The function of vertical and horizontal space to social group identity

ABSTRACT

Where an object or person is located in space can communicate important attributes, such as power, agency, or status. We theorized that people may use location to convey messages about social groups. In four studies, we examined whether women and men express ingroup bias or stereotypical bias in their placement of and memory for gendered objects. In Study 1, participants placed objects symbolizing their ingroup higher but not further left than objects symbolizing their outgroup. Vertical ingroup-bias emerged consistently in spatial placement (Studies 2 and 4) but not spatial memory (Studies 3 and 4). We discuss the influence of gender identity, the role of automaticity, and the value of vertical versus horizontal spatial location in communicating group bias.

KEYWORDS:

• Vertical location
• gender identity
• ingroup bias
• spatial agency bias

Acknowledgments

We would like to thank Stefano Andriolo for lending his expertise and constructing the web-based platforms for spatial placement and memory data collection. We would also like to thank Stephanie Lamme for photo-editing/creating the objects and backgrounds used in these tasks.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. We excluded participants who had incomplete data. Sixteen participants were excluded because they failed to complete the study. These participants did not follow the external link to complete the spatial placement task. We were therefore unable to run analyses including these participants’ data.

2. Results are similar regardless of the analytic method.

3. Magnet ratings from participants in Study 1 confirmed pretesting; feminine magnets were rated as significantly more feminine than masculine magnets (M_difference = 2.86, SD = 1.21), t(186) = 32.39, p <.001.

4. Significance testing with mixed effect models can be calculated several different ways. In this case, we estimated degrees of freedom using Satterthwaite’s procedures which are based on the number of participants rather than groups (Satterthwaite, 1946). Degrees of freedom may vary substantially within the same model based on which effect is evaluated. The Satterthwaite method has demonstrated reliability in mixed effect models with sufficient sample sizes and low Type I error rates (Manor & Zucker, 2004). See Kuznetsova, Brockhoff, and Christensen (2017) for more information on the implementation of this estimation procedure in R.

5. Including random slopes exceeded what the data could reliably estimate in models we report throughout the manuscript. We, therefore, have run all models with only random intercepts.

6. Seven participants were excluded for one of these two reasons: because they did not complete the placement portion of the study (n = 6), or because they identified as gender non-binary (n = 1).

7. Although multi-level models draw statistical power from both the sample size of participants and stimuli, this study and Study 3 are likely underpowered given the small number of stimuli. We address this power issue in Study 4.

8. Participants also completed several other exploratory measures. See Appendix for a list of these additional measures.

9. One person declined to provide their age; their data were included in analyses. Ten participants were excluded because they did not complete the study due to technical issues with the eye tracker.

10. Exploratory analyses were conducted on visual attention as part of a student’s thesis project. See Appendix for list of additional measures used in this thesis project.

11. The sample included 90 participants who identified as Asian/Pacific Islander, 45 who identified as Black/African American, 44 who identified as Hispanic/Latinx, 595 who identified as White/European American, 5 who identified as Native American, 3 who identified as Middle Eastern, and 21 who identified as Multiracial.

12. One person declined to provide their age and two people declined to provide their race; the rest of their data were still included in analyses. Twelve participants were excluded because they did not pass the attention checks. Specifically, they did not correctly plot their score on the graph during the identity threat manipulation (± 5 points) and failed to accurately recall their score (± 5 points). We reasoned that if people failed both of these checks that they were unlikely to have been paying attention to or have understood the experimental manipulation. See pre-registration for more information (https://osf.io/ez5vh/?view_only=82ccc390bcc54e5cb269a71b44b6a31a). We also excluded participants using the following pre-registered criteria: 20 identified as gender non-binary, 2 entered their gender incorrectly at the start of the study, 55 failed to complete a portion of the study (i.e., they did not follow the external link or started but did not complete the study), and 7 opted to exclude their data once they learned the purpose of the study.

Additional information

Funding

This research was supported by an NSF GRFP [DGE-1104602].