Effects of violent and nonviolent sexualized media on aggression‐related thoughts, feelings, attitudes, and behaviors: A meta‐analytic review

Abstract

Women are often depicted as sex objects rather than as human beings in the media (e.g., magazines, television programs, films, and video games). Theoretically, media depictions of females as sex objects could lead to negative attitudes and even aggressive behavior toward them in the real world. Using the General Aggression Model (Anderson & Bushman, 2002) as a theoretical framework, this meta-analytic review synthesizes the literature on the effects of sexualized media (both violent and nonviolent) on aggression-related thoughts, attitudes, and behaviors. Our sample includes 166 independent studies involving 124,236 participants, which yielded 321 independent effects. Overall, the effects were "small" to "moderate" in size (r = .16 [.14-.18]). Significant correlations were found in experimental, cross-sectional, and longitudinal studies, indicating a triangulation of evidence. Effects were stronger for violent sexualized media (r = .25 [.19-.31]) than for nonviolent sexualized media (r = .15 [.13-.17]), although the effects of nonviolent sexualized media were still significant and nontrivial in size. Moreover, the effects of violent sexualized media on aggression were greater than the effects of violent non-sexualized media on aggression obtained in previous meta-analyses. Effects were similar for male and female participants, for college students and non-students, and for participants of all ages. The effects were also stable over time. Sensitivity analyses found that effects were not unduly influenced by publication bias and/or outliers. In summary, exposure to sexualized media content, especially in combination with violence, has negative effects on women, particularly on what people think about them and how aggressively they treat them.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 60 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.