Attributions of Blame in Stranger and Acquaintance Rape: A Multilevel Meta-Analysis and Systematic Review

Abstract

Background: Victim blame, particularly in cases of acquaintance rape, presents an obstacle to criminal justice. Past research indicates that acquaintance rape results in more blame than stranger rape. However, there are inconsistencies in these findings (e.g., whether there is a linear relationship between victim blame and relationship closeness), partly due to methodological variation. Objectives: To examine the effect of victim–perpetrator relationship on victim blame, how this effect is impacted by rape myth acceptance (RMA) and ambivalent sexism (AS), and to establish what the methodological quality is of studies. Synthesis method: Studies were synthesized through a multilevel meta-analysis using the Metafor package in R (version 2.4-0), synthesizing findings from 47 individual studies. Studies compared victim blame between stranger and acquaintance rape, in isolation or in conjunction with RMA and AS, and were identified through a database search. Results: The review found higher levels of blame in acquaintance as compared to stranger rape, with a medium effect size. This effect was not moderated by RMA. AS was not included as a moderator in the meta-analysis, but the review indicated that benevolent sexism may be a particularly relevant variable. Implications: Future research should examine the relationship between AS and victim blame. The current review contributes to the evidence base on victim blame in rape cases by suggesting that methodological limitations can account for some of the past mixed findings in this area, particularly in a lack of consistency in vignette details. It is recommended that future sexual assault research uses rigorous methodology and increases transparency of research processes.

Blame, Control, or Responsibility? Validation of Sexual Assault Culpability Measures

“…The robust maximum likelihood estimation (MLR: Yuan & Bentler, 1997) was used due to the non-normality of our data (as determined through examining histograms). It is generally expected that data on victim and perpetrator blame demonstrate non-normality, as levels of victim blame tend to be low, and levels of perpetrator blame tend to be high (Persson & Dhingra, 2020;, and this was the case with the current dataset. Further details on the spread of the data can be found in Persson & Dhingra (2021).…”

“…Much previous research has measured various aspects of victim and perpetrator blame, but as we have highlighted in some of our past research (e.g., Persson & Dhingra, 2020;, there has at times been a lack of clarity in the conceptualisation of victim blame.…”

“…Items of blame tapped into beliefs that the victim/perpetrator should be held accountable for the assault; items of responsibility tapped into beliefs that the victim/perpetrator's choices or actions contributed to the assault; and items of control tapped into beliefs that the victim/perpetrator had the power to influence or direct people's behaviour or the course of events. Some items were adapted from other research (Abrams et al, 2003;Ayala et al, 2018;Bieneck & Krahe, 2011;Davies & McCartney, 2003;Franklin & Garza, 2018), and others were created following a comprehensive literature review (Persson & Dhingra, 2020).…”

“…All materials (including the vignettes developed and validated in conjunction with the scales) that were developed for this article can be found in repositories on the Open Science Framework (https://osf.io/92bq7/) as part of the larger, previously published research project (Persson & Dhingra, 2020), where this dataset is from. In addition, we also include links to repositories where the full datasets will be posted once all analyses are complete (https://osf.io/92bq7/), and the code that was used for the analysis.…”

“…In other words, "victim blame" and "perpetrator blame" have been used in a rather broad sense, with considerable variation in the exact items used to measure these attributions. In our previous review of the literature (Persson & Dhingra, 2020), we note that items used to index "victim blame" vary substantially. Some researchers have assessed blame, while others have assessed culpability, responsibility, guilt, and control.…”