The effects of gender and power on persuasion

Download data and study materials from OSF

Principal investigators:

Asia A. Eaton

Florida International University



Penny S. Visser

University of Chicago



Sample size: 610

Field period: 2007


Findings from the current line of research on the relationship between social power and persuasion show that there are clear descriptive and prescriptive norms for those in high-power social roles to be resistant to attitude change, and that activating the concept of social power in undergraduates causes them to be resistant to weak persuasive messages than control participants, but equally persuaded by strong arguments compared to controls (Eaton and Visser, under review 2007). Although this initial research with undergraduates failed to uncover any gender differences in expectations for or responses to power, there are well-documented differences in the access that men and women in the U.S. have to social power (e.g. Depret & Fiske, 1993; Johnson, 1976; Kanter, 1977; Lips, 1991; Lorber, 1998), and gender differences in implicit and explicit associations between the self and power (Haines & Kray, 2005). This raises the possibility that men and women in the general U.S. population may conceive of and enact powerful social roles in distinct ways.
Data from this study suggest that men respond to power cues by conforming to the high-power role norm to be resolute. Women, on the other hand, may respond to power cues by acting in accord with the expectations for those in low-power roles, continuing to process arguments and be persuaded by strong messages. Control women who possessed power in their real lives (i.e. high levels of income), however, responded to persuasive messages much like men primed with the concept of power, by resisting persuasion to both strong and weak messages.


Men and women may conceive of and enact powerful social roles in distinct ways.
Being in a low-power position facilitates malleability and openness to influence.

Experimental Manipulations

In this 2 x 2 x 2 design, male and female participants were given either a power-priming or control vocabulary task and were then exposed to strong or weak counter-attitudinal persuasive messages.
For the vocabulary task, participants were randomly assigned to complete one of two versions of the task, one that semantically primed the concept of social power by exposing participants to words related to the concept of power or one that did not contain any power-related words (the control version). In this task, participants were asked to decide which of three synonyms was most similar to a target word. The
one-page task included 8 target words in total, each accompanied by 3 synonyms. In the power-priming version of the task, 5 target words (e.g. "authority," "powerful") and their synonyms were related to the concept of social power, whereas the remaining 3 target words (e.g. "common," "blend") and synonyms were power-neutral words used to disguise the hypothesis. In the control version of the vocabulary task, all eight target words and synonyms were unrelated to the concept of power.
To manipulate message quality, a strong, compelling argument in-opposition to the death penalty was constructed using information disseminated by the American Civil Liberties Union (also called the ACLU), an organization that opposes the death penalty. A weaker, less compelling message of the same length and format was also constructed.


Attitude change was the primary dependent variable. Attitude change was calculated by subtracting participants' post-message attitude towards the death penalty from their pre-message attitude towards the death penalty.

Summary of Results

A 2 x 2 x 2 analysis of variance (ANOVA) uncovered a significant 3-way interaction between participant gender, the prime type, and argument quality on attitude change, F(7,545) = 4.14, p < .05. Planned contrasts revealed that males and females in the control conditions showed significant argument quality-differentiation, each being more persuaded by the strong than the weak arguments. Male participants primed with the concept of power failed to show argument quality differentiation, being no more persuaded by the weak than by the strong arguments. This finding is consistent with the prescriptive expectation for powerholders to be generally resistant to persuasion. In contrast, female participants primed with power continued to show the same level of argument quality differentiation as females in the control condition. Upon further examination, it was determined that the effect of the power prime on persuasion to strong and weak messages for women was moderated by women's household income. Specifically, women in the control condition showed less argument quality differentiation as their income increased, becoming increasingly resistant to the strong message.

Additional Information

All participants were pre-selected for having positive attitudes towards the death penalty (all were above the neutral midpoint on a seven-point scale).
For analysis, the TESS participant sample was first restricted to those participants who spent a reasonable amount of time on the survey, as the on-line response format permits participants to speed through the responses without answering them carefully or to complete part of the survey at one time point and to return to the other part at a later time point. The sample was restricted to those participants who took long enough on the survey to actually thoughtfully fill out the measures but not so long that the effects of the prime wore off (between 2 and 30 minutes).


Eaton, A.A., and P. S. Visser. 2008. "The effects of gender and power on persuasion." Presented at the conference of the Midwestern Psychological Association. Chicago, IL. May 1-3.

Eaton, Asia A., Penny S. Visser, Jon A. Krosnick and Sowmya Anand. 2009. Social Power and Attitude Strength Over the Life Course. Pers Soc Psychol Bull 35:1646. DOI: 10.1177/0146167209349114