Edited and new postscript on Dec/4/2019

Photo credit: https://www.theblot.com/pole-dancing-daddy-fun-acrobatics-7767007
Who’s your daddy?  Priming women with a disengaged father increases their willingness to have sex without a condom.

In a five study article, Danielle J. DelPriore and Sarah E. Hill from Texas Christian University wanted to examine the influence of a disengaged father on daughter’s sexual attitudes and behaviors.

It is difficult to study the determinants of sexual behavior in humans because it is neither practical nor ethical to randomly assign daughters to engaged and distant fathers to see how this influences daughters’ sexual attitudes and behaviors.

Experimental social psychologists believe that they have found a solution to this problem.  Rather than exposing individuals to the actual experiences in the real world, it is possible to expose individuals to stimuli or stories related to these events.  These studies are called priming studies.  The assumption is that priming individuals has the same effect as experiencing these events.  For example, a daughter with a loving and caring partner will respond like a daughter with a distant father if she is randomly assigned to a condition with a parental disengagement prime.

This article reports five priming studies that examined how thinking about a distant father influences daughters’ sexual attitudes.

Study 1 (N = 75 female students)

Participants in the paternal disengagement condition read the following instructions:

Take a few seconds to think back to a time when your biological father was absent for an important life event when you really needed him . . .. Describe in detail how your father’s lack of support—or his physical or psychological absence—made you feel.

Participants in the paternal engagement condition were asked to describe a time their father was physically or psychologically present for an important event.

The dependent variable was a word-stem completion task with words that could be completed with words related to sex (s_x;  _aked;  sex vs. six; naked vs. baked).

Participants primed with a disengaged father completed more word-stems in a sexual manner (M = 4.51, SD = 2.06) than participants primed with an engaged father (M = 3.63, SD = 1.50), F(1,73) = 4.51, p = .037, d = .49.

Study 2 (N = 52 female students)

Study 2 used the same priming manipulation as Study 1, but measured sexual permissiveness with the Sociosexual Orientation Inventory (SOI; Simpson & Gangestad, 1991).  Example items are “sex without love is OK,” and “I can imagine myself being comfortable and enjoying casual sex with different partners.”

Participants who thought about a disengaged father had higher sexual permissiveness scores (M = 2.57, SD = 1.88) than those who thought about an engaged father (M = 1.86, SD = 0.94), F(1,62) = 3.91, p = .052, d = .48.

Study 3 (N = 82 female students)

Study 3 changed the control condition from an engaged father to a disengaged or disappointing friend.  It is not clear why this condition was not included as a third condition Study 2 but ran as a separate experiment. The study showed that participants who thought about a disengaged dad scored higher on the sexual permissiveness scale (M = 2.90, SD = 2.25) than participants who thought about a disappointing friend (M = 2.09, SD = 1.19), F(1,80) = 4.24, p = .043, d = .45.

Study 4 (N = 62 female students)

Study 4 used maternal disengagement as the control condition. Again, it is not clear why the researchers did not run one study with four conditions (disengaged father, engaged father, disappointing friend, disengaged mother).

Participants who thought about a disengaged dad had higher scores on the sexual permissiveness scale (M = 2.85, SD = 1.84) than participants who thought about a disengaged mother (M = 1.87, SD = 1.16), F(1, 60) = 6.03, p = .017, d = .64.

Study 5 (N = 85 female students & 92 male students)

Study 5 could have gone in many directions, but it included women and men as participants and used disappointing friends as the control condition (why not using engaged and disengaged mothers/fathers in a 2 x 2 design to see how gender influences parent-child relationships?).  Even more disappointing was that the only reported (!) dependent variable was attitudes towards condoms. Why was the sexual attitude measure dropped from Study 5?

The results showed a difference between male and female participants who thought about a disengaged dad or friend.  Participants reported more negative attitudes towards condoms after thinking about a disengaged dad (M ~ 3.4 based on Figure) than participants who thought about a disengaged friend (M = 2.9 ~ based on Figure), F(1,172) = 5.10, p = .025, d = 0.33.  The interaction with gender was not significant, p = .58, but the effect of the manipulation on attitudes towards condoms was marginally significant in an analysis limited to women, (M = 3.07, SD = 1.30; M = 2.51, SD = 1.35), F(1, 172)= 3.76, p = .054, d = 0.42.  Although the interaction was not significant, the authors conclude in the general discussion section that “the effects of primed paternal disengagement on sexual risk were also found to be stronger for women than for men (Experiment 5)” (p. 242).

CONCLUSION

Based on this set of five studies, the authors conclude that “the results of the current research provide the first experimental support for PIT [Parental Investment Theory] by demonstrating a causal relationship between paternal disengagement cues and changes in women’s sexual decision making” (p. 242).

They then propose that “insight gained from this research may help inform interventions aimed at reducing some of the personal and financial costs associated with father absence, including teen pregnancy and STI risk” (p. 242)

However, before researchers or lay people get too excited about these experimental findings, it is important to examine whether they are even credible findings.  Five successful studies may seem like strong evidence for the robustness of this effect, but unfortunately the reported studies cannot be taken at face value because scientific journals report only successful studies and it is not clear how many failed studies or analysis are not reported.

To examine the credibility and replicability of these reported findings, I ran a statistical test of the reported results.  These tests suggest that the results are not credible and unlikely to replicate in independent attempts to reproduce these studies.

N	statistic	p	z	OP
75	F(1,73)=4.51	0.037	2.08	0.55
64	F(1,62)=3.91	0.052	1.94	0.49
82	F(1,80)=4.24	0.043	2.03	0.53
62	F(1,60)=6.03	0.017	2.39	0.67
177	F(1,172)=5.10	0.025	2.24	0.61

OP = observed power

The Test of Insufficient Variance (TIVA) shows that variance of z-scores is much less than random sampling error would produce, var(z) = 0.03 (expected 1.00), p < .01.   The median observed power is only 55% when the success rate is 100%, showing that the success rate is inflated. The Replicability Index is 55 – (100 – 55) = 10.  This value is below the value that is expected if only significant studies are selected from a set of studies without a real effect (22). A replicabilty index of 10 suggests that other researchers will not be able to replicate the significant results reported in this article.

Conclusion

This article does not contain credible evidence about the causes of male or female sexuality, and if you did grow up without a father or with a disengaged father it does not mean that this necessarily influenced your sexual attitudes, preferences, and behaviors.  Answers to these important questions are more likely to come from longitudinal studies of real family relationships than from priming studies that assume real world experiences can be simulated in a cheap laboratory study with priming.  However, social psychologists cherish the illusion that they are real scientists because they conduct research in ‘laboratories’ as if they are chemists or microbiologists.  Somebody needs to tell them that climate science, astronomy, and zoology are real sciences, too.  The rise and fall of priming studies will be an interesting chapter in the history of psychology. 

Postscript (Dec/4/2019)

I checked on the citations of this article and found that it had been cited only 9 times. Four citations were by the authors, including a JPSP article in 2018 that reported four more priming studies. The p-values of the focal tests were, p = .030, .049, .028, and .062. The last result was accompanied by a comment that “given the extant literature demonstrating reliable effects of paternal absence-disengagement on sexually proceptive behavior in women
(as reviewed in the Introduction), a one-tailed statistical test (p = .031) could be justified here, supporting a causal effect of paternal disengagement on flirting. My criticism of the 2013 article was posted in 2016 and there was widespread awareness about questionable research practices, and concerns about the replicability of priming studies. The fact that JPSP published a similar article in 2018 shows that some influential social psychologists are unable or unwilling to improve the quality of social psychological research.

Reference

DelPriore, D.J., & Hill, S.E. (2013). The effects of paternal disengagement on women’s sexual decision making: An experimental approach, Journal of Personality and Social Psychology, 105, 234-246. DOI: 10.1037/a0032784