interested in others.

33. Like order.

34. Change my mood a lot.

35. Am quick to understand things.

36. Don’t like to draw attention to
myself.

37. Take time out for others.

38. Shirk my duties.

39. Have frequent mood swings.

40. Use difficult words.

41. Don’t mind being the center of
attention.

42. Feel others’ emotions.

43. Follow a schedule.

44. Get irritated easily.

45. Spend time reflecting on things.

46. Am quiet around strangers.

47. Make people feel at ease.

48. Am exacting in my work.

49. Ofien feel blue.

50.Am full of ideas.

Appendix G. Intensity and Time Aflect Survey.

Next, we want to know how you feel in general, that is on average. To what extent do
you experience each of the following emotions in general?

w
1 - Very slightly or not at all
2 - A little
3 — Moderately
4 - Quite a bit
5 — Extremely
1. Affection 13. Caring
2. Joy l4. Contentment
3. Fear 15. Anxiety
4. Anger l6. Disgust
5. Shame 17. Regret
6. Sadness l8. Unhappiness
7. Love 19. Fondness
8. Happiness 20. Pride
9. Worry 21. Nervous
10. Irritation 22. Rage
11. Guilt 23. Embarrassment
12. Loneliness 24. Depression

82

Appendix H. Neutral Objects Questionnaire.

Please indicate the response that best represents your feelings about the following items.

Bile:
1 — Very Dissatisfied

2 - Somewhat Dissatisfied
3 - Neutral

4 — Somewhat Satisfied

5 — Very Satisfied

95”.“???pr

The city in which you live.
The residence where you live.
The neighbors you have.

The high school you attended.

The climate where you live.

The movies being produced today.

The quality of food you buy.
Today’s cars.

Local newspapers.

Your relaxation time.

. Your first name.
. The people you know.
. Television programs.

83

14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.

Local speed limits.

The way people drive.
Advertising.

The way you were raised.
Telephone service.
Public Transportation.
Restaurant food.
Yourself

Modern art.

Popular music.

8 V2” x 11” paper

Your telephone number.

Appendix 1. Global Self-Esteem (*) and Optimism.

Listed below are a number of statements concerning how you may see yourself. Please
read each item and use the scale below to rate how well each statement describes you.
Fill in the appropriate circle to indicate your response:

1 - Does not describe me

2 — Describes me a little

3 — Describes me moderately
4 — Describes me well

5 — Describes me very well

.‘O.°°.".‘"S"PP’.N."‘

I feel that I am a person of worth, at least on an equal basis with others.*
I feel that I have a number of good qualities.*

In uncertain times, I usually expect the best.

It’s easy for me to relax.

All in all, I am inclined to feel that I am a failure.*

If something can go wrong for me, it will.

I am able to do things as well as most other people.*

I’m always optimistic about my future.

I feel I do not have much to be proud of. *

I wish I could have more respect for myself."

. On the whole, I am satisfied with myself!“

. I enjoy my friends a lot.

. It’s important for me to keep busy.

. I take a positive attitude toward myself.“

. I hardly ever expect things to go my way.

. I certainly feel useless at times.*

. I don’t get upset too easily.

. I rarely count on good things happening to me.

. Overall, I expect more good things to happen to me than bad.
. At times, I think I am no good at all.*

84

Appendix J. Depression Scale.
How often you felt or behaved this way during the past week.

0 — Rarely/None

l — Some or a Little of the Time
2 - A lot ofthe Time

3 — Most of the Time

I was bothered by things that usually don’t bother me.

I felt that I could not shake off the blues even with the help of my friends or family.
I felt that I was just as good as other people.

I had trouble keeping my mind on what I was doing.

I felt that everything I did was an effort.

I felt hopeful about the future.

I felt my life had been a failure.

I felt fearful.

People were unfriendly.

I felt lonely.

999089509959:-

~

85

Appendix K. Anxiety Scale.

A number of statements which people have used to describe themselves are given
below. Read each statement and then select the appropriate number to indicate how you
generally feel. There are no right or wrong answers. Do not spend too much time on any
one statement but give the answer which seems to describe how you generally feel.

1 — Almost never
2 — Sometimes

3 — Often

4 — Almost always

Nv—ou—Iu—nu—tu—Iu—ot—s—no—n
999°89‘V'99’P"

u—n
Ppwflp‘S‘PP‘Nr‘

I feel pleasant

I tire quickly

I feel like crying

I wish I could be as happy as others seem to be

I am losing out on things because I can’t make up my mind soon enough
I feel rested

I am “calm, cool, and collected”

I feel that difficulties are piling up so that I cannot overcome them

I worry too much over something that really doesn’t matter

I am happy

. I am inclined to take things hard

I lack self-confidence

. I feel secure

I try to avoid facing a crisis or difficulty

. I feel blue

I am content

Some unimportant thought runs through my mind and bothers me

I take disappointments so keenly that I can’t put them out of my mind
I am a steady person

I become tense and upset when I think about my present concerns

86

Appendix L. Personality ratings of the Simpsons.

For the following individuals, please use the rating scale to indicate how well you think

each adjective describes him or her:

1 — Does not describe this person

2 — Describes this person a little

3 - Describes this person moderately
4 — Describes this person well

5 — Describes this person very well

Bart Simpson

1. Energetic

2. Trustful

3. Dependable
4. Vulnerable
5. Philosophical

Lisa Simpson

1. Energetic

2. Trustful

3. Dependable
4. Vulnerable
5. Philosophical

Homer Simpson

1. Energetic

2. Trustful

3. Dependable
4. Vulnerable
5. Philosophical

Marge Simpson

1. Energetic

2. Trustful

3. Dependable
4. Vulnerable
5. Philosophical

87

FPS-52”.“?

999°>IQ

PPWNF”

99”.“?

Talkative
Cooperative
Organized
Irritable
Intelligent

Talkative
Cooperative
Organized
Irritable
Intelligent

Talkative
Cooperative
Organized
Irritable
Intelligent

Talkative
Cooperative
Organized
Irritable
Intelligent

Appendix M. Variability of a friend across contexts.

Below is a list of adjectives that can be used to describe people. We would like you to
think a person that you know well. This person should be a close friend with whom you
feel close. For each of the following adjectives, using the scale below, please indicate
how each describes your friend IN GENERAL.

W

1 — Does not describe him or her

2 — Describes him or her a little

3 — Describes him or her moderately
4 - Describes him or her well

5 — Describes him or her very well

1. Energetic

2. Trustful

3. Dependable
4. Vulnerable
5. Philosophical

Next we would like you to think about how you see your friend when he or she is at work
or at a job. For each of the following adjectives, using the scale below, please indicate
how each describes your friend when he or she is BEING A WORKER.

Energetic
Trustfirl
Dependable
Vulnerable
Philosophical

9993"?“

Next we would like you to think about how you see your friend when he or she is with
people that he or she doesn’t know. For each of the following adjectives, using the scale
below, please indicate how each describes your fiiend when he or she is BEING A
STRAN GER.

l. Energetic

2. Trustful

3. Dependable
4. Vulnerable
5. Philosophical

88

 

 

   

I111:1Jillian