Sunteți pe pagina 1din 21

Personal Relationships, 21 (2014), 655675. Printed in the United States of America.

Copyright 2014 IARR; DOI: 10.1111/pere.12056

A prototype analysis of infidelity

DANA A. WEISER,a CAMILLE B. LALASZ,b DANIEL J. WEIGEL,b


b
AND WILLIAM P. EVANS
a Texas Tech University and b University of Nevada, Reno
Abstract
The current research tested whether the concept of infidelity is prototypically organized and whether laypeoples
conceptualizations of infidelity are consistent with how researchers have operationalized this construct. Across 4
studies, results indicated that infidelity is indeed prototypically organized as individuals are able to list and rate how
central certain features are to the infidelity construct. Furthermore, there was evidence that the centrality ratings
influenced how individuals processed information in a series of memory tasks and narratives about infidelity
experiences. Laypeople are less likely than researchers to consider the presence of particular behaviors (i.e., flirting,
kissing, and sexual intercourse) as defining qualities of infidelity. Instead, laypeople focus more on the concealment of
behaviors and the resulting emotional fallout from infidelity.

Infidelity is an extremely complex and intriguing construct to study as the vast majority
of individuals expect monogamy in their
romantic relationships (Sheppard, Nelson,
& Andreoli-Mathie, 1995; Treas & Giesen,
2000), yet a significant proportion of individuals engage in infidelity (Blow & Hartnett,
2005a). For instance, research on married
couples has indicated that 87% of newlywed
participants felt that cheating would never
be acceptable in their relationship because it
would be a breach of trust, commitment, and
the marital contract (Wiederman & Allgeier,
1996). Among studies utilizing large representative samples, however, approximately
Dana A. Weiser, Human Development & Family Studies, Texas Tech University, Lubbock; Camille B. Lalasz,
Interdisciplinary Social Psychology Program, University
of Nevada, Reno; Daniel J. Weigel, Human Development
& Family Studies/Cooperative Extension, University of
Nevada, Reno; William P. Evans, Human Development &
Family Studies, University of Nevada, Reno.
The co-first authors contributed equally to this project;
the order of authorship was determined by coin flip.
The researchers wish to thank Julianna Chomos, Bret
Davis, and Michael Doane for their assistance with coding
for Study 4.
Correspondence should be addressed to Dana A.
Weiser, Human Development & Family Studies, Texas
Tech University, Lubbock, TX 79409-1230, e-mail:
Dana.Weiser@ttu.edu.

22%25% of men and 11%15% of women


in marital relationships were found to have
engaged in sexual infidelity (Allen et al.,
2005), and the percentages of those engaging
in infidelity are even higher within dating
relationships (Wiederman & Hurd, 1999).
Individuals who discover that a partner
has engaged in an infidelity are likely to
experience intense emotional reactions, such
as jealousy, hurt, anger, and disgust (Becker,
Sagarin, Guadagno, Millevoi, & Nicastle,
2004). The discovery of a partners infidelity
is also associated with a reduction in psychological well-being, including an increased
incidence of depression and anxiety, as well as
reductions in self-esteem (Buunk, 1995; Cano
& OLeary, 2000). Likewise, individuals who
engage in infidelity often experience feelings
of remorse, guilt, and depression (Allen et al.,
2005) and express fears of hurting their partner
or experiencing disapproval from their family
and friends after the relationship transgression
(Allen & Baucom, 2006). Not surprisingly,
infidelity is one of the strongest and most
proximal predictors of relationship dissolution for both married and dating couples
(Amato & Previti, 2003; Hall & Fincham,
2006).

655

656

Due to the prevalence of infidelity and the


potential negative consequences associated
with this relationship transgression, personal
relationships researchers and clinicians have
amassed a substantial literature in order to
understand the experience of infidelity, treat
individuals who experience infidelity, and
prevent future infidelities. Perhaps one of
the clearest themes to emerge from a review
of the infidelity literature is that there is no
one singular definition, or even name, for the
construct of infidelity. Indeed, a variety of
terms have been used to even refer to infidelity,
including unfaithfulness (Roscoe, Cavanaugh,
& Kennedy, 1988; Yarab, Allgeier, & Sensibaugh, 1999), cheating (Brand, Markey, Mills,
& Hodges, 2007), affairs (Atkins, Dimidjian,
& Jacobsen, 2001), and extradyadic behavior
(Buunk, 1995). The purpose of this study is
to address this lack of conceptual clarity by
conducting a prototype analysis of infidelity.
In doing so, we hope to gain information about
how laypeople conceptualize infidelity and
explore how these lay concepts compare to
the conceptualization of infidelity utilized by
researchers.
Differing infidelity definitions
At the broadest level, scholars have defined
infidelity as any type of secret emotional,
sexual, or romantic behavior that violates the
exclusivity norms of a romantic relationship
(Glass, 2002). Researchers have emphasized
a variety of qualities that define the infidelity
construct, although these components are
not always consistent. For instance, Bernard
(1974) proposed that when individuals fail
to love, honor, and support their partners,
they are violating the vows of a committed
romantic relationship, and thus are engaging in infidelity. By contrast, Pittman and
Wagers (2005) argued that the hallmark of
infidelity is the secrecy and concealment of
extradyadic behaviors, rather than any one specific extradyadic behavior. Thompson (1983)
postulated a broad three-pronged definition
of infidelity that required the following: (a)
that extradyadic behavior is not permitted by
romantic partners, (b) that behavior occurs
outside the primary relationship, and (c) a

D. A. Weiser et al.

description of the actual extradyadic behavior


(intercourse, flirting, etc.).
Researchers also have distinguished three
types of infidelity: emotional only, sexual only,
and composite (i.e., emotional and sexual)
infidelities (Blow & Hartnett, 2005a; Hall
& Fincham, 2006). Under these definitions,
individuals may engage in emotional infidelity
by developing deep feelings for and a bond
with an extradyadic partner or sexual infidelity by becoming sexually involved with
an extradyadic partner. In addition, individuals can engage in a composite infidelity that
is characterized by both a deep emotional
and physical involvement with an extraneous relationship partner. Less commonly,
researchers have classified infidelities into
other general types, such as one-night stands
versus long-term relationships or Internet
relationships versus work relationships (Blow
& Hartnett, 2005a).
Other researchers have employed even narrower infidelity definitions by focusing on the
presence of specific behaviors, such as spending time with another individual, going on a
romantic date, provocative talk, flirting, kissing, fondling, oral sex, as well as sexual intercourse (Blow & Hartnett, 2005b; Luo, Cartun,
& Snider, 2010; Roscoe et al., 1988). Thinking a lot about or dreaming and fantasizing
about another individual besides the romantic
partner may also be considered components of
infidelity (Feldman & Cauffman, 1999; Yarab
et al., 1999). Wilson, Mattingly, Clark, Weidler, and Bequette (2011) developed the Perceptions of Infidelity Scale in order to measure
the extent to which individuals perceive certain activities to be infidelity. The researchers
found support for an infidelity scale composed
of three subscales: ambiguous behaviors (e.g.,
going someplace with another individual
and talking on the phone/Internet), deceptive
behaviors (e.g., lying and withholding information), and explicit behaviors (e.g., fondling,
dating, oral sex, and sexual intercourse).
A prototype approach to infidelity
In sum, there is great variability in how
researchers have defined, identified, and
measured infidelity. This state of affairs

Infidelity prototype

indicates that infidelity may not be definable


in a classical sense. According to the classical
view of definitions, category membership is
determined by necessary and sufficient conditions in an all-or-nothing manner in which
an instance that meets the criteria is a member
and all others are not (Fehr, 1988; Lambert,
Graham, & Fincham, 2009; Weigel, 2008).
Because each member is required to possess
the same set of attributes, all members of a category are also assumed to be equally representative of that category. However, there are many
phenomena of interest in which such simplistic
classifications are not warranted, and there are
many constructs of interest that necessitate far
more complex conceptualizations. Infidelity is
likely such a construct as there is so much lack
of agreement about what is infidelity. Infidelity
definitions vary widely in their scope as well as
content, which indicates that in order to bring
clarity to the concept, a different approach
must be undertaken in evaluating the infidelity
construct.
As an alternative to the classical view of
concepts, Rosch (1975, 1978) proposed a prototype theory to describe the fuzzy internal structure of categories. After examining
types of concepts, Rosch argued that many categories maintain an internal prototype structure
in which clearest cases are surrounded by other
members that can be ordered in terms of their
degree of similarity to the prototypical cases.
For example, when examining the concept of
furniture, Rosch (1975) found that furniture
types could be ordered from most prototypical (i.e., chair and dresser) to least prototypical (i.e., rug and stove), and that this internal structure influenced the way in which a
series of experimental tasks were performed.
Fehr (1988) extended this research by determining that like types of concepts, features of
concepts may also be prototypically structured.
Specifically, in an analysis of love and commitment, Fehr demonstrated that some features
of love (e.g., trust and caring) and commitment (e.g., loyalty and responsibility) are more
central than others (e.g., love: dependency and
scary; commitment: think about other all the
time and feel trapped) and that the two concepts
of love and commitment are largely overlapping.

657

Since these seminal studies were published,


the prototype approach has become a popular
tool to better understand laypeoples understandings of concepts. In fact, the prototype
approach has been used to study a variety
of relationship constructs, including jealousy (Sharpsteen, 1993), relationship quality
(Hassebrauck, 1997), romantic love (Regan,
Kocan, & Whitlock, 1998), respect (Frei
& Shaver, 2002), forgiveness (Kearns & Fincham, 2004), intimacy (Birnie-Porter & Lydon,
2013), family (Weigel, 2008), and gratitude
(Lambert et al., 2009). In these studies,
prototypes are defined as individuals understandings of the meaning of concepts, which
consist of interrelated sets of features that are
commonly ascribed to the concept (Surra &
Bohman, 1991), and the focus is on assessing
the centrality of these features (Kearns &
Fincham, 2004). In order for a construct to
demonstrate a prototype structure, two conditions must be met. First, individuals must be
able to name features relevant to the concept
and then reliably rate the centrality of these
features to the concept. Second, the centrality
rating of each feature should have implications
for how individuals then process and think
about the concept.
Birnie-Porter and Lydon (2013) argue
that a prototype analysis is warranted when
researchers attempt to answer the question:
What is it? and Harasymchuk and Fehr
(2013) advocated for a prototype approach
when there is lack of consensus on the definition of a concept. No simple infidelity
definition exists because the concept cannot be
classically defined in black and white terms.
To illustrate, Kruger et al. (2013) asked participants to rate to what degree 27 different
behaviors were indicative of cheating using a
11-point scale ranging from 0% to 100%. The
researchers found that no single behavior was
universally considered cheating, and that mean
ratings for items ranged from 8% to 98% as
indicative of cheating. These findings further
highlight that infidelity should be considered
a fuzzy construct as there are certain actions
and experiences that are more characteristic of
infidelity compared to others. Consequently,
we propose that infidelity is prototypically
organized, and there are numerous benefits

658

to understanding how laypeople think about


infidelity.
Why are lay conceptions of infidelity
important?
Currently, there has been little research that
examined individuals concepts of infidelity.
This is a limitation of the literature as there
are a number of ways in which lay conceptualizations may be useful for researchers
(Fehr & Russell, 1991). First, understanding
lay peoples conceptualizations of concepts,
such as infidelity, can provide researchers with
hypotheses to test (Fehr & Russell, 1991). A
great deal of infidelity research is concerned
with identifying variables that can be used
to predict infidelity. It is possible that how
an individual conceptualizes and understands
infidelity may relate to the likelihood that they
engage in this behavior. Second, a prototype
approach allows researchers to understand the
breadth of a concept (Gregg, Hart, Sedikides,
& Kumashiro, 2008) and potentially highlight
previously ignored components of the concept
(Harasymchuk & Fehr, 2013). Lay theories of
relationships shape individuals relationship
behaviors (Fletcher & Thomas, 1996), so
further identifying the various components of
these lay theories may again help us to better
understand infidelity.
Third, understanding lay conceptions of
infidelity may also have significant implications for research methods as it can help free
scientists from hidden assumptions and confusion (Fehr & Russell, 1991, p. 436). Many
researchers conduct studies in which they ask
participants to self-report experiences with
infidelity. A prototype approach may provide
better insight into what participants mean
when they say have or have not experienced an
infidelity. Finally, understanding lay conceptions of infidelity is important as most research
is done based on the assumption that what the
investigator is measuring corresponds to the
idea of infidelity in the minds of participants;
comparing laypeoples meanings to expert
definitions, therefore, is critical.
In addition to having implications for theory
and research, understanding lay conceptions
of infidelity may also play an important role

D. A. Weiser et al.

in applied settings. Infidelity is one of the


most difficult relationship problems to address
in therapy and further insight into how to
rebuild relationships following infidelity is
needed (Peluso & Spina, 2008; Whisman,
Dixon, & Johnson, 1997). Understanding how
everyday people conceptualize and experience
infidelity may help illuminate the many facets
of infidelity and allow clinicians to better
help their clients work through the aftermath of infidelity, as well as prevent future
infidelities.
Overview
Four studies were conducted to explore how
laypeople understand and define the concept of
infidelity. In Study 1, participants produced a
list of infidelity features. In Study 2, a separate
group of participants rated the perceived centrality of each feature from the list created in
Study 1. Studies 3 and 4 were then conducted
to test whether feature centrality would influence how individuals thought about infidelity.
All studies took place at a mid-sized university
in the Western United States.
Study 1: Generation of Prototype Features
Method
The purpose of Study 1 was to generate features of the concept of infidelity. Participants
were asked to list all of the features of infidelity
in a free-response format.
Participants
Participants were 155 undergraduate students
(64 male, 91 female) who ranged in age
from 17 to 55 years (M = 22.2). Participants
were recruited from sociology and human
development and family studies courses. The
sample included 109 Caucasians (70.3%), 13
Latinos (8.4%), 10 Asian Americans (6.5%),
9 African Americans (5.8%), 9 multiethnic individuals (5.8%), and 5 individuals of
other backgrounds (3.2%). Among participants, 56 individuals were not currently in
a romantic relationship (36.6%), whereas
10 individuals were dating multiple people (6.5%), 56 individuals were dating one

Infidelity prototype

individual exclusively (36.6%), 8 individuals


were cohabitating (5.2%), 10 individuals were
engaged (6.5%), and 13 individuals were married (8.5%). Participants currently involved
in a romantic relationship were in their relationships on average 32.9 months. In addition,
59 individuals (39.6%) indicated that they
had previously been cheated on in a romantic relationship, and 38 individuals (24.7%)
reported that they had committed a relationship
infidelity.
Procedure
Participants first viewed the following instructions (adapted from Frei & Shaver, 2002):
This questionnaire is part of a project on the
sorts of things we have in mind when we
hear and use words. The term INFIDELITY
means a variety of things to people. In this
study, we are interested in your views and
thoughts about the term infidelity and what
you believe are its defining features and
characteristics. For example, if you were
instructed to list the features of apple, you
might list red, sweet, fruit, juicy, healthy,
satisfies hunger, and so forth.
In the space below please list as many
features of INFIDELITY as comes to
mind. The features of INFIDELITY that
you list may include characteristics, components, functions, feelings, ideas, or
behaviorsanything that helps define
INFIDELITY. There are no right or wrong
answersjust give us your opinion.
Participants were then asked to list all features of infidelity that came to mind within a
3-min period. Participants were provided with
a sheet of paper with 20 lines. Upon finishing this task, participants provided information about their demographic background, relationship status, and infidelity history. All procedures were completed during regular class
time.
Results
One member of the research team first compiled a verbatim list of the features generated

659

by participants, only correcting for spelling.


Participants originally listed 1,237 items
and on average participants provided 7.86
responses. These features were then condensed by the first three authors, following
the procedure of Fehr (1988) and Kearns
and Fincham (2004). First, it was decided to
remove 14 items that appeared to be unrelated
to the concept of infidelity, such as terrorist,
Fidel Castro, and Fidelity Investments. Next,
the list of features was reduced by extracting
linguistic units. All single word units, such
as unfaithful and cheating, were classified
as distinct features. When participants provided phrases, the authors judged whether
the phrase should be considered one feature
or divided into multiple features; any phrase
that conveyed one single thought was left
intact, whereas phrases that conveyed several
thoughts were divided accordingly.
Next, two of the authors examined the
features to determine where they could be
combined. Features were grouped if they were
modified by an adjective or adverb, were
different forms of the same word, or shared a
similar meaning. For example, items such as
hurts, hurt feelings, and hurt were all grouped
into the attribution category hurt, and items
such as without morals and lack of morals
were grouped into the category immoral. In
contrast, items such as hurt and hurtful were
kept separate as one seemed to convey an
emotion and the other a characteristic of the
infidelity act. The researchers carefully coded
all items to be as conservative as possible
while also appropriately reducing the list of
features. Interrater reliability, for agreement
about grouping the features, was calculated
using Cohens Kappa and found to be good
( = .89). Any disagreements between the
two coders were resolved through discussion
among the authors. This coding procedure
resulted in a list of 434 features. To make
the list more manageable, features listed only
once (283) or twice (56) were excluded from
future analyses. Items mentioned 3 times or
more were retained and resulted in a final list
of 95 features (see Table 1). The most popular feature was cheating, which 51.61% of
participants mentioned, followed by unfaithful (mentioned by 27.10% of participants)

660

D. A. Weiser et al.

Table 1. Infidelity features generated in Study 1, sorted by Study 2 centrality rankings


Study 1
Features
Unfaithful
Cheating
Broken trust
Being with someone who is not your
partner
Loss of trust
Lies
Affair
Untrustworthy
Betrayal
Dishonest
Physically cheating
Cheater
Disloyal
Lying
Mistrust
Disrespectful
Heartbreak
Sneaking around
Adultery
Wrong
Hurtful
Sex
Secretive
Deceive
Secret
Liar
Hurt
Emotions
Selfish
Lust
Sneaky
Breakup
Bad
Pain
Emotional cheating
Backstabbing
Anger
Unforgivable
Immoral
Tears
Relationship
Temptation
Traitor
Guilt
Crying

Study 2

Frequency

Centrality rating

SD

42
80
4
8

27.10
51.61
2.58
5.16

6.90
6.87
6.79
6.78

1.78
1.81
1.71
1.97

4
12
4
16
16
24
3
15
9
24
8
3
11
4
6
30
25
24
3
5
3
18
12
7
11
10
9
3
17
10
3
3
22
6
15
3
14
5
3
4
4

2.58
7.74
2.58
10.32
10.32
15.48
1.94
9.68
5.81
15.48
5.16
1.94
7.10
2.58
3.87
19.35
16.13
15.48
1.94
3.23
1.94
11.61
7.74
4.52
7.10
6.45
5.81
1.94
10.97
6.45
1.94
1.94
14.19
3.87
9.68
1.94
9.03
3.23
1.94
2.58
2.58

6.77
6.67
6.62
6.60
6.60
6.60
6.58
6.58
6.55
6.53
6.52
6.44
6.43
6.42
6.42
6.30
6.30
6.25
6.23
6.22
6.19
6.19
6.14
6.12
6.09
6.05
6.04
5.95
5.93
5.92
5.86
5.86
5.82
5.81
5.81
5.78
5.77
5.75
5.72
5.70
5.67

1.67
1.74
1.92
1.78
1.88
1.77
1.94
1.98
1.89
1.82
1.75
1.86
1.87
1.87
2.02
1.95
1.87
1.93
1.83
1.92
1.86
1.94
1.92
1.79
1.99
1.95
1.88
1.90
1.96
2.02
2.18
2.08
1.95
2.00
2.16
1.98
2.00
2.01
2.19
2.03
2.06

Infidelity prototype

661

Table 1. Continued
Study 1
Features
Unfair
Sad
Uncommitted
Broken family
Cruel
Divorce
Intimacy
Broken
Unhappy
Asshole
Kissing
Sorrow
Trust
Mistake
Insecure
Weakness
Dirty
Lose
Regretful
Depression
Jealous
Lack of communication
Player
Deviance
Unsatisfied
Jerk
Not loving
Marriage
Sinful
Love
Males
Hate
Family
Coward
Dangerous
Lonely
Slut
Bad individual
Common
Rebellion
Females
Moral
Excitement
Children
Boredom
Cheap
Pointless
Alcohol
Crime
Physically violent behavior

Study 2

Frequency

Centrality rating

SD

6
22
3
4
3
11
3
3
11
7
6
6
4
3
3
6
5
7
3
7
4
4
4
4
4
3
5
8
3
5
15
10
3
3
3
5
7
3
6
4
8
3
7
3
4
3
3
3
3
5

3.87
14.19
1.94
2.58
1.94
7.10
1.94
1.94
7.10
4.52
3.87
3.87
2.58
1.94
1.94
3.87
3.23
4.52
1.94
4.52
2.58
2.58
2.58
2.58
2.58
1.94
3.23
5.16
1.94
3.23
9.68
6.45
1.94
1.94
1.94
3.23
4.52
1.94
3.87
2.58
5.16
1.94
4.52
1.94
2.58
1.94
1.94
1.94
1.94
3.23

5.66
5.64
5.64
5.63
5.62
5.56
5.53
5.51
5.47
5.46
5.40
5.40
5.38
5.37
5.33
5.32
5.32
5.26
5.26
5.24
5.21
5.20
5.19
5.17
5.17
5.16
5.12
5.08
5.08
5.08
4.98
4.97
4.96
4.81
4.79
4.75
4.74
4.51
4.44
4.36
4.32
4.32
4.26
4.26
4.19
4.18
4.11
4.04
3.61
3.41

2.08
2.01
2.16
2.04
2.02
2.11
2.20
2.21
2.04
2.31
2.14
2.11
2.37
2.02
2.09
2.01
2.19
2.17
2.03
2.11
1.94
2.14
2.27
2.04
2.13
2.15
2.15
2.28
2.39
2.39
2.25
2.16
2.34
2.11
2.15
2.17
2.31
2.03
2.12
2.05
2.22
2.49
2.31
2.47
2.14
2.26
2.20
2.20
2.32
2.18

662

and wrong (mentioned


participants).

D. A. Weiser et al.

by

19.35%

of

Discussion
These results indicate that there is a great
deal of variety in how individuals define and
construe infidelity. No single feature was
named by all participants, but certain themes
of features began to emerge. Just over half
of all participants indicated that infidelity
was characterized by cheating. This finding
is not surprising as infidelity and cheating
are typically considered synonymous. Most
participants also indicated that infidelity is
associated with some type of violation including being unfaithful, lying, and dishonest.
Participants also indicated that infidelity was
a morally inappropriate act as a number of
participants identified infidelity as wrong,
bad, and immoral. Participants also described
features of individuals including liar, cheater,
selfish, and untrustworthy. Participants did not
focus on specific behaviors, however, with the
exception that many participants listed sex as a
feature.
Many participants also identified emotions
or emotion-related features such as hurtful,
anger, and sad. In addition, many participants
reported various reasons for infidelity, such as
boredom, excitement, unhappy, lust, lack of
communication, as well as consequences of
infidelity including divorce, heartbreak, and
breakup. It is possible that the participants
intended features such as anger and sad as
causes of infidelity, whereas features such as
unhappy and lack of communication were
intended to be outcomes of infidelity. Due
to the nature of the data, such classifications
by the researchers are somewhat speculative.
Additional work that focuses on what lay
people perceive to be the causes and consequences of infidelity may be a fruitful avenue
to purse although beyond the scope of the
current prototype study. Thus, the features
listed by participants encompass much of the
definitions used by relationship researchers
(i.e., deceptive behaviors), but appear to be
broader and more inclusive as many other
additional features appear to be reflected in the
infidelity prototype.

Study 2: Centrality Rankings of Infidelity


Features
The purpose of Study 2 was to gather more
information about the centrality of the features
generated in Study 1. Specifically, we sought
to determine what features were deemed
to be central and which were considered to
be peripheral. We also sought to establish
whether or not there was agreement on these
judgments.
Method
Participants
The sample included 286 undergraduate students (83 males, 198 females, 5 refused to
answer) recruited from an online social psychology subject pool. Participants ranged in
age from 17 to 59 years (M = 22.20) and were
Caucasian (76.2%), Latino (8.7%), Asian
(4.9%), African American (3.8%), multiethnic
(2.8%), Pacific Islander (1.0%), and Native
American (1.0%). Just over half (54.5%) indicated that they were in a committed romantic
relationship, with 30.1% dating one person
exclusively, 12.2% living with their partner,
4.9% engaged, and 7.3% married. The remainder of the sample indicated that they were
not in a romantic relationship (38.5%), were
dating multiple people (5.6%), or declined to
provide this information (1.4%). Half (50.0%)
of the participants reported that they had been
in a relationship in which their partner committed an infidelity. However, 65.0% reported
that they had never committed a relationship
infidelity.
Procedure
Upon accessing the online survey, participants
were provided with the following instructions:
This questionnaire is part of a project on
the sorts of things we have in mind when
we hear and use words. In this study, we
are interested in your views of the term
INFIDELITY. The term INFIDELITY
means a variety of things to people. In a
previous study people listed a number of
features of the concept of INFIDELITY.
Some of the features that were listed might

Infidelity prototype

be considered more central to the concept


INFIDELITY than others.
In this study, we would like you to rate
how central each feature is of your concept of INFIDELITY. By central, we mean
that the feature is closer to the heart or
essence of what the term INFIDELITY
means to you. For example, consider the
word apple. Close your eyes and imagine an apple. Think about the features that
make it an apple, such as red, sweet, fruit,
juicy, healthy, and so forth. To you, some
of these features might be more central to
the real essence of apple than others. You
might feel that the features of red or juicy
are more central to the essence of apple
than are the features of shape or size.
With each feature of INFIDELITY listed
below, we would like you to rate how
central the feature is in your concept of
INFIDELITY. Please click on the number that corresponds to how central you
believe the feature is to the essence of INFIDELITY. We would like you to think not
only about your own experiences with infidelity but to the concept of INFIDELITY in
generalwhat you think are the defining
features of INFIDELITY. Dont worry
about why you think something is or isnt
centraljust give us your opinion.
Participants then rated the 95 infidelity
features obtained from Study 1 using a scale
ranging from 1 (not at all central) to 8
(extremely central). Participants randomly
received one of three versions of the survey
based on the day of the month they were
born; the versions were identical except for
the order in which the infidelity features were
presented. There was slight concern about
participant fatigue and by counterbalancing
the presentation of items through the three versions we sought to minimize how this potential
fatigue might impact ratings. A multivariate
analysis of variance (MANOVA) was run to
compare centrality ratings across versions
and was significant; Wilkss = .16, F(190,
246) = 1.96, p < .001. In total, there were 17
significant univariate differences although no
discernible pattern was discovered, indicating

663

that counterbalancing should negate any


meaningful version differences.
Results
Mean centrality rankings for the 95 infidelity
features are listed in Table 1; the most central
features of infidelity were unfaithful, cheating,
broken trust, being with someone who is not
your partner, and loss of trust. To assess the
reliability of these means, an interclass correlation coefficient, which is equivalent to the
average of all possible split-half coefficients,
was computed and found to be significant
(ICC = .98, p < .001). In comparing the centrality ratings obtained in this study with
the frequency percentages from Study 1, we
observed that overall the two were positively
correlated (r = .42, p < .001). For instance,
the first and second most highly mentioned
features by Study 1 participants, cheating
(51.6%) and unfaithful (27.1%), received the
highest centrality ratings by Study 2 participants (6.87 and 6.90, respectively). The main
purpose of this study was to determine what
features should be deemed central and which
should be considered to be peripheral to the
concept of infidelity. Following procedures
common in prototype research, we conducted
a median split of the centrality ratings and
concluded that features with a centrality rating of 5.64 or higher should be considered
central and any features below 5.64 would
be considered peripheral. This classification
of features as either central or peripheral was
used in Studies 3 and 4. Although necessary
for the present purposes, we acknowledge that
such a division is artificial and that centrality should be considered a continuum rather
than a dichotomy (e.g., Kearns & Fincham,
2004).
Next, we explored whether centrality ratings differed based on whether participants had
ever been cheated on or not, whether a partner
had ever cheated or not, and gender. An overall
mean centrality score was calculated separately
for central and peripheral features. Mean centrality ratings did not significantly differ based
on either participants own infidelity history or
whether a partner had cheated. T tests revealed
that women, on average, provided higher

664

centrality ratings compared to men for both


central, t(281) = 2.79, ps = .006, 6.37 versus
5.88, and peripheral, t(281) = 2.56, ps = .011,
5.05 versus 4.67, features. A series of analyses of variance (ANOVAs) were then run to
more specifically examine gender differences.
Results indicated that women and men significantly differed on their ratings of 36 of the
95 items; for all of these items, women ranked
the feature as more central than men. These
features included the following: weakness,
boredom, affair, cruel, divorce, hurt, guilt,
heartbreak, insecure, sad, unhappy, pain,
disrespectful, male, player, lack of communication, anger, coward, lonely, jerk, lying,
emotions, intimacy, trust, breakup, unforgivable, unsatisfied, crying, broken, sorrow,
hurtful, loss of trust, tears, lose, secretive, and
being with someone who is not your partner.
Given the number of analyses, a Type I error
may certainly be of concern, so specific univariate results should be cautiously interpreted.
Discussion
Study 2 demonstrated that participants considered some features to be more prototypical
of the concept of infidelity than other features and exhibited a high level of agreement
in these ratings, thereby fulfilling the first
criterion for demonstrating that a concept is
prototypically organized. Furthermore, those
features that were rated as more central in
Study 2 were elicited with greater frequency
in free-response form in Study 1. This finding
adds to the argument that the features generated from Study 1 adequately capture the
structure of the infidelity concept. Also of
note was the finding that women rated both
the central and peripheral features higher than
men and 38% of the individual features as
more central than males. These findings will
be further examined in the General Discussion
section as meaningful gender differences were
also found in the subsequent study.

D. A. Weiser et al.

organized and that the identified structure


affects cognitive functions, such as recall and
recognition memory tasks. This idea is based
on research by Cantor and Mischel (1977,
1979), which demonstrated that the activation
of a prototype causes central features, which
are more closely associated with the prototype,
to be more easily accessible in memory than
peripheral features, which are not as closely
associated with the prototype. As a result,
individuals: (a) have a tendency to recall features that are more closely associated with the
prototype than those that are not as closely
associated with the prototype and (b) show
a bias toward recognizing nonpresented, but
highly related, features of the prototype under
investigation.
Based on this research, and consistent
with other prototype studies (e.g., Fehr, 1988;
Kearns & Fincham, 2004), a single study
was conducted to investigate both recall and
recognition memory of the identified infidelity
prototype features. This study is based on the
assumption that being presented with infidelity attributes during an acquisition phase
activates the prototype of infidelity, which
acts as an organizing principle for processing
subsequent information. As such, participants
should have a difficult time distinguishing
between central features of infidelity that were
presented and those central features that were
not presented but are closely associated with
the prototype that was activated. However,
participants should have less difficulty distinguishing between presented and nonpresented
peripheral features of infidelity, as they are
less closely associated with the concept that
is activated. Therefore, participants who were
exposed to infidelity prototype features were
expected to: (a) correctly recall and recognize
more central features that were presented and
(b) falsely recall and recognize more central
features that were not presented.
Method
Participants

Study 3: Recall and Recognition Memory


for Infidelity Features
The purpose of Study 3 was to provide further evidence that infidelity is prototypically

Potential participants were recruited from


three sociology and criminal justice classes.
Although all willing individuals in the classrooms were allowed to participate in the

Infidelity prototype

study (N = 177), only those 146 students who


reported that they had ever been in a romantic
relationship were included in the analyses. As
infidelity history was considered a potential
variable of interest in subsequent analyses, past
opportunity to engage in or be the recipient of
infidelity was deemed to be a necessary qualification for inclusion. Participants included 62
males and 84 females. Participants ranged in
age from 17 to 58 years (M = 20.0) and were
Caucasian (74.7%), Latino (11.6%), Asian
(4.1%), multiethnic (4.1%), African American
(2.7%), and Pacific Islander (2.7%). Nearly
51% of participants indicated that they were
not currently in a romantic relationship, 6.2%
were dating multiple people, 41.1% were
dating one person exclusively, and 2.1% were
married. Just under half of the participants
(46.9%) reported that they had been in a
relationship in which a partner committed an
infidelity and 31.5% reported that they had
personally committed an infidelity in the past.
Procedure
Following the procedure of Kearns and Fincham (2004), participants viewed a series of
Microsoft PowerPoint slides that were presented with a projector on a screen in the front
of the classroom. The first slide informed participants that they would be shown a series
of slides that would display infidelity characteristics and instructed them to pay attention
to these characteristics because they would be
answering questions about them later in the
study. Each subsequent slide contained one
infidelity statement, with participants viewing
20 slides total (Task 1). Slides were automatically advanced by the computer, so each slide
was presented for precisely 4 s.
The infidelity characteristics that were
presented to participants were 20 randomly
selected central features and 20 randomly presented peripheral features identified in Study
2; although these features were randomly
selected, an effort was made to ensure that the
final set of 40 features was of roughly equal
length (see Table 2 for features used in this
study). Once the 40 features were identified,
they were then inserted into the following
structure: Infidelity is characterized as (by)

665

Table 2. Features presented in Study 3


Central

Peripheral

Version 1 (alphabetized)
Betrayal
Cheating
Dishonest
Disloyal
Disrespectful
Heartbreak
Loss of trust
Mistrust
Physically cheating
Untrustworthy

Alcohol
Children
Crime
Excitement
Family
Hate
Lonely
Moral
Pointless
Slut

Version 2 (alphabetized)
Adultery
Affair
Broken trust
Cheater
Hurtful
Lies
Lying
Sneaking around
Unfaithful
Wrong

Bad individual
Boredom
Cheap
Common
Coward
Dangerous
Females
Love
Males
Rebellion

_________ (e.g., cheating) to create infidelity


statements. These statements were then randomly divided into two groups, such that each
group contained 10 central and 10 peripheral
infidelity features, and placed in a random
order. Therefore, some classrooms were
shown one set of 20 random infidelity statements (Group 1), whereas others received a
completely different set of 20 random infidelity
statements (Group 2). The average central and
peripheral ratings from Study 2 for Group
1 were 6.60 and 4.40, respectively, and for
Group 2 they were 6.55 and 4.57, respectively.
After viewing the 20 infidelity statements,
participants were asked to open a manila envelope that they had been given and take out the
yellow packet (Task 2); this packet contained
an interference task in which participants were
given 4 min to list, in alphabetical order, as
many of the states in the United States and their
capitals as possible. Next, participants were
instructed to take the blue packet out of the
envelope and complete the recall task (Task 3).

666

For this task, participants were given a blank


piece of paper with the following directions at
the top: Please list as many of the infidelity
characteristics presented in the first task as you
can. You have 3 minutes. Finally, participants
were asked to retrieve the white packet and
complete Task 4, the recognition task. For this
task, participants were presented with the set
of 40 infidelity characteristics identified for
Task 1 (just the characteristics were listed);
this set consisted of 20 characteristics that the
participants viewed during Task 1 as well as
the other 20 characteristics that were presented
to the other group of participants. Therefore,
during the recognition task, each participant
was presented with 20 characteristics that he
or she had previously seen and 20 characteristics that he or she had not previously seen.
Participants were instructed to indicate if they
saw the infidelity characteristics during the
first task by circling yes or no next to the
characteristic and then provide demographic
information. Each packet in the envelope contained a cover page so that participants would
not be aware of the purpose of the study or
information about tasks to come.
Results
Recall memory task
For the recall memory task (Task 3), participants were asked to freely list as many
infidelity characteristics as they could. The
first two authors coded each of the participants
responses into one of four categories: central
hits (i.e., central features the participant was
presented with during Task 1), peripheral hits
(i.e., peripheral features the participant was
presented with during Task 1), central false
alarms (i.e., central features the participant was
not presented with during Task 1), and peripheral false alarms (i.e., peripheral features the
participant was not presented with during Task
1). Characteristics that were recalled but were
not part of the infidelity prototype identified
in Study 1 were not included in the coding,
although such instances were infrequent. Interrater reliability, as established using Cohens
Kappa, was high ( = .97), and the few discrepancies between coders were resolved by

D. A. Weiser et al.

the authors. Univariate ANOVAs were first


conducted to explore differences in central
hits, peripheral hits, central false alarms, and
peripheral false alarms for the recall data as
a function of group, gender, and infidelity
history. Results revealed that Group 2 recalled
significantly more central false alarms, .99 versus .41, F(1, 175) = 19.67, p < .001, in comparison to Group 1. In addition, in comparison to
males, females recalled a significantly greater
number of central hits, 3.95 versus 3.33, F(1,
171) = 6.51, p = .012, and peripheral hits, 4.71
versus 4.18, F(1, 171) = 4.97, p = .027. No significant differences were found as a function
of whether participants had ever cheated and
whether participants had ever been cheated on.
Based on this information, the recall data were
analyzed in a series of mixed-model repeated
measure ANOVAs with group (1 or 2) and
gender (male or female) as between-subjects
variables and feature centrality (central or
peripheral) as the within-subjects variable.
Our first prediction for the recall task was
that participants would provide more central
hits than peripheral hits (i.e., correctly recall
more central features that were presented
than peripheral features). Contrary to our
prediction, results indicated that participants
correctly recalled more peripheral features
(M = 4.39) than central features (M = 3.64),
F(1, 142) = 21.34, p < .001, partial 2 = .13.
The main between-subjects effect of gender was also found to be significant, F(1,
142) = 5.18, p = .02, partial 2 = .04, such that
overall women provided a higher number of
total hits (M = 4.25) in comparison to men
(M = 3.78). There were no other significant
main effects or interactions for hits in the
recall task.
Our second prediction for the recall task
was that participants would provide more central false alarms than peripheral false alarms
(i.e., falsely recall more central features that
were not presented than peripheral features
that were not presented). Results supported this
prediction, F(1, 142) = 31.80, p < .001, partial
2 = .18, as participants provided a mean of
.68 central false alarms and .22 peripheral false
alarms. A significant Group False Alarm
interaction qualified this main effect, F(1,
142) = 10.62, p = .001, partial 2 = .07. Group

Infidelity prototype

2 was found to provide significantly more central recall false alarms (M = .97) than Group
1 (M = .40), but did not differ significantly in
the mean number of peripheral false alarms
recalled (Group 1 M = .21; Group 2 M = .24).
A significant between-subjects main effect
for group was also found, F(1, 142) = 6.36,
p = .001, partial 2 = .07, such that Group 2
provided more false alarms (M = .60) than
Group 1 (M = .30). There were no other significant main effects or interactions for false
alarm in the recall task.
Recognition memory task
For the recognition memory task, participants
were asked to indicate (yes/no) which features they previously saw as part of the PowerPoint presentation. Based on the group they
were in, these responses again were coded into
one of four categories: central hits, peripheral hits, central false alarms, and peripheral
false alarms. Univariate ANOVAs were first
conducted to explore differences in central
hits, peripheral hits, central false alarms, and
peripheral false alarms for the recognition data
as a function of group, gender, and infidelity
history. Results revealed that Group 2 recognized significantly more central false alarms,
4.43 versus 3.32, F(1, 175) = 9.36, p = .003,
and peripheral false alarms, 1.31 versus 90,
F(1, 175) = 4.82, p = .029, in comparison to
Group 1. In addition, females recognized significantly more peripheral hits, 7.71 versus
7.09, F(1, 171) = 4.55, p = .034, than males.
Once again, no significant recognition differences were found as a function of whether participants had ever cheated and whether participants had ever been cheated on. Based
on this information, the data were analyzed
in a series of mixed-model repeated measure
ANOVAs with group (1 or 2) and gender (male
or female) as between-subjects variables and
feature centrality (central or peripheral) as the
within-subjects variable.
Our first prediction for the recognition
memory task was that participants would
provide more central hits than peripheral hits
(i.e., correctly recognize more central features
that were presented than peripheral features).
This prediction was supported; results indicated that participants correctly recognized

667

more central features (M = 7.95) than peripheral features (M = 7.36), F(1, 142) = 11.94,
p < .001, partial 2 = .08. A significant Gender
Hits interaction qualified this main effect,
F(1, 142) = 5.30, p = .02, partial 2 = .04, such
that men (M = 7.05) recognized significantly
less peripheral hits than women (M = 7.67).
There were no other significant main effects or
interactions for hits in the recognition task.
Our second prediction for the recognition
task was that participants would provide more
central false alarms than peripheral false
alarms (i.e., falsely recognize more central
features that were not presented than peripheral features that were not presented). Results
supported this prediction, F(1, 142) = 218.62,
p < .001, partial 2 = .61. Participants recognized a mean of 4.00 central false alarms and
1.08 peripheral false alarms. A significant
Group False Alarm interaction qualified
these results, F(1, 142) = 4.46, p = .04, partial
2 = .03. A simple effects analyses revealed
that Group 2 (M = 4.61) provided significantly more central recognition false alarms
than Group 1 (M = 3.39), but did not differ significantly in the average number of
peripheral false alarms recognized (Group 2
M = 1.27; Group 1 M = .89). A significant
between-subjects main effect for group was
also found, F(1, 142) = 8.84, p = .003, partial
2 = .06, such that overall Group 2 recognized
a greater number of false alarms (M = 2.94) in
comparison to Group 1 (M = 2.14). There were
no other significant main effects or interactions
for false alarms in the recognition task.
Finally, to provide additional support for
the prototype structure, correlations were run
between the frequencies of Study 1 and centrality ratings from Study 2 with the memory
indices in this study (see Table 3 for correlations among all measures across studies).
Feature frequency was positively associated
with recall false alarms, recognition hits,
and recognition false, and frequency was
marginally related to recall hits. Centrality
rating was positively associated with both
recall false alarms and recognition false
alarms. There was a trend that centrality rating
was positively related to recognition hits,
although centrality rating was not significantly
related to recall hits. These results suggest that

668

D. A. Weiser et al.

Table 3. Convergence correlations across studies

1. Study 1 frequency
2. Study 2 centrality rating
3. Study 3 recall hits
4. Study 3 recall false alarms
5. Study 3 recognition hits
6. Study 3 recognition false alarms
7. Study 4 narrative frequency
p < .10.

.42**
.27
.77**
.37*
.70**
.57**

.11
.29*
.27
.68**
.11

.25
.58**
.14
.36*

.20
.62**
.62**

.29
.17

.41**

*p < .05. **p < .01.

participants are more likely to falsely recall


and recognize those features most central to
the prototype infidelity, and that they are also
more likely to correctly identify features most
central to the prototype.
Discussion
To confirm that infidelity is prototypically
structured, it is necessary to demonstrate
that the centrality of infidelity features is
related to cognitions. Study 3 provided this
evidence as participants correctly recognized
more central features. In addition, participants
falsely reported more central features than
peripheral features for both the recall and
recognition tasks. Notably, the effect size for
the mean number of central features rather
than peripheral features that were incorrectly
recognized was particularly high (partial
2 = .61). These results suggest that the acquisition phase may have activated the prototype
of infidelity, which then influenced memory
and identification performance.
Contrary to our predictions, we found that
participants correctly recalled more peripheral
than central features. This finding could be
attributable to the fact that the groups of central features tended to contain similar words
(e.g., lies and lying, loss of trust and mistrust,
and cheating and physically cheating) and
participants tended to recall only one word
from these sets of words. In comparison, the
groups of peripheral features tended to contain
words that were more unique (e.g., slut and
alcohol) and, consequently, may have made
a greater impression on participants and been

easier to recall after the interference task. This


result is not entirely surprising, given that
other prototype research has identified this
reverse finding for recall, but not recognition,
memory tasks (e.g., Fehr, 1988; Hassebrauck,
1997). Furthermore, this finding does not
change the fact that the overall results of this
study provide evidence that central features
of infidelity are more salient in memory and,
thus, affect memory tasks.
It is also important to note that the analyses revealed that Group 2 recalled and recognized significantly more central false alarms
than Group 1. Post hoc group analyses do
not lead us to believe that these findings are
attributable to meaningful differences in group
composition (e.g., race and age); rather, we
speculate that these differences may be the
result of slight differences in environmental
conditions (e.g., different classroom environment and time of day) that were present at the
time of data collection. It is also possible that
features presented to Group 2 were more memorable. Given that both groups followed the
same pattern of results (i.e., recalling and recognizing more central false alarms than peripheral false alarms), these results do not lead us
to question our finding that infidelity features
affect cognition. However, Study 4 was conducted to provide further support to our claim
that infidelity is prototypically organized.
Study 4: Generation of Infidelity Narratives
The purpose of Study 4 was to gain additional
evidence to support previous findings that infidelity feature centrality is related to cognition.

Infidelity prototype

Based on the idea that some features are more


central to a prototype than others, we expected
that participants would use more central than
peripheral features when describing a personal
experience with infidelity.
Method
Participants
Participants included 51 undergraduate students (15 males, 36 females) recruited from
an online social psychology subject pool.
Participants ranged in age from 18 to 44 years
(M = 22.5) and were Caucasian (76.5%),
African American (5.9%), Latino (9.8%),
multiethnic (5.9%), and Native American
(2%). Over half of participants (60.8%) indicated that they were currently in a committed
romantic relationship; 45.1% reported that
they were dating one person exclusively, 3.9%
were engaged, and 11.8% were married. All
participants had been involved in a romantic
relationship at some point in their lifetime, and
the remainder of the sample indicated that they
were not currently in a romantic relationship
(37.3%) or were dating multiple people (2.0%).
About half of the participants (52.9%) reported
that they had been in a relationship in which
a partner had committed an infidelity, while
only 29.4% reported that they had a personal
history of committing a relationship infidelity.

669

Group 2, rather than central and peripheral.


The coders were then instructed to read each
narrative and list all features used in each
narrative, and whether these features were
included in Group 1 or Group 2. Coders
were instructed to include features that were
conceptually similar (i.e., heartbroken was
coded as heartbreak). This coding procedure
therefore provides frequencies for central and
peripheral features, as well as a frequency for
each feature of the prototype.
There was a 54.9% agreement rate between
coders, and moderate interrater reliability was
observed with a Kappa value of .56. Due to
the large number of features and that coders
were not limited in the number of codes per
narrative, a high level of agreement between
coders was not expected. Such an agreement
level is similar to the agreement rate found by
Lambert et al. (2009) for their narrative study
within their prototype analysis of gratitude.
Most disagreements were due to one coder
catching a feature the other missed. Discrepancies were easily resolved by the first two
authors. A total of 125 central features and
64 peripheral features were generated from the
narratives. For example, among all narratives
the central feature breakup was listed 21 times,
whereas the peripheral feature unhappy was
listed only once.
Results

Procedure
Upon accessing the online survey, participants
were provided with the following instructions
(adapted from Lambert et al., 2009): Please
think back to a time when you, a friend, or
a family member experienced infidelity in a
romantic relationship. Please write a paragraph
about the experience. After generating a narrative about their experiences with infidelity,
participants provided demographic information and were thanked for their time.
Two coders blind to the study hypotheses
coded the 51 narratives. They were each given
a list of the 95 infidelity features generated
from Study 1 with the features divided into
two groups (central and peripheral). In order
to keep the coders blind to the purpose of the
study, the groups were labeled Group 1 and

Overall, it was found that central features were


more likely to be present compared to peripheral features, t(50) = 4.60, p < .001. Univariate
ANOVAs were first conducted to explore
differences in use of central and peripheral
features as a function of gender and infidelity
history. No significant differences were found
for gender or whether participants had ever
been cheated on by a romantic partner. Participants who had ever cheated themselves,
however, reported a significantly higher
number of peripheral features in their narratives compared to individuals who had never
cheated, 1.87 versus 1.00, F(1, 49) = 5.75,
p = .020. Based on this finding, whether participants had ever cheated or not was included
as a between-subjects factor in a mixed-model
repeated measures ANOVA comparing the

670

frequency of use of central and peripheral


features in the narratives. Consistent with
expectations, participants used significantly
more central features (M = 2.45) in their narratives than peripheral features (M = 1.25), F(1,
49) = 14.04, p < .001, partial 2 = .22. No other
main effects or interactions were significant.
Finally, we also explored whether feature frequency from Study 1 and the centrality ratings
from Study 2 were correlated with frequency
of prototype feature use in the narratives (see
Table 3). Feature frequency from Study 1
was positively associated with Study 4 narrative frequency, although prototype centrality
was unrelated to narrative frequency. Results
indicate that centrality score is unrelated to
how often a particular feature was utilized,
although as a whole participants were more
likely to use central features in their narratives.
Discussion
Once more, results confirmed that feature centrality was related to participants cognitions
about infidelity. As predicted, when asked to
generate narratives about a time when they,
a friend, or family member experienced infidelity, participants utilized more central than
peripheral features to convey these events.
These findings provide further evidence that
infidelity has a prototype structure that is accurately captured by this research. Notably, participants not only described the actual events
(i.e., kissing, sexual intercourse) but also many
indicated how the infidelity was discovered, the
emotions experienced as a result of the infidelity, and how the relationship was impacted
as a result of the infidelity. It makes sense
that as participants were asked to share an
account of infidelity, they included a description of the actual infidelity behavior as this
may be considered a major component of an
event. Although participants in the previous
studies did not list many behavioral features,
and only sex was a central feature, the demands
of the task at hand necessitated the participants
provide some information about the infidelity
behavior. These findings further signify that
laypeople view infidelity in a broader manner
than researchers and that infidelity is prototypically organized.

D. A. Weiser et al.

In addition, although personal infidelity history was not significant in the main analysis,
the preliminary univariate ANOVA found that
individuals who had ever cheated used significantly more peripheral features to describe the
infidelity experience compared to those who
have never cheated. As participants were able
to write about theirs or anothers infidelity
experience, it is difficult to assess what this
difference truly means. Participants appear to
have avoided writing about their own experiences as a cheater, as one third of participants
indicated they had cheated in a relationship,
yet only four narratives were about the participants own infidelity. As direct experience
and frequency of experience influences the
development of mental models (Fletcher &
Thomas, 1996), it is possible that individuals
who cheated talk and think about infidelity
differently. However, such a conclusion is
beyond the scope of this study.
General Discussion
Together, these four studies demonstrate
that infidelity does indeed have a prototype
structure as participants were able to make
consistent and meaningful distinctions about
which features were central or peripheral to
the infidelity construct. In Study 1, participants
were able to generate a list of infidelity features
and in Study 2 participants were able to reliably rate the centrality of these features. Thus,
the findings of Studies 1 and 2 provide support
for the first condition necessary to demonstrate
a prototype structure (Rosch, 1975). Studies 3
and 4 also indicated that feature centrality was
related to how individuals thought about and
processed infidelity information. This finding
provides support for the second condition
needed to demonstrate a prototype structure
(Rosch, 1975). Specifically, in Study 3 participants correctly recognized a greater number
of central features than peripheral features,
and had a greater number of central feature
false alarms in both the recall and recognition
memory tasks. In Study 4, participants generated narratives about a time they, a friend,
or a family member experienced infidelity in
a romantic relationship. Within these narratives, participants used a significantly greater

Infidelity prototype

number of central features than peripheral features to describe these infidelity experiences.
Therefore, these studies provide substantive evidence that infidelity is prototypically
organized.
Components of the infidelity prototype
Overall, this prototype approach has illuminated a great deal about how laypeople
conceptualize infidelity. Our results suggest
that laypeople perceived infidelity in a much
more complex, broad, and multifaceted manner
than typically defined by researchers, although
there were a number of similarities between
laypeople and researchers conceptualizations
as well. We propose that five broad themes
emerged: violation, secretiveness, immorality,
consequences, and emotional outcomes. First,
participants reported unfaithful, cheating, and
affair as central characteristics of infidelity,
which capture the violating nature of infidelity.
Interestingly, all of these central features are
terms that researchers have previously used
to describe infidelity (i.e., Atkins, Dimidjian,
et al., 2001; Brand et al., 2007; Roscoe et al.,
1988). In some regards, these features may be
viewed as synonyms for infidelity although
we argue that such characteristics belong as
features of the infidelity construct. Affair, for
example, begins to get at that composite type
of infidelity in which there is a prolonged
emotional and sexual violation. In contrast,
unfaithful appears related to a deceptive,
violating component of infidelity.
Second, participants indicated dishonest,
lying, sneaking around, deceive, and secretive are central components of the infidelity
concept. Such features are consistent with
researchers definitions of infidelity that have
emphasized concealment of behaviors, lying,
deceit, and secretiveness (i.e., Glass, 2002;
Pittman & Wagers, 2005). Third, participants overwhelmingly used moral language to
describe infidelity, including wrong, betrayal,
disrespectful, selfish, bad, and immoral, indicating that most individuals hold unfavorable
views of infidelity. In a study of individuals
in the United States, the vast majority of
individuals expect sexual exclusivity from
their romantic relationships (Treas & Giesen,

671

2000); therefore, it is consistent with expectations that an infidelity prototype should


contain numerous features that capture this
disapproval for infidelity. It is particularly
interesting that laypeople have little qualms
about making strong, moral judgments about
the concept of infidelity. Within the infidelity
literature, researchers have attempted to be
somewhat neutral about how they conceptualize infidelity (e.g., some researchers opt for
the term extradyadic activity to avoid moral
implications). Overall, laypeople clearly view
infidelity as wrong, perhaps tying into multiple
domains of morality, including dimensions
of loyalty and fairness, harm prevention, and
sanctity (Haidt, 2013).
Fourth, participants list a number of consequences of infidelity including breakup,
broken trust, unforgiveable, and loss of trust.
Fifth, participants list a number of highly emotional outcomes related to infidelity including
pain, crying, sad, hurt, guilt, and heartbreak.
Researchers have commonly studied both consequences and emotional reactions as outcome
variables of infidelity (i.e., Becker et al., 2004;
Cano & OLeary, 2000; Hall & Fincham,
2006). In contrast to researchers, it appears
that laypeople actually view these outcomes of
infidelity as central components of infidelity.
Essentially, these results begin to suggest that
laypeople do not separate the negative consequences of infidelity from the construct of
infidelity. Further work such as script analysis
may shed light on how these features fit into
a timeline of infidelity and which features are
most strongly identified as causes and which
as consequences of infidelity.
Finally, perhaps the most surprising finding
to emerge from the prototype approach is the
fact that, in contrast to researchers, laypeople
did not include a number of behaviors as
defining infidelity. In fact, sex and kissing
were the only specific behaviors mentioned
by participants, and sex was the only behavior
that was central to the infidelity concept. The
one exception to this pattern is that participants tended to include infidelity behaviors in
their narratives. This likely occurred because
participants viewed these actions as an important component when telling their infidelity
story. Thus, while particular behaviors are

672

an essential component when discussing infidelity, especially in particular tasks, such


behaviors are not the sole defining feature of
the infidelity construct according to laypeople.
Instead, laypeople appear to conceptualize
infidelity in a much broader, holistic manner
than merely behaviors.
Role of gender and infidelity history
It is also important to note the meaningful
gender differences that emerged in Studies 2
and 3. In Study 2, women were more likely to
rate a number of negatively valenced features
as more central to the infidelity construct.
These findings are not surprising as a great
deal of research has found gender differences
when investigating infidelity. For instance,
researchers have found that women classify
a greater array of behaviors (e.g., forming a
romantic attachment with another individual,
flirting) as infidelity compared to men (Brand
et al., 2007; Yarab et al., 1999). Furthermore,
men tend to view extradyadic activities more
favorably, express greater desire to engage
in infidelity, and are more likely to actually
engage in infidelity (Atkins, Dimidjian, et al.,
2001; Oliver & Hyde, 1993; Petersen & Hyde,
2010; Seal, Agostinelli, & Hannett, 1994). It
is therefore consistent with expectations that
women would consider a greater number of
negatively valenced words (i.e., cruel, hurt,
heartbreak, insecure, sad, unhappy, pain,
disrespectful, jerk, and lying) as more central
to the prototype of infidelity.
In Study 3, women accurately reported
more central and peripheral hits during the
recall task and more peripheral hits during the
recognition task. Taken together, these results
suggest that women not only consider more
negative features as central to the prototype of
infidelity but they also may have a more complex and easily activated conceptualization of
infidelity compared to men. Women are typically socialized to place greater value on social
cohesion and relationships, whereas men are
typically socialized to be more independent
and autonomous (Crawford & Unger, 2004).
Because of this different socialization, it is
likely that men and women will form different
mental models of relationships (Acitelli &

D. A. Weiser et al.

Young, 1996), with women possessing more


complex relationships beliefs (and strong,
negative beliefs about infidelity). Strong relationship beliefs may be chronically accessible
so that mental models are more easily activated and more frequently used to process
perceptions (Fletcher & Fitness, 1993). In
addition, the more constructs involved in an
individuals belief, the more easily activated
the mental model will be and it will have a
greater opportunity to color perceptions (Higgins, 1996). The findings of this study suggest
that womens infidelity mental models were
more easily activated and they were thus able
to remember features more accurately.
These results suggest that infidelity is a
construct that is viewed by the genders differently. This is a departure from other prototype analyses of relational qualities that do
not find gender differences. For example, Fehr
(1988) found no differences between mens
and womens prototypes of love, which is consistent with the love literature that finds that
overall men and women view love similarly
(Fehr & Broughton, 2001; Hendrick & Hendrick, 1995). The findings of this study suggest that men and women may not just differ
in their acceptance of infidelity, as established
by the infidelity literature, but potentially differ in how they conceptualize infidelity altogether. This may be due to the fact that women
are less likely to engage in infidelity (Atkins,
Dimidjian, et al., 2001) or that relationships
are more likely to end due to a womans infidelity (Brand et al., 2007). Thus, while men
and women may experience concepts like love,
commitment, and forgiveness in a similar manner, there is support for the idea that women
conceptualize, and possibly experience, infidelity differently. These differences, whether
through direct experience or broader socialization, may then shape a uniquely female view of
infidelity.
Infidelity history only emerged as a significant individual difference in Study 4, as
individuals who had cheated used significantly
more peripheral features to describe the infidelity experience. Future work may explore
whether individuals prompted to write about
their own infidelity experience avoid using
the moral language, which was identified as

Infidelity prototype

central to the infidelity prototype, and instead


use peripheral features to focus on justifications for their actions. The previous
literature finds individuals accounts of
ones own transgressions tend to utilize a
number of self-serving biases and individuals tend to downplay their problematic
behaviors (Baumeister, Stillwell, & Wotman,
1990; Kearns & Fincham, 2005; Stillwell &
Baumeister, 1997). An interesting avenue for
future work will be to further explore whether
individuals discuss and evaluate infidelity differently after engaging in infidelity. Another
fruitful avenue for researchers will be to see
if individuals conceptualizations of infidelity
are predictive of infidelity behavior. Certain
individuals may be predisposed to viewing
morality as less central to the infidelity construct, and these individuals may be more
likely to engage in infidelity.
Limitations and conclusions
As with any research, some limitations should
be noted. All four studies utilized convenience
sampling, which ultimately limits the generalizability of results. Participants were mainly
young, unmarried, undergraduate students and
their infidelity prototypes may differ from
married, older, or less educated individuals as
age, marital status, and education level have all
been linked with infidelity behavior (Atkins,
Baucom, & Jacobsen, 2001; Treas & Giesen,
2000). In addition, it should be noted that
in all studies females were overrepresented.
Although we do not believe this gender composition changes the meaning of our findings,
there is a rich avenue of research to be pursued
on how men and women conceptualize and
experience infidelity.
In conclusion, although many researchers
have argued for greater consistency and clarity
when defining infidelity (e.g., Blow & Hartnett, 2005b; Tsapelas, Fisher, & Aron, 2011),
the current studies illuminate the challenges in
reaching this goal. Infidelity is a complex construct that cannot be easily broken down into
categories of infidelity or not infidelity. Our
findings reveal individuals conceptualize infidelity as a broad, multifaceted construct that
is characterized by violations, secretiveness,

673

and immorality with serious consequences and


emotional outcomes. These studies indicate
that infidelity does have a prototype structure
and that laypeople do indeed conceptualize
infidelity in a broader manner than researchers.
As a result, researchers may need to re-evaluate
how they operationalize and measure infidelity
in their work. Simply asking participants
about their behaviors may not provide the best
assessment of infidelity, as lay people view
infidelity as possessing qualities besides just
the presence of certain behaviors. Finally, new
avenues of research may be pursued to understand how individuals conceptualizations of
infidelity relate to actual infidelity behavior.
Perhaps the most meaningful contribution of
any infidelity research is understanding why
some individuals have a greater likelihood
of committing infidelity. By illuminating this
mechanism, it may be possible for clinicians
and relationship educators to reduce the occurrence of infidelity, or at least allow couples to
better negotiate the infidelity experience.

References
Acitelli, L. K., & Young, A. M. (1996). Gender and thought
in relationships. In G. J. O. Fletcher & J. Fitness
(Eds.), Knowledge structures in close relationships
(pp. 147168). Mahwah, NJ: Erlbaum.
Allen, E. S., Atkins, D. C., Baucom, D. H., Snyder, D.
K., Gordon, K. C., & Glass, S. P. (2005). Intrapersonal, interpersonal, and contextual factors in engaging
in and responding to extramarital involvement. Clinical Psychology: Science and Practice, 12, 101130.
doi:10.1093/clipsy/bpi014
Allen, E. S., & Baucom, D. H. (2006). Dating, marital,
and hypothetical extradyadic involvement: How do
they compare? Journal of Sex Research, 43, 307317.
doi:10.1080/00224490609552330
Amato, P. R., & Previti, D. (2003). Peoples reasons for
divorcing: Gender, social class, the life course, and
adjustment. Journal of Family Issues, 24, 602626.
doi:10.1177/0192513X03024005002
Atkins, D. C., Baucom, D. H., & Jacobson, N. S. (2001).
Understanding infidelity: Correlates in a national
random sample. Journal of Family Psychology, 15,
735749. doi:10.1037/0893-3200.15.4.735
Atkins, D. C., Dimidjian, S., & Jacobsen, N. S. (2001).
Why do people have affairs?: Recent research and
future directions about attributions for extramarital
involvement. In V. Manusov & J. H. Harvey (Eds.),
Attribution, communication behavior, and close relationships (pp. 305319). Cambridge, England: Cambridge University Press.

674
Baumeister, R. F., Stillwell, A., & Wotman, S. R. (1990).
Victim and perpetrator accounts of interpersonal conflict: Autobiographical narratives about anger. Journal
of Personality and Social Psychology, 59, 9941005.
doi:10.1037/0022-3514.59.5.994
Becker, D. V., Sagarin, B. J., Guadagno, R. E., Millevoi,
A., & Nicastle, L. D. (2004). When the sexes need
not differ: Emotional responses to the sexual and emotional aspects of infidelity. Personal Relationships, 11,
529538. doi:10.1111/j.1475-6811.2004.00096.x
Bernard, J. (1974). Infidelity: Some moral and social
issues. In J. R. Smith & L. G. Smith (Eds.), Beyond
monogamy: Recent studies of sexual alternatives in
marriage (pp. 130158). Baltimore, MD: Johns Hopkins University Press.
Birnie-Porter, C., & Lydon, J. E. (2013). A prototype
approach to understanding sexual intimacy through its
relationship to intimacy. Personal Relationships, 20,
236258. doi:10.1111/j.1475-6811.2012.01402.x
Blow, A. J., & Hartnett, K. (2005a). Infidelity in committed relationships II: A substantive review. Journal of Marital and Family Therapy, 31, 217233.
doi:10.1111/j.1752-0606.2005.tb01556.x
Blow, A. J., & Hartnett, K. (2005b). Infidelity in committed relationships I: A methodological review. Journal of Marital and Family Therapy, 31, 183216.
doi:10.1111/j.1752-0606.2005.tb01555.x
Brand, R. J., Markey, C. M., Mills, A., & Hodges,
S. D. (2007). Sex differences in self-reported infidelity and its correlates. Sex Roles, 57, 101109.
doi:10.1007/s11199-007-9221-5
Buunk, B. P. (1995). Sex, self-esteem, dependency, and
extradyadic sexual experience as related to jealousy
responses. Journal of Social and Personal Relationships, 12, 147153. doi:10.1177/0265407595121011
Cano, A., & OLeary, K. D. (2000). Infidelity and separations precipitate major depressive episodes and symptoms of nonspecific depression and anxiety. Journal
of Consulting and Clinical Psychology, 68, 774781.
doi:10.1037/0022-006X.68.5.774
Cantor, N., & Mischel, W. (1977). Traits as prototypes: Effects on recognition memory. Journal of
Personality and Social Psychology, 35, 3848.
doi:10.1037/0022-3514.35.1.38
Cantor, N., & Mischel, W. (1979). Prototypicality and
personality: Effects on free recall and personality
impressions. Journal of Research in Personality, 13,
187205. doi:10.1016/0092-6566(79)90030-8
Crawford, M., & Unger, R. (2004). Women and gender: A feminist psychology (4th ed.). New York, NY:
McGraw-Hill.
Fehr, B. (1988). Prototype analysis of the concepts
of love and commitment. Journal of Personality and Social Psychology, 55, 557579.
doi:10.1037/0022-3514.55.4.557
Fehr, B., & Broughton, R. (2001). Gender and personality
differences in conceptions of love: An interpersonal
theory analysis. Personal Relationships, 8, 115136.
doi:10.1111/j.1475-6811.2001.tb00031.x
Fehr, B., & Russell, J. A. (1991). The concept of love
viewed from a prototype perspective. Journal of

D. A. Weiser et al.
Personality and Social Psychology, 60, 425438.
doi:10.1037/0022-3514.60.3.425
Feldman, S. S., & Cauffman, E. (1999). Your
cheatin heart: Attitudes, behaviors, and correlates of sexual betrayal in late adolescents.
Journal of Research on Adolescence, 9, 227252.
doi:10.1207/s15327795jra0903_1
Fletcher, G. J. O., & Fitness, J. (1993). Knowledge
structures and explanations in intimate relationships.
In S. Duck (Ed.), Individuals in relationships (pp.
121143). Newbury Park, CA: Sage.
Fletcher, G. J. O., & Thomas, G. (1996). Close relationship
lay theories. Their structure and function. In G. J. O.
Fletcher & J. Fitness (Eds.), Knowledge structures in
close relationships (pp. 324). Mahwah, NJ: Erlbaum.
Frei, J. R., & Shaver, P. R. (2002). Respect in close relationships: Prototype definition, self-report assessment, and
initial correlates. Personal Relationships, 9, 121139.
doi:10.1111/1475-6811.00008
Glass, S. P. (2002). Couple therapy after the trauma of infidelity. In A. S. Gurman & N. S. Jacobson (Eds.), Clinical handbook of couple therapy (3rd ed., pp. 488507).
New York, NY: Guilford Press.
Gregg, A. P., Hart, C. M., Sedikides, C., & Kumashiro, M.
(2008). Everyday conceptions of modesty: A prototype
analysis. Personality and Social Psychology Bulletin,
34, 978992. doi:10.1177/0146167208316734
Haidt, J. (2013). The righteous mind: Why good people
are divided by politics and religion. New York, NY:
Random House Digital.
Hall, J. H., & Fincham, F. D. (2006). Relationship dissolution following infidelity. In M. A. Fine & J. H. Harvey
(Eds.), Handbook of divorce and relationship dissolution (pp. 153168). Mahwah, NJ: Erlbaum.
Harasymchuk, C., & Fehr, B. (2013). A prototype analysis of relational boredom. Journal of
Social and Personal Relationships, 30, 627646.
doi:10.1177/0265407512464483
Hassebrauck, M. (1997). Cognitions of relationship
quality: A prototype analysis of their structure and
consequences. Personal Relationships, 4, 163185.
doi:10.1111/j.1475-6811.1997.tb00137.x
Hendrick, S. S., & Hendrick, C. (1995). Gender
differences and similarities in sex and love.
Personal Relationships, 2, 5565. doi:10.1111/j.14756811.1995.tb00077.x
Higgins, E. T. (1996). Knowledge activation: Accessibility,
applicability, and salience. In A. Kruglanski & E. T.
Higgins (Eds.), Social psychology: Handbook of basic
principles (pp. 133168). New York, NY: Guilford
Press.
Kearns, J. N., & Fincham, F. D. (2004). A prototype
analysis of forgiveness. Personality and Social
Psychology Bulletin, 7, 838855. doi:10.1177/01461
67204264237
Kearns, J. N., & Fincham, F. D. (2005). Victim and
perpetrator accounts of interpersonal transgressions:
Self-serving or relationship-serving biases? Personality and Social Psychology Bulletin, 31, 321333.
doi:10.1177/0146167204271594

Infidelity prototype
Kruger, D. J., Fisher, M. L., Edelstein, R. S., Chopik, W.
J., Fitzgerald, C. J., & Stout, S. L. (2013). Was that
cheating? Perceptions vary by sex, attachment anxiety,
and behavior. Evolutionary Psychology, 11, 159171.
Lambert, N. M., Graham, S. M., & Fincham, F. D.
(2009). A prototype analysis of gratitude: Varieties of gratitude experiences. Personality and
Social Psychology Bulletin, 35, 11931207.
doi:10.1177/0146167209338071
Luo, S., Cartun, M. A., & Snider, A. G. (2010). Assessing
extradyadic behavior: A review, a new measure,
and two new models. Personality and Individual
Differences, 49, 155163. doi:10.1016/j.paid.2010.
03.033
Oliver, M. B., & Hyde, J. S. (1993). Gender differences
in sexuality: A meta-analysis. Psychological Bulletin,
114, 2951. doi:10.1037/0033-2909.114.1.29
Peluso, P. R., & Spina, P. (2008). Understanding
infidelity: Pitfalls and lessons for couples counselors. The Family Journal: Counseling and
Therapy for Couples and Families, 16, 324327.
doi:10.1177/1066480708323282
Petersen, J. L., & Hyde, J. S. (2010). A meta-analytic
review of research on gender differences in sexuality, 1993-2007. Psychological Bulletin, 136, 2138.
doi:10.1037/a0017504
Pittman, F. S., & Wagers, T. P. (2005). The relationship, if any, between marriage and infidelity. Journal of Couple & Relationship Therapy, 4, 135148.
doi:10.1300/J398v04n0212
Regan, P. C., Kocan, E. R., & Whitlock, T. (1998).
Aint love grand! A prototype analysis of the
concept of romantic love. Journal of Social
and Personal Relationships, 15, 411420.
doi:10.1177/0265407598153006
Rosch, E. (1975). Cognitive representations of
semantic categories. Journal of Experimental
Psychology:
General,
104,
192233.
doi:10.1037/0096-3445.104.3.192
Rosch, E. (1978). Principles of categorization. In E. Rosch
& B. B. Lloyd (Eds.), Cognition and categorization
(pp. 2748). Hillsdale, NJ: Erlbaum.
Roscoe, B., Cavanaugh, L. E., & Kennedy, D. R.
(1988). Dating infidelity: Behaviors, reasons, and
consequences. Adolescence, 23, 3543.
Seal, D. W., Agostinelli, G., & Hannett, C. A. (1994).
Extradyadic romantic involvement: Moderating effects
of sociosexuality and gender. Sex Roles, 31, 122.
doi:10.1007/BF01560274
Sharpsteen, D. J. (1993). Romantic jealousy as an
emotion concept: A prototype analysis. Journal
of Social and Personal Relationships, 10, 6982.
doi:10.1177/0265407593101005

675
Sheppard, V. J., Nelson, E. S., & Andreoli-Mathie, V.
(1995). Dating relationships and infidelity: Attitudes
and behaviors. Journal of Sex & Marital Therapy, 21,
202212. doi:10.1080/00926239508404399
Stillwell, A. M., & Baumeister, R. F. (1997). The
construction of victim and perpetrator memories:
Accuracy and distortion in role-based accounts.
Personality and Social Psychology Bulletin, 23,
11571172. doi:10.1037/0022-3514.82.4.675
Surra, C. A., & Bohman, T. (1991). The development of
close relationships: A cognitive perspective. In G. J.
O. Fletcher & F. D. Fincham (Eds.), Cognition in close
relationships (pp. 281305). Hillsdale, NJ: Erlbaum.
Thompson, A. P. (1983). Extramarital sex: A review of the
research literature. Journal of Sex Research, 19, 122.
doi:10.1080/00224498309551166
Treas, J., & Giesen, D. (2000). Sexual infidelity
among married and cohabitating Americans.
Journal of Marriage and the Family, 62, 4860.
doi:10.1111/j.1741-3737.2000.00048.x
Tsapelas, I., Fisher, H. E., & Aron, A. (2011). Infidelity: When, where, why. In W. R. Cupach & B. H.
Spitzberg (Eds.), The dark side of close relationships
II (pp. 175195). New York, NY: Routledge/Taylor &
Francis.
Weigel, D. J. (2008). The concept of family: An analysis of
laypeoples views of family. Journal of Family Issues,
29, 14261447. doi:10.1177/0192513X08318488
Whisman, M. A., Dixon, A. E., & Johnson, B.
(1997). Therapists perspectives of couple problems and treatment issues in couple therapy.
Journal of Family Psychology, 11, 361366.
doi:10.1037/0893-3200.11.3.361
Wiederman, M. W., & Allgeier, E. R. (1996). Expectations and attributions regarding extramarital
sex among young married individuals. Journal
of Psychology & Human Sexuality, 8, 2135.
doi:10.1300/J056v08n03_02
Wiederman, M. W., & Hurd, C. (1999). Extradyadic
involvement during dating. Journal of Social
and Personal Relationships, 16, 265274.
doi:10.1177/0265407599162008
Wilson, K., Mattingly, B. A., Clark, E. M., Weidler,
D. J., & Bequette, A. W. (2011). The gray area:
Exploring attitudes towards infidelity and the development of the perceptions of dating infidelity
scale. Journal of Social Psychology, 151, 6386.
doi:10.1080/00224540903366750
Yarab, P. E., Allgeier, E. R., & Sensibaugh, C. C. (1999).
Looking deeper: Extradyadic behaviors, jealousy,
and perceived unfaithfulness in hypothetical dating
relationships. Personal Relationships, 6, 305316.
doi:10.1111/j.1475-6811.1999.tb00194.x

S-ar putea să vă placă și