Summary of:
Penke, L. & Asendorpf, J.B. (2008). Evidence for Conditional Sex Differences in Emotional but
Not in Sexual Jealousy at the Automatic Level of Cognitive Processing. European Journal of
Personality, 22:3-30.
Summary by Haley Andrews, Elizabeth Haralambos, and Andrea Hollingshead
For Dr. Mill’s Psyc 310 class, Spring 2012
As we have discussed in class, evolutionary psychological hypotheses suggest that in the
case of infidelity, women tend to be more jealous of emotional infidelity, whereas men tend to be
more jealous of sexual infidelity. This article reviews two evolutionary hypotheses regarding
jealousy towards infidelity, and proposes a further study to combat some of the problems with
previous studies on this issue. The authors suggest that inconsistencies lie in the results of
previous findings, and they can be resolved when the two hypotheses are evaluated separately
and when the underlying cognitive processes are considered.
‘Jealousy’ is a concept in many cultures that, in its broadest meaning, describes affective
responses to a real or imagined situation where a personally highly valued possession is
threatened to be lost to someone else (e.g. Brochhaus-Enxyklopadie, 1996; Merriam-Webster’s
Collegiate Dictionary, 1993). According to Penke and Asendorpf, sex differences in romantic
jealousy have been hotly debated, arguably because they serve as a prominent testing ground for
an evolutionary psychological approach to sex differences in social cognition and emotion
(Penke & Asendorpf, 2008; e.g. Harris, 2003, 2005; Sagarin, 2005). Penke and Asendorpf
summarize the classic evolutionary hypothesis: it assumes that men and women react differently
to sexual and emotional infidelity because these two types of infidelity posed different adaptive
problems. Men are expected to be more jealous to sexual infidelity that women in order to
minimize investment in genetically unrelated offspring. Women, on the other hand, are expected
to be more jealous to emotional infidelity that men in order to minimize loss of paternal
investment in their offspring (Penke & Asendorpf, 2008). In a critique of a previous study by
Buss et al., these are sex-typical adaptive problems faced by all mammalian species with internal
fertilization and biparental care and Buss et al have assumed that humans have evolved
psychological mechanisms that sensitize men to cues of their mate’s sexual infidelity and women
to cues of their mate’s emotional infidelity. This focuses on domain specific cognitive modules,
in which Buss proposed sex-specific cognitive modules that solved these sex-specific adaptive
problems, also known as the “jealousy as a specific ‘innate’ module hypothesis” by Harris, 2000.
In the case of this hypothesis, this evolutionary account consists of two independently derived
hypotheses, one being sexual infidelity and the other emotional infidelity, and that none of them,
one of them, or both of them may be true.
The contrasting evolutionary theory by Harris (2000, 2003) proposed a ‘social-cognitive
theory of jealousy’, where jealousy is the result of an evolved, but domain-general appraisal
mechanism (Penke &Asendorpf, 2008). In this module, Harris argues that this mechanism is
sensitive to all kinds of threats posed by rivals. In other words, a mate’s sexual or emotional
infidelity poses a threat to a relationship, and therefore another variable such as culturally
determined gender roles may play a part.
The authors go in depth about prior research, offering all possible limitations and
interactions that may have an effect on the significance of previous findings to give way to their
own study. To test the two evolutionary hypotheses Buss et al., as well as other subsequent
studies, used a forced-choice paradigm. In these studies, participants are presented with a
hypothetical dilemmas that contrast sexual and emotional infidelity and are forced to choose the
more distressing alternative. In this study, there were four sub hypotheses of a sex by type of
infidelity interaction: (a) men react more jealous to sexual infidelity that women, (b) women
react more jealous to emotional infidelity that men, (c) men react more jealous to sexual
infidelity that to emotional infidelity, and (d) women react more jealous to emotional infidelity
than to sexual infidelity. According to Harris, a-d should be expected from the specific ‘innate’
model perspective of sexual and emotional jealousy, the two sex differences (a) and (b), and the
two within sex differences (c) and (d) (Penke & Asendorpf, 2008). The problem with this study,
according to Penke and Asendorpf, is that the forced choice paradigm only tests whether one of
the two evolutionary hypotheses or both are supported, not whether both are confirmed. Penke
and Asendorpf propose to reconstruct this study for four reasons: (1) there are two different
evolutionary hypotheses, (2) confirming both requires a crossover sex type of infidelity
interaction, (3) ordinal sex by type of infidelity interactions violate at least on of the two
evolutionary hypotheses and (4) sex differences in the forced choice paradigm may be due to
particularly strong male sexual jealousy, to particularly strong female emotional jealousy, or
both. Thus an evaluation of the two hypotheses requires additional tests that evaluate the sex
difference for sexual jealousy and the sex difference for emotional jealousy separately (Penke &
Asendorpf, 2008). Additionally, the evolutionary approach has empirically falsifiable
hypotheses, whereas the social-cognitive theory does not predict any specific between-sex or
within-sex differences in emotional and sexual jealousy.
As far as empirical evidence goes, the forced-choice paradigm in the Harris study
confounds sex differences in sexual and emotional jealousy preventing strong conclusions
because most studies had only two dilemmas to choose from and it gives no information about
the cognitive processes that underlie the sex differences in choices. As an alternative to the
forced-choice paradigm, other studies used the separate evaluation of sexual and emotional
infidelity to make sure that jealousy was not confounded. Using negative emotional ratings to
evaluate sex differences in sexual and emotional infidelity, however, failed to confirm the
expected sex by type of infidelity interaction. Few other studies looked at the cognitive
processing of infidelity cues, in which a reaction time was measured between the presentation of
the forced-choice dilemma and the decision. These modules were seemingly consisted with
evolutionary hypotheses for emotional jealousy, but not for sexual jealous. Thus, the
insignificant findings coupled with the small amount of research on cognitive processes gives
way to Penke and Asendorpf’s present study.
The aim of their study was to test the two evolutionary hypotheses with both the forcedchoice method and continuous emotion ratings for many dilemmas in a sample of sexually
experienced young adults. To approach the underlying cognitive processes, Penke and Asendorpf
studied the automatic of the responses in the forced-choice task in two conditions: cognitive
constraint condition and deliberation condition. In the cognitive constraint condition, participants
were instructed to rely on their spontaneous preferences, while simultaneously remembers a
sequence of numbers. In the deliberation condition, participants were instructed to take their
time, vividly imagining scenarios as to make a careful decision. In this case decisions, emotion
ratings, and response times were all recorded. This study was thus designed to evaluate previous
studies’ designs separately with respect to the two evolutionary hypotheses when processing
infidelity cues (Penke &Asendorpf, 2008). Additionally, three tasks were presented in the same
order for all participants to make a comparison of results for different methods possible within
the same sample. First, participants completed the forced-choice under cognitive constraint
(strongest influence). Second, they completed emotion ratings with vivid imagination
(intermediate influence). And third, they completed forced-choice with deliberation and vivid
imagination (weakest influence). Penke and Asendorpf expected (1) the classic sex difference in
the forced-choice paradigm under both conditions, particularly under cognitive constraint, (2)
that women rate emotional infidelity situations more negatively than men, but that men do not
rate sexual infidelity situations more negatively than women and (3) that women process
emotional infidelity cues faster than men, but that men do not process sexual infidelity cues
faster than women (Penke & Asendorpf. 2008). The authors also looked at extraneous factors of
romantic relationship status and education on these effects.
The study was comprised of 284 participants (71 unmarried couples and 142 singles, 141
men and 143 women, age M = 23.7 years, SD = 2.7) who were recruited for a study on “Love,
Sexuality, and Personality” by flyers and postings in a number of public places with the incentive
of sixteen euro and personal feedback in return for participation.
The study began with the forced-choice with cognitive constraint, and participants were
instructed to evaluate situations that might occur in a relationship when you are distracted by
another task. Participants were presented with a six digit series that they were asked to keep in
mind, presented with descriptions that might occur in a relationship, and then asked to recall the
six digits. They were asked to spontaneously decide each time which of the two presented
dilemmas upset or distressed them more, as well as remember the digits. They were given 14
different pairs of situations that included six infidelity dilemmas adapted from the Buss et al.
(1992, 1999) and eight distractors, and the order was identical for all participants. The decision,
reaction time between the presentation of all 14 dilemmas and the participant’s decision, and the
correctness of the remembered number were all recorded.
In the next portion of the study, continuous emotion ratings, participants were instructed
to report without being distracted what they felt in six situations. They were asked to imagine the
situations vividly and realistically and then rate the situations for the amount of anger, anxiety,
jealousy, and humiliation on a five-point scale from ‘not at all’ to ‘extremely’. They were
presented in a fixed alternating order of an event of emotional or sexual infidelity. Ratings and
reaction times were recorded for each individual rating.
Lastly, participants completed the forced-choice with deliberation portion of the study.
They were once again presented with the six pairs of situations and asked to choose which one
would distress or upset them more. In this category, participants were instructed to take their
time to imagine the situation as vividly and realistically as possible to make a decision according
to their respective feelings. The same six jealousy dilemmas used in the cognitive load condition
were presented in the same fixed order, and the decisions and reaction times for each response
were recorded.
In the forced-choice under cognitive constraint condition, 16% of the participants
correctly remembered the six-digit number for all six dilemmas, 61% made two or more errors
and two participants failed every time to remember the number. A series of t tests were used to
analyze sex differences in the preference scores for emotional infidelity relative to sexual
infidelity. The test revealed a significant, moderately large effect for all choices of all
participants, confirming that more women (77%) than men (63%) judged emotional infidelity as
more distressing than sexual infidelity (Penke &Asendorpf, 2008).
Through a sex x education ANOVA indicated that the sex difference was moderated by
the educational level of the participants. The less educated participants demonstrated a stronger
sex difference, where as the better educated participants showed a small sex difference. Sex
differences in the reaction times were analyzed separately for sexual and emotional infidelity
choices. It was found that, decisions for sexual infidelity were as fast as decisions for emotional
jealousy and men and women did not differ in their decision time. However, the sex difference
was different for sexual and emotional infidelity. Later t tests showed that women were faster
than men when they chose emotional infidelity, where as an opposite inclination for sexual
infidelity was not significant (Penke &Asendorpf, 2008).
In the emotion ratings condition there was a significant, moderately large main effect of
sex, in that women rated the infidelity situations more negatively than men. It was then
suggested to analyze the sex effects separately for each infidelity type. Post hoc t tests showed a
significant effect of sex for ratings of jealousy, anxiety, and humiliation but not for anger. It was
revealed that women reported overall more negative emotions for both sexual and emotional
infidelity, with a slightly larger effect size for emotional jealousy. It was also found that the
education effect was mainly driven by stronger emotional jealousy of less educated women. In
regards to the sex differences in the reaction times, the reaction time data for the emotion ratings
did not show any significant effects of sex, type of infidelity, type of emotion, education or their
interactions (Penke &Asendorpf, 2008).
In the forced choice with deliberation condition, a significant, moderately large effect
confirmed that more women than men judged emotional infidelity as more distressing than
sexual infidelity. A sex x education ANOVA showed that this sex differenced was moderated by
the participant’s educational level. This indicated that the less educated participants showed a
strong sex difference where as better-educated participants showed only a small, marginally
significant sex difference (Penke &Asendorpf, 2008). In regards to reaction times the sex
difference for emotional infidelity was close to being significant. Penke and Asendorpf (2008)
state that:
The results for the two forced-choice tasks suggest consistently that the sex
difference for the classic forced-choice task is due to fast, spontaneous decisions,
rather than due to long deliberation and that it is more pronounced in less
educated participants (18).
The present study brought to light three main findings, two of which confirmed the
hypotheses derived from previous literature on sexual and emotional jealousy. The first finding
that women are more likely than men to pick emotional infidelity over sexual jealousy as the
more distressing choice when they were forced to decide, seems to be the result of spontaneous,
automatic reactions, not controlled deliberation (Penke &Asendorpf, 2008).
Secondly, the results for response measures insinuate that the sex difference in the
forced-choice condition is almost only driven by sex differences in emotional jealousy. Women
reported more negative emotions than men for both types of infidelity, with a slightly significant
stronger sex difference for emotional infidelity. It was also found that women chose emotional
infidelity faster than men in both forced-choice tasks, where as the tendency of men to choose
sexual infidelity was not significant (Penke &Asendorpf, 2008).
Lastly, attained education level turned out to be a strong moderator of the sex difference
in sexual versus emotional jealousy. Lower educated participants showed a large sex difference
where as the higher educated showed a small difference. The continuous emotion ratings
presented that this relative effect was due to a strong sex difference for emotional jealousy
among the lower educated participants (Penke &Asendorpf, 2008).
In regards to sexual jealousy, the early evolutionary discussions of jealous centered on
male sexual jealous as an evolved adaptation to minimize cuckoldry. But, the evidence that
human males get more jealous to a mates sexual infidelity than human females is mixed. A sex
difference such as this was not significantly confirmed in the present study or by the 10 similar
studies reviewed by Sagarin (2005). The reactive time measure signified that while in all four
cases, men chose sexual infidelity faster and processes cues for it faster than women, all
differences failed to reach statistical significance (Penke &Asendorpf, 2008). It is thought that
the emotion-rating task was affected by the general sex difference that women experience
emotions more intensely than men, in the context of relationship infidelity. Penke and Asendorpf
(2008) stated that:
Men have indeed evolved a higher sensitivity for sexual infidelity, this tendency
seems to interact with so many other factors that the resulting sex difference is
minimal, at least in the forced-choice and rating paradigms (21).
In contrast, the results for emotional jealousy consistently showed that women react more
jealous to emotional infidelity than men. In the emotion ratings and reaction times of forcedchoice decisions under cognitive load, women chose emotional infidelity as the more distressing
alternative faster than men. The finding that there is a weak sex difference for sexual jealousy,
suggest that the robust sex difference in the forced-choice paradigm is mainly driven by a sex
difference in emotional jealousy. The studies reviewed by Sagarin (2005) and the reaction time
studies by Schiitzwohl (2004, 2005) are also consistent with this view.
To address environmental influences regarding cognitive processing, the effect of
educational level on sex differences in emotional jealousy was found in all three conditions. This
suggests that emotional jealousy is already conditional to education-related environment factors
on the automatic level of processing.
Outline
A. Introduction
1. Debate over sex differences in romantic jealousy
a. Classic evolutionary hypothesis: men and women react differently to sexual
and emotional infidelity because these two types of infidelity posed different
adaptive problems
aa. Men are expected to be more jealous of sexual infidelity than women in
order to minimize investment in genetically unrelated offspring
bb. Women are expected to be more jealous of emotional infidelity than men
in order to minimize loss of parental investment in their offspring
cc. ‘Jealousy as a specific ‘innate’ module hypothesis” (Harris, 2000): this
evolutionary account consists of two independently derived hypotheses, one
for sexual and one for emotional infidelity, where one, none, or both may be
true
b. “Social cognitive theory of jealousy”: jealousy is the result of an evolved, but
domain general appraisal mechanism; this mechanism is sensitive to all kinds
of threats posed by rivals
aa. jealousy is aroused when a rival outdoes someone in domains that are
particularly important to the self
bb. this theory does not expect universal sex differences in sexual and
emotional jealousy
cc. they may or may not exist, depending on culturally determined gender
roles
2. Closer look at the two evolutionary hypotheses
a. both are weak tests of evolutionary hypotheses because they confound sex
differences in sexual jealousy and emotional jealousy
b. four cases of a sex by type of infidelity
aa. men react more jealous to sexual infidelity than women
bb. women react more jealous to emotional infidelity than men
cc. men react more jealous to sexual infidelity than to emotional infidelity
dd. women react more jealous to emotional infidelity than to sexual infidelity
c. forced-choice only tests whether one of the two evolutionary hypotheses or
both are supported; it does not test whether both hypotheses are confirmed
d. social-cognitive theory does not predict any specific between-sex or withinsex differences in emotional or sexual jealousy
3. The necessity to study cognitive processes
a. it is the aim of evolutionary psychology to study the design of cognitive
modules as the mediating mechanisms between evolution and adaptive
behavior
b. this requires and integration of the study of ultimate evolved functions and
proximate cognitive processes
c. spontaneity versus deliberation in terms of making a decision to cues of
emotional and sexual infidelity is important, as well as the extent to which
emotional versus sexual infidelity violate gender roles within a specific
culture
4. Empirical evidence
a. Forced-choice paradigm
b. Separate ratings of sexual and emotional infidelity
c. Cognitive processing of infidelity cues
B. The present study
1. aim was to test the two evolutionary hypotheses using a combination of
a. forced-choice method
b. continuous emotion ratings
2. cognitive processes
a. automaticity of responses in the forced-choice task
b. decision rating
3. hypotheses
a. the classic sex difference in the forced-choice paradigm under both
conditions, particularly under cognitive constraint
b. women rate emotional infidelity situations more negatively than men, but that
men do not rate sexual infidelity situations more negatively than women
c. women do not rate emotional infidelity cues faster than men, but that men do
not process sexual infidelity cues faster than women
C. Method
1. Sample
a. German native speakers between 20-30 years of age
b. Recruited in a large city by flyers and postings in various public places
2. Design
a. forced dilemmas with cognitive constraint and the instruction to respond
spontaneously
b. continuous emotion ratings with the instruction to vividly imagine each
situation
c. deliberate forced-choice dilemmas with the instruction to vividly imagine
each alternative and take enough time for the decision
D. Results
1. Data recording and screening
a. The order of sexual versus emotional infidelity within each of the six
dilemmas was counterbalanced.
b. Responses were recorded such that decisions for sexual and emotional
infidelity were always coded as 0 and 1.
2. Consistency Across Dilemmas
a. All three conditions were found to consistent across the dilemmas and
situations.
d. b. Individual log transformed reaction times formed sufficiently reliable
scales.
3. Forced-choice under cognitive constraint
a. First task: 16 % of the participants correctly remembered the six-digit
number for all six dilemmas, 61% made two or more errors and two
participants always failed to correctly remember the number.
b. For all choices of all participants, a significant, moderately large effect
confirmed that more women (77%) than men (63%) judged emotional
infidelity as more distressing than sexual infidelity.
c. Sex and education interactions
d. Sex differences in the reaction times
4. Emotion Ratings
a. There was a significant, moderately large main effect of sex, with women rating
the infidelity situations overall more negatively than men.
b. Post hoc t tests showed a significant effect of sex for ratings of jealousy,
anxiety, and humiliation but not for anger.
c. Women reported overall more negative emotions for both sexual and emotional
infidelity, with a slightly larger effect size for emotional jealousy.
d. Post hoc t tests showed that for the less educated participants, the sex difference
for sexual infidelity was small and not significant, but the sex difference for
emotional infidelity was significant and fairly large.
i. The education effect was mainly driven by stronger emotional jealousy of
less educated women.
5. Forced-choice with deliberation
a. A significant, moderately large effect confirmed that more women than men
judged emotional infidelity as more distressing than sexual infidelity.
6. Forced- choice task summary
a. The sex difference for the classic forced-choice task is due to fast, spontaneous
decisions, rather than due to long deliberation and that it is more pronounced in
less educated participants.
7. Relationship Effects
a. All participants had at least once experienced a committed, sexual relationship
that lasted at least 1 month.
b. Neither the total number of committed sexual relationships the participants had
so far, nor the duration of the current relationships of the paired participants were
significantly correlated with any of the jealousy measures.
E. Discussion
1. Three main findings; two confirmed the hypotheses derived from previous literature
a. Women are more likely than men to choose emotional infidelity over sexual
jealousy as the more distressing alternative when they are forced to make a choice
b. The sex difference in the forced-choice paradigm is almost exclusively driven
by sex differences in emotional jealousy. Women reported more negative
emotions than men for both sexual and emotional infidelity.
c. Attained education level turned out to be a strong moderator of the sex
difference in sexual versus emotional jealousy.
2. Sexual Jealousy
a. Early evolutionary discussions of jealousy focused on male sexual jealousy as
an evolved adaptation to minimize cuckoldry. The evidence that human males
react more jealous to a mate’s sexual infidelity than human females is mixed.
b. Men have evolved a high sensitivity for sexual infidelity, this tendency seems
to interact with so many other factors that the resulting sex differences is minimal.
3. Emotional Jealousy
a. The results for emotional jealousy consistently suggested that women react
more jealous to emotional infidelity than men.
4. Environmental Influences
a. Germany can be regarded as having a high degree of gender equality.
b. Concerning cognitive processing, the effect of educational level on sex
differences in emotional jealousy was found in all three conditions.
5. Strengths and limitations of the present study
a. Strengths
i. Systematic variation of the automaticity of the jealousy responses
ii. Greater reliability of the jealousy assessments
iii. the recording of the response times
iiii. Validity checks
b. Limitations: Only one culture was studied
Critical Review Items
(a) Interesting points
1. Interesting point: Although it was not the main concern of the researchers, education
played an important role in some of the findings. Penke and Asendorpf found larger
sex differences for their lower educated participants that were mainly driven by
stronger emotional jealousy of less educated women.
2. The fixed order of the three conditions in the present study is very important because
asking participants for spontaneous decisions after they have made deliberate ones on
the same scenarios could lead to skewed results.
3. The study aims to benefit evolutionary psychology by spending more effort of the
specific description and empirical testing of psychological mechanisms. The authors
suggest that their study is a step in the right direction, and they hope it will inspire
evolutionary psychologists interested in this and other domains of psychological
functioning to move from the ultimate expectations to the detailed study of proximate
cognitive mechanisms.
(b) Weaknesses
1. The present study only included young adults, who might weigh emotional versus
sexual infidelity more strongly according to their early developmental environments,
than older adults, who might increasingly weigh them according to their own
experiences
2. The findings are limited to the fact that one culture was studied. Only native German
participants were included in this study. Therefore, it does not allow researchers to
provide strong support got the adaptive plasticity of human jealousy.
3. The dilemmas provided in each condition of the study were hypothetical situations.
Penke and Asendorpf note that higher-level cognitive processes might be necessary to
anticipate the threat of sexual infidelity from more subtle cues that the straightforward
descriptions of sexual infidelity acts used in the study. The alternative of drawing
from memory rather than description, however, produces bias.
Test Questions
1. The aim of the present study was to
a. disprove previous findings
b. to test the two evolutionary hypotheses using the forced choice method
c. to test the two evolutionary hypotheses using continuous emotion ratings
d. a combination of b and c
2. To test the two evolutionary hypotheses, researchers have used a forced-choice
paradigm. What is the purpose of the forced-choice paradigm?
a. participants rate how they feel about dilemmas
b. participants take their time to reflect on whether they are more jealous of
emotional or sexual jealousy
c. participants are confronted with dilemmas that contrast hypothetical situations of
emotional and sexual infidelity, and they choose the more distressing.
3. What was one of the differences between the forced-choice with cognitive constraint
condition and the forced-choice with deliberation condition?
a. the hypothetical situations were different in each condition
b. the forced-choice with cognitive constraint condition included a six digit series
that participants were instructed to recall once they finished reading through the
dilemmas
c. the forced-choice with deliberation condition asked participants to rate their
emotions of anger, anxiety, jealousy, and humiliation on a 5-point scale in
addition to reading the dilemmas
4. The classic evolutionary hypothesis assumes that men and women react differently to
sexual and emotional infidelity because these two types of infidelity posed different
adaptive problems. (T/F)
5. All participants in the study were not in a serious committed relationship. (T/F)
6. Participants in the forced-choice with deliberation condition were presented with
hypothetical situations of emotional and sexual infidelity for the first time. (T/F)
Answer Key
1.
2.
3.
4.
5.
6.
d
c
b
T
F
F