Statistics Math


The primary purpose of this paper is to examine a research article titled “One woman’s behavior affects the attractiveness of others” by Mishra, Clark and Daly (2007) published in Evolution of Human Behavior. The article was written as an attempt to replicate previous research on “contrast effect” in evaluating attractiveness of relationship partners and strangers. While the study that was conducted failed to replicate previous results, it made an important contribution to psychological science. This paper is organized as follows. The first section introduces the problem. The second section describes the null and alternative hypothesis of the study. The third section describes dependent and explanatory variables. The fourth section is dedicated to internal validity in the study. The fifth section describes the sample that was used for the research. The sixth section describes the type of study. The two final sections are dedicated to the results of the study and its primary implications.

Problem Description

In their research article Mishra, Clark and Daly (2007) examine the so-called “contrast effect” that manifests itself in adverse evaluation of typical women and own relationship partners by men who were looking at photographs of attractive women. These findings are based on studies by Kenrick and Gutieres (1980), and Kenrick, Gutieres, and Goldberg (1989). These studies explain this “contrast effect” through suggesting that looking at photographs of attractive females initiates local mate pool assessment. Interestingly, a similar effect could not be induced in women. However, women would demonstrate adverse evaluation of typical men or their own relationship partners when provided with photos supplemented by biographical information that indicated high dominance of the pictured male. This suggests that men rely on visual stimuli when determining potential mating partners, while women rely on social and material success in such determination. This is the first idea that Mishra, Clark and Daly (2007) decided to test. Another idea tested by Mishra, Clark and Daly (2007) was “contrast effect” that manifested itself when experiment subjects watched videos that demonstrated proceptive behavior of members of opposite sex. Thus, their study was structured as two separate experiments.

Null and Alternative Hypothesis of the Study

The study conducted by Mishra, Clark and Daly (2007) consisted of two experiments. In the first experiment, participants who were visually isolated from each other were asked to answer a set of biographical questions and then rate a set of 15 photographs. These photographs were assigned to study participants randomly to determine one of the stimulus sets received by the participant. The first stimulus set consisted of attractive upper-body photographs of underwear models of opposite sex. The second set consisted of abstract artworks. Study participants who did not have a relationship partner answered Question Set B that asked them to rate average members of the opposite sex. Those participants of the study that were engaged in a relationship were asked to rank their relationship partner on 17 attributes. Therefore, the following set of null and alternative hypotheses for the first experiment were tested:

H01: Men would rate their relationship partners and average members of the opposite sex as less physically attractive having viewed photographs of attractive models, and would change after they were exposed to abstract art. This would confirm the existence of “contrast effect”.

HA1: Men’s evaluation of their partners or average females would not change after viewing photographs of underwear models.

H02: Similarly to the results of the previous studies, women would not rate their relationship partners or average members of the opposite sex adversely having been exposed to visual stimuli such as photographs of underwear models. This would confirm findings by other studies conducted in similar settings.

HA2: Visual exposure to attractive members of the opposite sex would make women rate their relationship partners as less physically attractive.

In the second experiment, settings were similar. However, instead of being exposed to visual stimuli such as artwork and photographs, participants of the experiment were shown short videotaped interviews of an opposite-sex student. In one video, the actor was behaving proceptively, while in the other, the actor behaved unreceptively. Having watched the video, participants were asked to rate the photograph of this actor and then answered the same sets of questions as in Experiment 1. Thus, the following set of hypotheses was tested for this experiment:

H01: Men would rate their relationship partners and average members of the opposite sex as less physically attractive having watched the video in which the actor behaved friendly.

HA1: Men’s evaluation of their partners or average females would not change after viewing videos with friendly actors.

H01: Women would not rate their relationship partners or average males adversely having been exposed to videos in which the actor behaved proceptively.

HA1:  Women would rate their relationship partners or average males adversely having been exposed to videos in which the actor behaved friendly.

Independent and Dependent Variables

Experiment 1: In this experiment, study participants evaluate their relationship partners in terms of physical attractiveness having been exposed to photographs of underwear models or art, the dependent variable was average evaluation of relationship partners (partner attractiveness rating), or stranger attractiveness (date worthiness of a stranger). Stimulus set was the set of explanatory variables. It may be inferred that stranger attractiveness or partner attractiveness could be impacted by stimuli that induce mate-seeking mechanisms. Therefore, there should be a co-variation between the two variables.

Experiment 2:  In this experiment, the dependent variable was partner attractiveness rating or stranger date-worthiness (for single study participants). Set of visual stimuli was used as a set of explanatory variables. It may be suggested that friendly behavior of a member of the opposite sex would trigger mechanisms that stimulate mate-seeking activity in both males and females. Therefore, the two variables should demonstrate co-variation.

Internal Validity in the Study

This study attempted to establish a cause-effect relationship. Internal validity of the study was increased through random assignment of stimulus sets to study participants. This reduced the risk of systematic differences between levels of the independent variable. Furthermore, participants of the study were visually isolated in both experiments. Thus, the study demonstrated the causal relationship between independent and explanatory variables by showing that:

  1. Stranger attractiveness rating and partner attractiveness rating changes after being exposed to visual stimuli (temporal precedence necessary to establish a cause-and-effect relationship);
  2. Dependent and independent variable demonstrate co-variation;
  3. Other possible explanations for changes in ratings of strangers or relationship partners were eliminated from the experimental setting.

Sample Composition and Sample Selection

Experiment 1: 87 male and 66 female undergraduate students (mean age 19,1; standard deviation 2,3 years) enrolled in an introductory psychology course.

Experiment 2: 72 male and 79 female undergraduate students (mean age 20,0; standard deviation 3,3 years) enrolled in a first-year psychology course.

Type of Study

The study that was conducted by Mishra, Clark and Daly (2007) was experimental. Experimental studies involve manipulating one variable to observe whether changes in this variable induce changes in another variable. In this study, visual stimuli were manipulated to observe changes in the set of mate attractiveness ratings. In this study, experiments were conducted in a controlled setting, and other factors that could affect changes in partner ratings were eliminated. This study involved random assignment of participants to levels of explanatory variable. This means that different study participants received different sets of visual stimuli.

Statistics Used and Results of the Study

The study used two-by-two analysis of variance (ANOVA) and means ratings in order to assess evaluation of photographs of underwear models and art in the first experiment. In the second experiment, two-by-two analysis of variance (ANOVA) was used. The study used t-statistics to check significance of the results.

Experiment 1: ANOVA analysis did not yield a significant result for both composite measure of partner evaluation and for averages of two subsets of evaluation characteristics (p-values were greater than .1). However, a marginally significant result was obtained: men who viewed photographs of models rather than abstract art rated their partners as less physically attractive (p-value < .05). Women did not demonstrate any significant difference in partner rating (p-value for variance analysis was greater than .5). For both sexes, attractiveness of strangers was affected by viewing photographs of attractive opposite-sex models (for men, t57=2,77, p<.01; for women t41=4,30, p < .001). Contrast effect was stronger in women than it was in men.

Experiment 2: In this experiment, both sexes rated the actors photograph higher if he or she was behaving friendly. For men, mean rating for a friendly female actor was 4,8 and for unfriendly actor, 2,8. For women, ratings were 3,5 and 3,0 respectively. The study found that men who viewed a proceptive video rated their partners lower on the composite measure of partner attractiveness (Sex/Stimulus condition interaction, F(1,67)=83,72, p < .001) while for women results were insignificant. Furthermore, men rated their partners as less physically attractive having watched a video with a friendly actor. Analysis of the results of the experiment for unmated participants yielded the following. Men’s ratings of stranger attractiveness and date worthiness were affected by the type of video that they observed. Study results reported that for men’s rating of stranger attractiveness, t-statistics was t37=2,86 (p < .01), and for men’s rating of stranger’s date-worthiness, t-statistics was t37=2,889 (p < .01). Women’s ratings were not affected by the type of video that they observed.

Implications of the Study and External Validity

In the first experiment, it was demonstrated that men did not downgrade their relationship partners on the composite score of attributes. Thus, results of the previous studies were not replicated. The second experiment demonstrated that men downgraded their relationship partners on a composite score of desirable attributes having viewed a video featuring a proceptively behaving female actor. This suggests that men may shift their preferences in order to court immediate partners (those behaving friendly towards them). Actors that behaved proceptively were rated higher by both sexes; however, men’s ratings of photographs were affected by the actor’s behavior than ratings of women. Furthermore, only men demonstrated contrast effects in the second experiment. This suggests that similarly to the animal world in the human world fitness of males is determined by the number of partners willing to mate. This also suggests that men reallocate mating effort based on cues received from attractive women. Furthermore, a proceptively-behaving female may be envisioned not only as physically attractive, but also as more pleasant and sociable. In context of the contrast effects discovered in the second experiment, this suggests that men’s downgrading of their relationship partners may occur not only in the dimension of their overall physical attractiveness, but also in the dimensions of likeability, sociability, and pleasantness.

External validity of the study is thus the following. First of all, men who are constantly exposed to attractive women at work or in another environment may develop high standards of rating females. This may suggest that they will have difficulties finding the ultimate mating partner in a situation when women to whom those men are exposed behave friendly or flirt. On the other hand, women who are exposed to physically attractive young men (for example, at the gym) will not have high standards of potential partners. Second, men’s mating preferences may be easily manipulated by females behaving proceptively. In other words, a woman willing to court a potential partner should behave friendly and interested towards him. Thus, this study proved a famous saying by Marlene Dietrich that “The average man is more interested in a woman who is interested in him than he is in a woman with beautiful legs”.