Looking for Blame: Rape Myth Acceptance and Attention to Victim and Perpetrator

Abstract

In two studies, the authors examined the influence of rape myth acceptance (RMA) on participants’ attention toward the potential victim versus perpetrator in a rape case. In Study 1 (N = 90), participants selected information that focused on either the male defendant or the female victim. With increasing RMA, participants preferred information that focused on the victim rather than the defendant. In Study 2 (N = 41), participants viewed photographs depicting both victim and defendant while their eye movements were recorded. With increasing RMA, participants spent less time inspecting the defendant relative to the victim. In both studies, higher RMA predicted stronger anti-victim and pro-defendant judgments, replicating previous research. Taken together, these results support the assumption that RMA guides participants’ attention, leading to a focus on the alleged rape victim and away from the alleged perpetrator. Implications of the current research and future directions are discussed.

Keywords

eye tracking rape case rape myths sexual violence visual attention

A plethora of research has demonstrated that humans actively interpret their reality in accordance with their attitudes, world views, and their cultural background. In doing so, we not only generate individual interpretations of reality (Vygotsky, 1930-1934/1978) but we are also able to “go beyond the information given” (Bruner, 1957) by drawing inferences based on knowledge and past experience. Stimulated by work that demonstrated such influences of culture on cognition, recently research has turned to perception and attention as processes traditionally conceptualized as rather universal across people and time and demonstrated effects of a participant factor such as culture on these (e.g., Miyamoto, Nisbett, & Masuda, 2006; Nisbett & Miyamoto, 2005). The present research builds on this work and examines how people attend to cases of sexual violence, how people selectively search for information based on their attitudes, and how they literally “see” a rape case given a certain world view.

In a rape case, police officers, jurors, and judges usually receive information about two opposing parties: the potential perpetrator and the alleged victim. It is their behavior (e.g., prior alcohol consumption), their characteristics (e.g., status, attractiveness), as well as their relation to one another (i.e., strangers, acquaintances, or intimate partners) that influence the verdict rendered by jurors or research participants (for an overview, see Temkin & Krahé, 2008). Depending on the case at hand, this decision might be rather easy with strong agreement across people (as in the case of a violent stranger rape), or very difficult and ambiguous with a lot of variation between people’s responses (as in the case of an acquaintance rape). Past research has shown that individual-difference variables such as stereotypical rape-related attitudes bias information processing not only in face of ambiguous cases (Frese, Moya, & Megias, 2004) but also in rather clear-cut rape cases (Abrams, Viki, Masser, & Bohner, 2003). In the present research, we take one step back by examining the influence of such attitudes on individuals’ attentional preferences when encountering new information in an alleged rape case. Thus, the present research investigates where people look and search for relevant information before processing that information as a possibly important route to victim blaming that hitherto has not been investigated. Concerning the direction of a potential bias, we posit that stereotypical rape-related attitudes go hand in hand with a focus away from the perpetrator and toward the alleged rape victim. Below, we briefly introduce the concept of rape myth acceptance (RMA) with an emphasis on research studying its influences on decision making in a juridical context; then we turn to the present research.

Rape myth acceptance

Rape myths can be defined as beliefs “that serve to deny and justify male sexual aggression against women” (Lonsway & Fitzgerald, 1994, p. 134). Thus, they address stereotypes about victims and perpetrators as well as the contexts in which an assault would occur (Bohner, 1998). Apart from studies on the attribution of responsibility and blame in mock-juries (e.g., Krahé, 1991), rape myths play a role in a variety of other important contexts, such as men’s self-reported likelihood of raping (Bohner, Siebler, & Schmelcher, 2006) or the recovery process of rape survivors (Burt & Katz, 1988; Littleton, Axsom, Radecki Breitkopf, & Berenson, 2006). Research on rape myth acceptance—that is, the extent to which these beliefs are endorsed—has focused on various issues, including the correlational links of RMA to other constructs (e.g., Süssenbach & Bohner, 2011; for reviews, see Bohner, 1998; Lonsway & Fitzgerald, 1994), general and gender-specific functions of RMA (e.g., Bohner & Lampridis, 2004; Bohner et al., 2006; Bohner, Weisbrod, Raymond, Barzvi, & Schwarz, 1993; for a review, see Bohner, Eyssel, Pina, Siebler, & Viki, 2009), measurement issues resulting in the development of several RMA scales (Burt, 1980; Cowan & Quinton, 1997; Gerger, Kley, Bohner, & Siebler, 2007; Payne, Lonsway, & Fitzgerald, 1999), and has spawned a variety of intervention programs (e.g., Berkowitz, 2003; Foubert & Marriott, 1997).

RMA and Juridical Decision Making

Research investigating the factors that influence participants’ decision-making processes in a rape case typically features information about an alleged case that is presented to participants in text vignettes to systematically vary aspects of the case. Prior research in the tradition of these mock-jury paradigms has shown that a host of information pertaining to either perpetrator or victim may influence verdicts in these text cases. To give some examples, participants’ judgments may be influenced by whether witness or defendant (or both) consumed alcohol (Cameron & Stritzke, 2003; Schuller & Wall, 1998), whether the victim wore revealing clothing (Whatley, 1996), whether the victim was attractive (Deitz, Littman, & Bentley, 1984; Jacobson & Popvich, 1983), or whether the defendant was similar to themselves (Helmke, Kobusch, Rees, Meyer, & Bohner, 2014). In fact, research by Rempala and Bernieri (2005) suggests that, irrespective of content, more information about one of the targets will generally lead to increased blaming of that party as it becomes more salient (for alternative findings, cf. Eyssel & Bohner, 2011). Rempala and Bernieri further argue that rape trials may systematically generate a greater focus on the alleged victim, who will routinely be extensively cross-examined as a witness, than on the defendant, who may exert his right to remain silent (see also Madigan & Gamble, 1991).

In addition, prior research also demonstrated that the influence exerted by such information interacts with participants’ levels of RMA. For example, Krahé, Temkin, and Bieneck (2007) showed that with increasing RMA, their research participants became more sensitive to an experimental manipulation concerning the prior victim–perpetrator relationship, and attributed more blame to the victim the closer her prior relationship to the perpetrator had been. Accordingly, RMA is conceptualized in terms of a cognitive schema that “guides and organizes an individual’s interpretation of specific information about rape cases” (Eyssel & Bohner, 2011, p. 1581, see also Süssenbach, Bohner, & Eyssel, 2012)

The Present Research

Our research aims were to investigate the influence of rape-related attitudes on the type of information participants are interested in and consider relevant for their judgment. Thus, different from prior research, the focus of this article is not on biased information processing but rather on biased information search, with a particular emphasis on biased visual attention as one form of biased information search. To test this notion, we conducted two studies: In Study 1, we exposed participants to a short vignette depicting a rape case and then provided them with the opportunity to select additional information about the case. Participants could select information that focused on either the witness (i.e., the potential victim) or the defendant (i.e., the potential perpetrator), but in fact never received the information chosen. In Study 2, we tracked participants’ eye movements while they viewed photographs depicting witness and defendant in a rape case after participants had read about the case. We assumed that with increasing RMA, participants would wish to receive more information about the potential victim relative to the defendant (Study 1) and pay more attention to her than to him (Study 2), thereby looking for ways to assign blame to the victim. It is important to note that, as participants either never received the information they were interested in (Study 1) or viewed only neutral information (Study 2), biased information processing was obstructed in the present research. Thus, RMA-related biases in information search were not expected to aggravate the influence of RMA on case judgments (however, in the presence of relevant stimuli [Süssenbach et al., 2012] or the feeling of having received valuable information [Eyssel & Bohner, 2011], one would expect an aggravated effect of RMA on case judgments). Thus, the present research disentangles biased processing from biased information search via providing no or only irrelevant information.

The prediction that RMA leads to an attentional shift away from the perpetrator toward the victim is consistent with research on social hypothesis testing, showing that people restrict their attention to a single favored hypothesis and systematically search for information potentially confirming it, while neglecting alternative hypotheses (e.g., Klayman, 1995; Nickerson, 1998). Our prediction is also consistent with research on the confirmation bias more generally, demonstrating that individuals look for or interpret information in a way that confirms their preconceptions (Bassok & Trope, 1984; Devine, Hirt, & Gehrke, 1990). Because most rape myths are centered on women (e.g., “A lot of women lead a man on and then they cry rape,” “Many women secretly desire to be raped,” “Any healthy woman can successfully resist a rapist if she really wants to”; see Burt, 1991, for a taxonomy of content categories regarding rape myths), participants with higher RMA should show higher interest for information pertaining to the potential victim relative to information pertaining to the potential perpetrator to find evidence congruent with their belief system.

Study 1

Method

Participants, procedure, and design

Participants (N = 90; 50% female) were recruited on the campus of Bielefeld University, Germany, for a laboratory experiment on judgmental processes in jurors’ decision making. Participants’ mean age was 25.69 (SD = 6.11) years. Most participants were students (n = 83; 20 majoring in psychology). Participants were randomly assigned to one of three conditions: information selection A (n = 30), information selection B (n = 30), and control (n = 30)—see below for details. They individually completed the study materials, which were presented on a desktop computer via MediaLab (Jarvis, 2005). On completion, participants were thanked, debriefed, and received 2 Euros as compensation for their time.

Materials

RMA

All participants completed a German 11-item short version of the Acceptance of Modern Myths About Sexual Aggression scale (AMMSA; Gerger et al., 2007), an instrument that was successfully used in prior research (e.g., Süssenbach, Eyssel, & Bohner, 2013).¹ This scale was designed to measure contemporary myths regarding sexual violence (e.g., “Women like to play coy. This does not mean that they do not want sex”; “Many women tend to exaggerate the problem of male violence”). Each item was rated on a 7-point response scale ranging from 1, completely disagree, to 7, completely agree.

Rape case

Participants read a short vignette about an alleged rape case and were asked to take the perspective of a lay juror. In the following, a summary of the presented scenario is given: Defendant (“Jens E.”) and witness (“Karin S.”) had met during a party held by the company for which they both worked. Although they knew each other from other occasions, they had never talked to each other before that night. During the party they engaged in a lively conversation. Later the same night, the defendant and witness went for a walk at a riverside close-by. They continued their conversation at the shore of the river and both parties confirmed that they started kissing there. However, their statements diverge with regard to the subsequent events. Whereas the witness stated that the defendant had raped her, the defendant claimed that consensual sexual intercourse had taken place.

Additional information

Following the case information, participants in the two information selection conditions (n = 60) were instructed to select five out of 10 additional pieces of information whose labels were presented on the computer screen. Participants checked the labels of those pieces of information that they would like to subsequently inspect. Half of the labels indicated information that focused on the witness, the other half on the defendant (“Psychological expertise on Karin S./Jens E.,” “Information on the relationship status/the sex life of Karin S./Jens E.,” “Credibility report on Karin S./Jens E.,” “Detailed background information on Karin S./Jens E.,” and “Statement of the lawyer of Karin S./Jens E.”). Witness- and defendant-focused information was presented in mixed order. The two information selection conditions A and B differed only with regard to the order in which information labels were presented on the screen (B presenting items in reverse order relative to A). A control condition in which participants did not have the opportunity to select additional information was included to make sure that the mere opportunity of selecting additional information had no influence on case judgments. Participants in the control condition (n = 30) directly continued with the case judgments.

Case judgments

After reading about the rape case (control condition) or selecting additional information (information selection conditions), participants were asked to take the perspective of a lay juror in judging the rape case. In all conditions, they were told that we were interested in their current opinion regarding the case before they would receive any further information. Participants first provided a verdict by indicating the likelihood of the defendant’s guilt (“In your opinion, how probable is the defendant’s guilt?” from 1, not at all probable, to 7, very probable) and subsequently recommended a sentence length (“What sentence length do you consider appropriate?,” from 1, acquittal, to 7, 6 years in prison). Blame attributions for defendant and witness (i.e., victim and perpetrator blame) were assessed separately with four items measuring attributions of responsibility and influence (“How responsible is he/she for what has happened?,” from 1, not at all responsible, to 7, fully responsible, and “How much influence did he/she have on the outcome of the situation?,” from 1, no influence at all, to 7, very much influence).

Perception of additional information

After providing case judgments, participants in the information selection conditions were again shown the labels of all 10 pieces of information they had chosen from, and rated the expected informativeness of each piece (“How informative is this piece of information?” from 1, not at all, to 7, very). Subsequently, for each of the 10 pieces of information, participants indicated their expectation of whether this information would (a) exonerate or incriminate the defendant and (b) suggest contributory fault on the part of the witness (“To what extent will this information rather incriminate or exonerate Jens E.?,” from 1, not incriminate at all, to 7, strongly incriminate, and “To what extent will this information speak for or against a contributory fault of Karin S.?,” from 1, against contributory fault, to 7, for contributory fault.).

Results

Preliminary analyses

Scoring of variables

The mean across the 11 AMMSA items was defined as a participant’s RMA score (Cronbach’s α = .89; M = 3.05, SD = 1.18). Participants’ six case judgments were reverse-coded where appropriate and then averaged to form a case judgment index with higher values indicating less perpetrator blame, more victim blame, and more lenient verdicts (Cronbach’s α = .77; M = 3.62, SD = 0.98). Participants’ selection of additional information was scored such that the number of selected witness-focused pieces of information was counted (e.g., a score of 3 would reflect the selection of three witness-focused and two defendant-focused pieces of information). Therefore, this score (“type of information”) can range from 0 to +5, with higher scores indicating that more witness-focused relative to defendant-focused pieces of information were chosen (M = 2.27, SD = 0.61).

Comparison of condition means

The order in which the labels of additional information pieces were presented (information selection conditions A vs. B) had no effect on the selection of information nor on the case judgment index; both Fs in the univariate analysis of variance (ANOVAs) contrasting information selection conditions A and B were less than 1; therefore, the two conditions were combined. Whether or not participants had been asked to select additional information before providing case judgments (i.e., combined information selection conditions vs. control condition) also had no influence on the case judgment index; the univariate ANOVA contrasting the control condition from the combined information selection condition yielded an F < 1.

Main analyses

Predicting participants’ information selection and case judgments

As hypothesized, RMA significantly predicted participants’ case judgments, β = .58, t(89) = 6.75, p < .001, accounting for 34.1% of the variance in a regression analysis with case judgments being regressed on RMA. To test our main hypothesis, a regression analysis was performed with type of information as the dependent variable and RMA as the predictor. In support of our rationale, within the combined information selection conditions, participants’ RMA also predicted which type of information participants wished to receive in addition, with higher RMA predicting a focus on the witness relative to the defendant, β = .35, t(59) = 2.82, p < .01; R² = .12. Including sex of participant as a covariate in the above-reported regression analyses had no influence on the case judgments nor information selection, both ps > .60.

Expectation regarding the selected and non-selected additional information

We analyzed participants’ expectations regarding the perceived likelihood that the information would incriminate the defendant and regarding the likelihood that the information would suggest contributory fault on the part of the witness. We also related these expectations to participants’ endorsement of rape myths. An analysis of these expectations sheds light on whether or not participants had different expectations for information that either focused on the victim or on the perpetrator and might explain why people high in RMA prefer information about one over the other.

On average, participants reported that witness-focused information (M = 4.60, SD = 0.93) would incriminate the defendant to a greater degree than defendant-focused information would (M = 4.00, SD = 0.74), t(59) = 6.06, p < .001, d = .78 (dependent t test comparing the incrimination of the defendant expected for witness- and defendant-focused information). Likewise participants expected that witness-focused information (M = 3.76, SD = 0.97) would be less likely to suggest a contributory fault on the part of the witness than would defendant-focused information (M = 4.09, SD = 0.69), t(59) = 2.81, p < .01, d = .36 (dependent t test comparing the contributory fault expected for witness- and defendant-focused information). Whereas expectations on whether or not a piece of information would incriminate the defendant or not was not systematically related to RMA (ps > .20; two Pearson correlations relating the incrimination of the defendant expected for witness- and defendant-focused information to RMA), participants with higher RMA expected witness-focused information, r(58) = .27, p < .05, as well as defendant-focused information, r(58) = .26, p < .05, to imply higher contributory fault on the part of the witness. To summarize, participants high on RMA expected both victim- and defendant-focused information to suggest a contributory fault on the part of the victim (with both effects being equally strong). On average (that is, irrespective of RMA), information about the victim was rated to be more favorable for her, whereas on average information about the perpetrator was perceived as more favorable for him.

Interestingly, information on the witness (M = 5.04, SD = 0.86) was perceived as somewhat more informative than information on the defendant (M = 4.79,SD = 0.84), t(59) = 2.52, p < .05, d = .33 (dependent t test comparing the informativeness expected for witness- and defendant-focused information); however, participants’ level of RMA was not related to these informativeness ratings, ps > .22 (two Pearson correlations relating the perceived informativeness of witness- and defendant-focused information to RMA).

Thus, according to these self-reported expectations, there should be no reason for someone high in RMA to favor witness-focused information over defendant-focused information. Quite on the contrary, it would be more rational for someone high in RMA to select information on the defendant because such information would ostensibly be more suited to exonerate him and blame her as an analysis of the means of the incrimination and contributory fault ratings suggest.

Discussion

In this first study, the hypothesis was tested that higher RMA is associated with a preference for information about an alleged victim in a rape case. Accordingly, participants were given the opportunity to select additional information about witness and defendant in a rape case. As hypothesized, RMA predicted a substantial proportion of participants’ selection choices in the expected direction. Concerning participants’ expectations regarding the available additional information, we found that participants in general assumed that witness-focused relative to defendant-focused information would speak against the potential perpetrator (i.e., the information would incriminate him) and for the potential victim (i.e., the information would exonerate her). We furthermore found that high-RMA participants’ preference for information about the potential victim over information about the potential perpetrator cannot be explained by differences in expectations due to RMA (i.e., only expectations concerning contributory fault were related to RMA, but this effect was nearly identical in size for victim- and perpetrator-focused information). Therefore, high-RMA participants’ information selection preferences are seemingly at odds with their reported expectations (as long as we assume that participants are truthfully reporting their expectations): Like participants low in RMA, they expected that victim-focused information would speak for the potential victim and against the potential perpetrator. These results do however support our notion that RMA functions in terms of a cognitive schema that actively guides information search when individuals are confronted with a rape case, resulting in a focus away from the perpetrator and toward the victim, whose characteristics and behavior are at the heart of most rape myths.

Nevertheless, the results of this study are subject to a number of limitations. For one, its operationalization might be considered somewhat artificial with participants basing their selection choices on headings of additional pieces of information. Although this procedure maximizes control, it could be argued that information in the real world is less pre-structured (and of course less readily available), thus questioning the generalizability of the present findings.

Furthermore, one might argue that participants’ RMA-related preference for victim-focused information is not the consequence of a biased information selection process, but rather a strategic means to exonerate the potential perpetrator. In line with such an objection, Bohner (2001) demonstrated that greater RMA was associated with the use of the passive voice when research participants wrote about a rape case. He argues that such a linguistic bias serves to reduce the salience of the actor to be able to exonerate him. Selecting information that focuses on the victim of a rape case might serve the same purpose: To illustrate, Rempala and Bernieri (2005) demonstrated that responsibility attributions in rape cases may be influenced by target salience. Increasing the amount of information about either victim or perpetrator in a rape case led to higher responsibility and guilt attributions for the party that was made more salient. In the context of the present study, it is possible that participants high in RMA chose victim-focused information not as a result of a biased information search but as a strategic means to redirect attention away from the potential perpetrator toward the victim.

To address these concerns, we conducted a second study. Study 2 therefore investigated participants’ information search processes in a less reactive manner by using eye tracking as the method of choice.

Study 2

The aims of this second study were to demonstrate biased information search in participants’ overt gaze behavior. To explore this, participants viewed three photographs after having read about a rape case. Ostensibly, these photographs depicted witness and defendant on the night of the incident. To measure subtle and automatic responses to the visual information, we tracked participants’ eye movements during inspection of the three pictures. We chose the analysis of viewing patterns of complex and realistic photographs as a non-reactive way to study participants’ information search preferences. While this technique minimizes possible demand effects (i.e., participants are less aware of the research hypothesis), it also reduces the plausibility of the aforementioned limitations. It is rather unlikely that participants use their gaze behavior as a strategic means to justify their opinion because they are neither directly nor indirectly communicating with anyone. As indicators of visual attention, we used participants’ overall dwell time on the victim and the perpetrator. To summarize, we hypothesized that higher RMA would predict greater interest in the alleged victim relative to the perpetrator of the rape case, resulting in greater visual attention paid to her than to him.

Method

Participants and design

A sample of 41 male students (seven psychology majors)² with an average age of 27.74 years (SD = 5.86) from Bielefeld University, Germany, participated. Participants were approached on campus and invited to take part. The study featured a between-subjects manipulation of social norms, that is, participants received feedback about their fellow students’ responses to the RMA questionnaire. Depending on condition, this false feedback either implied fellow students having low (n = 21) or high levels of RMA (n = 20). To anticipate, the social norm manipulation had no effect on participants’ judgments or viewing patterns and is therefore not discussed.

Procedure

Participants took part in two ostensibly unrelated studies. They were told that in the first study, they would complete a variety of attitude questionnaires and that they would receive feedback about some of them afterward. Thus, participants completed a variety of self-report measures, including a scale assessing RMA, which were computer-administered via MediaLab (Jarvis, 2005). On completion of the self-report package, participants learned about their fellow students’ alleged RMA responses, which, depending on condition, constituted either high-RMA or low-RMA normative feedback. After completion, participants were escorted to a different laboratory to take part in a second study on information processing and decision making. They were then familiarized with the eye tracker. After calibration of the eye tracker (using a standard 13-point calibration; see SMI, 2009), participants read about a rape case and then viewed three photographs that displayed defendant and witness of the rape case. Presentation time of all three photographs (each for 6 s) as well as entry point to the photographs (fixation point in the center of the screen) was held constant to ensure comparability of participants’ eye tracking data. After viewing the photographs, participants provided verdict, blame, and responsibility judgments using a paper-and-pencil questionnaire. Participants completed the questionnaire package on their own in the laboratory to avoid experimenter effects. While they were seated in front of the eye tracker, a separating wall divided participants from the experimenter. On completion of the study, participants completed an open-ended suspicion probe. (None of the participants reported that the two allegedly independent studies were connected.) Then participants were thanked, debriefed, and reimbursed with 2 Euros.

Apparatus

Eye movements were recorded monocularly at 240 Hz with an I-View X-High-Speed system (SMI, Berlin) using pupil locations as well as corneal reflections. The experiment was presented on a 365 mm (1280 pixel) wide by 270 mm (1024 pixel) high CRT monitor refreshing at 60 Hz. The computer screen was positioned 700 mm in front of the participant, who sat with head supported by the chin and forehead rest of the iView tracking column. Integrated software was used for stimulus presentation (SMI Experiment Center; see SMI, 2010b) as well as data analysis (SMI BeGaze; see SMI, 2010a).

Materials

RMA

Participants completed a 16-item short version of the AMMSA that included the 11 items used in Study 1 plus 5 items that were selected from the 30-item scale based on their item-to-total correlations (Cronbach’s α = .93; M = 2.95, SD = 1.32).

RMA feedback

After completing the questionnaire package, participants were told that they would be shown the responses of male Bielefeld students to one of the self-report measures they had just filled out themselves. To provide an explanation for the feedback, they were informed that past test takers had often expressed a wish to know what other people thought about these topics. All participants then received feedback on the RMA questionnaire. Each of the 16 AMMSA items was presented individually on the screen, and above the item wording the following text was displayed: “The mean value of male Bielefeld students is [value].” Depending on condition, the respective “value” was either 1 SD below or 1.5 SDs above the mean of the item-wise descriptive statistics of a prior study. The mean of the aggregated feedback value was 1.55 in the low feedback condition and 5.43 in the high feedback condition (on a scale ranging from 1 to 7). This procedure followed other successful uses of social norm manipulations (see Bohner, Pina, Viki, & Siebler, 2010; Bohner et al., 2006; Eyssel, Bohner, & Siebler, 2006).

Rape case

After calibration of the eye tracker, participants were instructed to take the perspective of a lay juror and read a short vignette describing a rape case. In the following, a summary of the scenario is given: The male defendant and female witness had met at the birthday party of a common friend. Both stated that they engaged in a prolonged conversation and danced for some time, which was confirmed by other guests of the party. Later that night, the defendant and witness left for the witness’s home where they intended to drink a glass of wine. Both parties agreed that they continued their conversation in the witness’s apartment and then started kissing. However, the statements diverge with regard to the subsequent events. Whereas the defendant claimed that consensual sexual intercourse had taken place, the witness stated that she had been raped.

Photographs

After reading the case information, participants were told that they would view three photographs that had supposedly been taken by other guests of the birthday party and that depict witness and defendant. The faces of both witness and defendant on those photographs were pixelated, ostensibly for legal reasons associated with data privacy. All photographs showed both parties in front of a bar counter in varying postures (see Figure 1).

Figure 1.

The first of three photographs depicting witness and defendant that participants viewed.

Visual interest

We conducted an analysis of areas of interest. In such an analysis, certain parts (areas) of a display (showing photographs in the present study) are defined and only eye movements that fall in these areas are analyzed. Areas of interest were defined for each of the three photographs. One comprised the body of the witness; the other the body of the defendant. For each photograph, the overall dwell time on the defendant (i.e., the total time participants spent looking at the male protagonist; across the three photographs: M = 2,272.43 ms, SD = 827.34 ms) was subtracted from the overall dwell time on the witness (i.e., the total time participants spent looking at the female protagonist; across the three photographs: M = 2,427.33 ms, SD = 598.10 ms). These difference scores were then aggregated across the three photographs. Thus, higher scores reflect more visual attention paid to the witness relative to the defendant (M = 154.90 ms, SD = 1,251.03 ms).

Data of 14 participants (8 in the low-RMA, 6 in the high-RMA feedback condition) had to be excluded for analyses involving eye tracking because of imprecise eye tracking. Including badly tracked participants does not affect results.³

Case judgments

Items used were identical to the ones in the first study (Cronbach’s α = .83; overall M = 3.76, SD = 1.16).

Results

Mean scores were calculated by averaging across the corresponding items (and reverse coding where necessary).

Case judgments

In a multiple regression analysis with RMA and condition as independent variables, only RMA was a significant predictor of participants’ case judgments, β = .36, t(40) = 2.43, p < .05, accounting for 13.0% explained variance. Participants in the low-RMA feedback condition (M = 3.54; SD = 1.07) did not differ significantly from participants in the high-RMA feedback condition (M = 3.98, SD = 1.23), β = .19, t(40) = 1.27, p = .21, R² = .04.

Visual interest

A multiple regression analysis with RMA and condition as predictors of participants’ visual attention toward the three photographs (i.e., the relative score integrating participants’ dwell times on witness and perpetrator) similarly found no effect of condition. Low (M = 36.21 ms, SD = 732.09 ms) and high norm (M = 115.74 ms, SD = 533.41 ms) feedback conditions did not differ significantly, β = .10, t(24) = 0.52, p = .61, R² = .01. Importantly, however, participants’ RMA predicted visual attention, with higher RMA resulting in a stronger focus on the witness relative to the defendant as hypothesized, β = .41, t(24) = 2.19, p = .04, R² = .17.⁴ To analyze this significant effect more deeply, identical regression analyses were conducted for the two areas of interest, that is, visual attention toward the witness and toward the defendant, separately. As with the relative score, norm feedback had no influence on viewing patterns, ps > .20. More importantly, whereas higher RMA was positively (but not significantly) related to a longer dwell time on the witness, β = .22, t(24) = 1.13, p = .27, R² = .05, it was strongly related to reduced looking at the defendant (i.e., a shorter dwell time), β = −.46, t(24) = −2.62, p = .02, R² = .21.

Discussion

The aim of our second study was to demonstrate that rape-related attitudes affect participants’ visual attention, resulting in a focus away from the potential perpetrator and toward the potential victim. The present results suggest that participants’ endorsement of rape myths not only predicted their explicit case judgments but also their gaze behavior as hypothesized: Individuals with higher RMA did prefer information about the potential victim over information about the potential perpetrator. This bias was evident in participants’ overall dwell time on the targets. However, a more precise analysis of participants’ dwell time data suggests that this effect might be driven by a RMA-based redirection of visual attention away from the potential perpetrator. Thus, the viewing patterns of high-RMA participants might be more appropriately characterized as an avoidance of looking at the potential perpetrator and less so as an orientation toward the potential victim. However, this finding needs to be corroborated by future research.

General Discussion

In the present research, we examined effects of RMA on information search guided by the premise that attitudes and belief systems can actively influence attentional processes (Nisbett & Miyamoto, 2005). To do so, we provided participants with the opportunity to select case-related information (Study 1) and to inspect photographs depicting witness and defendant of a rape case (Study 2). In both studies—with the first providing information in a highly standardized manner and the second doing so in a much more open form—we demonstrated an influence of rape-related attitudes on the type of information participants focused on. Taken together, our results suggest that RMA exerts a biasing effect well before the processing of any relevant information, already at the stage of information search. As predicted, participants with higher RMA (a) preferred information about the potential victim to information about the potential perpetrator and (b) shifted their visual interest away from the potential perpetrator to the potential victim. Such RMA-based information search preferences are consistent with research on the confirmation bias (Nickerson, 1998): Because most rape myths focus on the role of the woman in a potential rape case, a focus on the witness relative to the defendant is the possible result of an attitude-congruent information search for people who endorse rape myths.

In a typical rape case, this preference is easily satisfied because there is usually more information available about the alleged victim (who is usually the only eye witness to the events) than about the defendant (who may exert his right to silence and thus avoid to testify). According to participants’ expectations (Study 1), such a one-sided information search should not be detrimental to the alleged victim as witness-focused information (compared with defendant-focused information) is expected to speak in her favor (more defendant guilt; less contributory fault). However, the extant literature on how information in rape cases is interpreted suggests that a biased information search is likely to have the opposite effect in the end. As even information that is legally irrelevant such as the victims’ clothing (Whatley, 1996), attractiveness (Deitz et al., 1984), or housing and job situation (Eyssel & Bohner, 2011) is processed in a biased manner and leads to more victim blaming and more lenient case judgments, greater attention paid to the alleged victim might very well be detrimental to her case.

Even in the absence of biased processing, increased attention to the alleged victim may affect case judgments: Research on target salience (Taylor, Crocker, Fiske, Sprinzen, & Winkler, 1979; Taylor & Fiske, 1975) suggests that people have a strong tendency to attribute causality and control to salient actors. Similarly, illusory causation (Lassiter, 2002; Lassiter, Geers, Munhall, Ploutz-Snyder, & Breitenbecher, 2002) posits that individuals’ causal attributions are influenced by what captures their visual attention. Following this reasoning, Rempala and Bernieri (2005) found that greater victim salience in a rape case resulted in more lenient verdicts, presumably because the potential victim became the attributional target for the events under consideration. Whereas Rempala and Bernieri manipulated salience directly by providing more or less information, we have shown that participants’ RMA predicts a preference for information as well as an attentional focus on the potential victim relative to the potential perpetrator. Thus, our findings suggest that RMA may cause higher salience of the potential victim relative to the perpetrator. Future research should investigate the consequences of such attitude-congruent biased attention. In particular, it would be interesting to see whether high-RMA individuals’ preference for victim-focused information leads them to form even more biased judgments as a result. It should be addressed whether such an effect would already be evident in the presence of a lot but irrelevant information (as implied by a target salience account; but see Eyssel & Bohner, 2011) or whether it would need the presence of information that potentially fits a rape myth schema such as the victim wearing revealing apparel (and thus would require biased processing).

With regard to biased processing, Krahé and colleagues (2007) have shown that an accountability instruction, that is, asking participants to prepare for justifying their decisions, diminished rape-myth consistent processing of subsequent information. Future research should address whether such an intervention can also successfully reduce another aspect of the confirmation bias—biased information search. However, we suspect that biased information search might be more difficult to counter because, for one thing, it is not perceived as detrimental to the alleged victim. For another thing, rape cases might be particularly suited for such a bias to unfold. Because in actual rape cases more information is available on the victim than on the perpetrator (Madigan & Gamble, 1991), the nature of rape trials might provide a fertile ground for biased information search to unfold.

In conclusion, the present research adds to the growing body of research demonstrating individual differences in cognition and attention due to participant factors (e.g., Miyamoto, 2013). It also supports the argument that people’s attitudes toward sexual violence and rape not only influence related cognitive processes, as suggested by prior research (e.g., Eyssel & Bohner, 2011; Krahé et al., 2007), but also guide relevant attentional processes.

Footnotes

Acknowledgements

We thank Annika Jungwirth for her assistance in conducting this research.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported in part by a grant from the Deutsche Forschungsgemeinschaft to Gerd Bohner (BO 1248/6-1) and by the Cluster of Excellence Cognitive Interaction Technology ‘CITEC’ (EXC 277) at Bielefeld University, which is funded by the German Research Foundation (DFG).

Notes

Author Biographies

Philipp Süssenbach is a post-doc researcher at the department of psychology, Philipps-University Marburg, Germany. He holds a PhD in Psychology from the University of Bielefeld and teaches in the fields of research methods and social psychology. He conducts basic and applied research in social psychology.

Friederike Eyssel is a Visiting Associate Professor of Psychology at New York University, Abu Dhabi. Further, she leads the Research Group “Gender and Emotion in Cognitive Interaction“ at the Center of Excellence in Cognitive Interaction Technology (CITEC) at Bielefeld University. Dr. Eyssel has earned her Masters Degree in psychology from University of Heidelberg in 2004, and has received a PhD in psychology from Bielefeld University in 2007. She has held visiting professorships for social psychology at the University of Münster, TU Dortmund, and University of Cologne. In her work, Dr. Eyssel merges experimental social psychology and a more applied research focus. With her team, she exploresunderpinnings of gender prejudice, attitude measurement and change,and prejudice reduction. She is also interested in psychological aspects of human-machine interaction as well as social and ethical implications associated with the development and use of smart home environments.

Jonas Rees has studied Applied Social Psychology and Psychology at the University of Sussex and Bielefeld University where he completed his PhD in 2015. He currently holds a post-doc position at Bielefeld University’s Center for Interdisciplinary Research. His research interests include intergroup processes and discrimination, political and environmental attitudes, and collective action and protest.

Gerd Bohner, Dr. phil. (Heidelberg 1990) habil. (Mannheim 1997), is a social psychologist. After his graduation in 1986 he held academic positions at Heidelberg, Mannheim, Canterbury (UK), and Würzburg, as well as a postdoctoral fellowship at New York University. Currently (since 2001) he is Professor of Social Psychology at the University of Bielefeld. His research interests include basic and applied aspects of attitudes and social cognition, gender, and sexual aggression.

References

Abrams

Viki

G. T.

Masser

B. M.

Bohner

(2003). Perceptions of stranger and acquaintance rape: The role of benevolent and hostile sexism in victim blame and rape proclivity. Journal of Personality and Social Psychology, 84, 111-125. doi:10.1037/0022-3514.84.1.111

Bassok

Trope

(1984). People’s strategies for testing hypotheses about another’s personality: Confirmatory or diagnostic? Social Cognition, 2, 199-216. doi:10.1521/soco.1984.2.3.199

Berkowitz

A. D.

(2003). Fostering men’s responsibility for preventing sexual assault. In Schewe

P. A.

(Ed.), Preventing violence in relationships (pp. 163-196). Washington, DC: American Psychological Association.

Bohner

(1998). Vergewaltigungsmythen [Rape myths]. Landau, Germany: Verlag Empirische Pädagogik.

Bohner

(2001). Writing about rape: Use of the passive voice and other distancing text features as an expression of perceived responsibility of the victim. British Journal of Social Psychology, 40, 515-529. doi:10.1348/014466601164957

Bohner

Eyssel

Pina

Siebler

Viki

G. T.

(2009). Rape myth acceptance: Cognitive, affective and behavioural effects of beliefs that blame the victim and exonerate the perpetrator. In Horvath

Brown

(Eds.), Rape: Challenging contemporary thinking (pp. 17-45). Cullompton, UK: Willan.

Bohner

Lampridis

(2004). Expecting to meet a rape victim affects women’s self-esteem: The moderating role of rape myth acceptance. Group Processes & Intergroup Relations, 7, 77-87. doi:10.1177/1368430204039974

Bohner

Pina

Viki

G. T.

Siebler

(2010). Using social norms to reduce men’s rape proclivity: Perceived rape myth acceptance of out-groups may be more influential than that of in-groups. Psychology, Crime & Law, 16, 671-693. doi:10.1080/1068316X.2010.492349

Bohner

Siebler

Schmelcher

(2006). Social norms and the likelihood of raping: Perceived rape myth acceptance of others affects men’s rape proclivity. Personality and Social Psychology Bulletin, 32, 286-297. doi:10.1177/0146167205280912

10.

Bohner

Weisbrod

Raymond

Barzvi

Schwarz

(1993). Salience of rape affects self-esteem: The moderating role of gender and rape myth acceptance. European Journal of Social Psychology, 23, 561-579. doi:10.1002/ejsp.2420230603

11.

Bruner

J. S.

(1957). Going beyond the information given. In Bruner

J. S.

Brunswik

Festinger

Heider

Muenzinger

K. F.

Osgood

C. E.

Rapaport

(Eds.), Contemporary approaches to cognition (pp. 41-69). Cambridge, MA: Harvard University Press.

12.

Burt

M. R.

(1980). Cultural myths and support of rape. Journal of Personality and Social Psychology, 38, 217-230. doi:10.1037/0022-3514.38.2.217

13.

Burt

M. R.

(1991). Rape myths and acquaintance rape. In Parrot

Bechhofer

(Eds.), Acquaintance rape: The hidden crime (pp. 26-40). New York, NY: Wiley.

14.

Burt

M. R.

Katz

B. L.

(1988). Coping strategies and recovery from rape. Annals of the New York Academy of Sciences, 528, 345-358. doi:10.1111/j.1749-6632.1988.tb50876.x

15.

Cameron

C. A.

Stritzke

W. G. K.

(2003). Alcohol and acquaintance rape in Australia: Testing the presupposition model of attributions about responsibility and blame. Journal of Applied Social Psychology, 33, 983-1008. doi:10.1111/j.1559-1816.2003.tb01935.x

16.

Cowan

Quinton

(1997). Cognitive style and attitudinal correlates of the perceived causes of rape scale. Psychology of Women Quarterly, 21, 227-245. doi:10.1111/j.1471-6402.1997.tb00110.x

17.

Deitz

Littman

Bentley

(1984). Attribution of responsibility for rape: The influence of observer empathy, victim resistance, and victim attractiveness. Sex Roles, 10, 261-280. doi:10.1007/BF00287780

18.

Devine

P. G.

Hirt

E. R.

Gehrke

E. M.

(1990). Diagnostic and confirmation strategies in trait hypothesis testing. Journal of Personality and Social Psychology, 58, 952-963. doi:10.1037/0022-3514.58.6.952

19.

Eyssel

Bohner

(2011). Schema effects of rape myth acceptance on judgments of guilt and blame in rape cases: The role of perceived entitlement to judge. Journal of Interpersonal Violence, 26, 1579-1605. doi:10.1177/0886260510370593

20.

Eyssel

Bohner

Siebler

(2006). Perceived rape myth acceptance of others predicts rape proclivity: Social norm or judgmental anchoring? Swiss Journal of Psychology, 65, 93-99. doi:10.1024/1421-0185.65.2.93

21.

Foubert

J. D.

Marriott

K. A.

(1997). Effects of a sexual assault peer education program on men’s belief in rape myths. Sex Roles, 36, 259-268. doi:10.1007/BF02766271

22.

Frese

Moya

Megias

J. L.

(2004). Social perception of rape: How rape myth acceptance modulates the influence of situational factors. Journal of Interpersonal Violence, 19, 143-161. doi:10.1177/0886260503260245

23.

Gerger

Kley

Bohner

Siebler

(2007). The acceptance of modern myths about Sexual Aggression Scale: Development and validation in German and English. Aggressive Behavior, 33, 422-440. doi:10.1002/ab.20195

24.

Helmke

Kobusch

P.-R.

Rees

J. H.

Meyer

Bohner

(2014). Beliefs about the Strauss-Kahn case in France and Germany: Political orientation and sexual aggression myths as local versus global predictors (urn:nbn:de:0070-ijcv-2014167). International Journal of Conflict and Violence, 8, 171-186. Retrieved from http://www.ijcv.org/index.php/ijcv/article/view/375/pdf_106

25.

Holmqvist

Nyström

Andersson

Dewhurst

Halszka

van de Weijer

(2011). Eye tracking: A comprehensive guide to methods and measures. Oxford, UK: Oxford University Press.

26.

Jacobson

Popvich

(1983). Victim attractiveness and perceptions of responsibility in an ambiguous rape case. Psychology of Women Quarterly, 8, 100-104. doi:10.1111/j.1471-6402.1983.tb00621.x

27.

Jarvis

B. G.

(2005). MediaLab (Version 2006) [Computer software]. New York, NY: Empirisoft.

28.

Klayman

(1995). Varieties of confirmation bias. Psychology of Learning and Motivation, 32, 385-418. doi:10.1016/S0079-7421(08)60315-1

29.

Krahé

(1991). Social psychological issues in the study of rape. European Review of Social Psychology, 2, 279-309.

30.

Krahé

Temkin

Bieneck

(2007). Schema-driven information processing in judgements about rape. Applied Cognitive Psychology, 21, 601-619. doi:10.1002/acp.1297

31.

Lassiter

G. D.

(2002). Illusory causation in the courtroom. Current Directions in Psychological Science, 11, 204-208. doi:10.1111/1467-8721.00201

32.

Lassiter

G. D.

Geers

A. L.

Munhall

P. J.

Ploutz-Snyder

R. J.

Breitenbecher

D. L.

(2002). Illusory causation: Why it occurs. Psychological Science, 13, 299-305. doi:10.1111/j.0956-7976.2002.x

33.

Littleton

H. L.

Axsom

Radecki Breitkopf

Berenson

A. B.

(2006). Rape acknowledgement and post assault behaviors: How acknowledgment status relates to disclosure, coping, worldview, and reactions received from others. Violence and Victims, 21, 761-778. doi:10.1891/vv-v21i6a006

34.

Lonsway

K. A.

Fitzgerald

L. F.

(1994). Rape myths: In review. Psychology of Women Quarterly, 18, 133-164. doi:10.1111/j.1471-6402.1994.tb00448.x

35.

Madigan

Gamble

N. C.

(1991). The second rape: Society’s continued betrayal of the victim. New York, NY: Lexington Books.

36.

Miyamoto

(2013). Culture and analytic versus holistic cognition: Toward multilevel analyses of cultural influences. Advances in Experimental Social Psychology, 47, 131-188. doi:10.1016/B978-0-12-407236-7.00003-6

37.

Miyamoto

Nisbett

R. E.

Masuda

(2006). Culture and the physical environment: Holistic versus analytic perceptual affordances. Psychological Science, 17, 113-119. doi:10.1111/j.1467-9280.2006.01673.x

38.

Nickerson

R. S.

(1998). Confirmation bias: A ubiquitous phenomenon in many guises. Review of General Psychology, 2, 175-220.

39.

Nisbett

R. E.

Miyamoto

(2005). The influence of culture: Holistic versus analytic perception. Trends in Cognitive Science, 9, 467-473. doi:10.1016/j.tics.2005.08.004

40.

Payne

D. L.

Lonsway

K. A.

Fitzgerald

(1999). Rape myth acceptance: Exploration of its structure and its measurement using the Illinois Rape Myth Acceptance Scale. Journal of Research in Personality, 33, 27-68. doi:10.1006/jrpe.1998.2238

41.

Rempala

D. M. F.

Bernieri

F. M.

(2005). The consideration of rape: The effect of target information disparity on judgments of guilt. Journal of Applied Social Psychology, 35, 536-550. doi:10.1111/j.1559-1816.2005.tb02134.x

42.

Schuller

R. A.

Wall

A.-M.

(1998). The effect of defendant and complainant intoxication on mock jurors’ judgments of sexual assault. Psychology of Women Quarterly, 22, 555-573. doi:10.1111/j.1471-6402.1998.tb00177.x

43.

SMI. (2009). iView X System Manual Version 2.4 [Computer software]. Berlin, Germany: Author.

44.

SMI. (2010a). BeGaze 2.4 [Computer software]. Berlin, Germany: Author.

45.

SMI. (2010b). Experiment Center 2 [Computer software]. Berlin, Germany: Author.

46.

Süssenbach

Bohner

(2011). Acceptance of sexual aggression myths in a representative sample of German residents. Aggressive Behavior, 37, 374-385. doi:10.1002/ab.20390

47.

Süssenbach

Bohner

Eyssel

(2012). Schematic influences of rape myth acceptance on visual information processing: An eye-tracking approach. Journal of Experimental Social Psychology, 48, 660-668. doi:10.1016/j.jesp.2011.12.005

48.

Süssenbach

Eyssel

Bohner

(2013). Metacognitive aspects of rape myths: Subjective strength of rape myth acceptance moderates its effects on information processing and behavioral intentions. Journal of Interpersonal Violence, 28, 2250-2272. doi:10.1177/0886260512475317

49.

Taylor

S. E.

Crocker

Fiske

S. T.

Sprinzen

Winkler

J. D.

(1979). The generalizability of salience effects. Journal of Personality and Social Psychology, 37, 357-368. doi:10.1037/0022-3514.37.3.357

50.

Taylor

S. E.

Fiske

S. T.

(1975). Point of view and perceptions of causality. Journal of Personality and Social Psychology, 32, 439-445. doi:10.1037/h0077095

51.

Temkin

Krahé

(2008). Sexual assault and the justice gap: A question of attitude. Oxford, UK: Hart.

52.

Vygotsky

L. S.

(1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press. (Original work published 1930-1934)

53.

Whatley

M. A.

(1996). Victim characteristics influencing attributions of responsibility to rape victims: A meta-analysis. Aggression and Violent Behavior, 1, 81-95. doi:10.1016/1359-1789(95)00011-9