Thinspiration: Self-Improvement Versus Self-Evaluation Social Comparisons with Thin-Ideal Media Portrayals

Abstract

Much research has demonstrated negative impacts of idealized-body imagery exposure on body satisfaction. Yet, paradoxically, media with such imagery attract mass audiences. Few studies showed women’s body satisfaction increased due to thin-ideal exposure. The kind of social comparison women engage in (self-evaluation vs. self-improvement) may explain these inconsistent findings and the paradoxical attraction to thin-ideal messages. Across 5 days, thin-ideal messages were presented to 51 women; self-evaluation and self-improvement social comparisons as well as body satisfaction were measured each day. A linear positive change in body satisfaction emerged. Greater self-improvement social comparisons increased this change, whereas greater self-evaluation social comparisons reduced it. Extent of both social comparison types changed during the prolonged exposure. A greater tendency to compare one’s body with others’ improved body satisfaction through self-improvement social comparisons and fostered weight-loss behaviors through self-evaluation social comparisons.

ACKNOWLEDGMENTS

The author thanks Jodi Whitaker for her assistance with an earlier version of this article and Melissa Kaminski for her help with finalizing the article.

Notes

¹ The labeling of the present work as “longitudinal” deserves elaboration: Other work on media effects in the body image context that used more than one data collection session did not specifically expose participants to stimuli and instead used survey measures. An interesting exception is a study by Stice, Spangler, and Agras (2001), which utilized a 15-month subscription to a fashion magazine with adolescent American girls—actual exposure, however, was self-reported. Research on social comparison with idealized body shapes in the media has generally worked with single-session designs and much fewer stimuli: According to Want’s (2009) meta-analysis of 47 studies, the use of about 10 images or 5 minutes of exposure is typical in this research domain. Another meta-analysis (Hausenblas et al., 2013) that did not have a social comparison focus showed that studies with more than 25 stimuli are very rare—only three unpublished works with 28, 30, or 60 stimuli were identified among 33 experimental studies. In contrast, the present study used five actual media stimuli exposure sessions, which entailed viewing 80 magazine pages.

² Only a handful of studies, which all used participants in the child to adolescent age range, examined long-term effects of exposure to idealized images or media exposure in general (Dohnt & Tiggemann, 2006; Harrison & Hefner, 2006; Schooler, & Trinh, 2011; Stice, Spangler, & Agras, 2001; Tiggemann, 2006), with inconsistent and conflicting results. Further, Hargreaves and Tiggemann (2003a, 2003b) examined the role of responsiveness for exposure effects across a 2-year time span as well as appearance–schema activation with adolescent samples. A study with adult women on delayed effects of ideal imagery exposure was done by Hausenblaus, Janelle, and Gardner (2004): Their sample of 30 college-aged women viewed full-body pictures slides, which depicted either the individual participants themselves (eight “self slides” with different-angle shots) or a model representing the thin ideal (with body mass index [BMI] at 19.75, 16.7% body fat), as eight “model” slides with different-angle shots. However, this exposure context presents a stark contrast to everyday encounters with thin ideal messages in the media.

³ Groesz et al. (2002, p. 7) had “expected larger doses of exposure (e.g., 25 thin media images) to have a more negative effect than smaller amounts of stimuli presented (e.g., five images)” but found a significant effect that contradicted this hypothesis: “as the number of stimuli presented expands, participants are less affected by the presentation of thin media images.”

⁴ Holmstrom (2004, p. 209) specified this pattern as follows: “For experimental studies—in which ‘length of time’ denotes the amount of time participants are exposed to images during experimental procedures (in this case, from 45 seconds to 26 minutes)—the relationship between effect size and length of exposure is similar to the overall correlation between these variables (r = –.07). For surveys—in which ‘length of time’ refers to the participants’ reports of weekly media exposure in hours—the correlation between effect size and length of exposure is negative but much larger (r = –.39). This correlation suggests that the more time participants report viewing media, the better they feel about their bodies.”

⁵ An additional six participants did not complete the study and were not included in analyses.