'The Role of Labels versus Appearances in Gender-Based Inferences' (AsPredicted #200,706)
Author(s) Jenna Alton (University of Maryland) - jalton@umd.edu Andrei Cimpian (New York University) - andrei.cimpian@nyu.edu Lucas Butler (University of Maryland) - lpbutler@umd.edu
Pre-registered on 2024/11/21 20:22 (PT)
1) Have any data been collected for this study already? It's complicated. We have already collected some data but explain in Question 8 why readers may consider this a valid pre-registration nevertheless.
2) What's the main question being asked or hypothesis being tested in this study? This study investigates the impact of gender labels ("girl," "boy") and gendered appearances (i.e., specifically one's gendered dress and hairstyle; feminine, masculine) on 4- and 5-year-old children's inductive inferences about gender-related attributes.
3) Describe the key dependent variable(s) specifying how they will be measured. Participants will be asked to rank characters (varying in gendered appearance and gender labels) in terms of whom they associate most with eight gender-related attributes (four masculine, four feminine). Four characters will be presented on each trial (i.e., for each attribute): a masculine boy, feminine boy, feminine girl, and masculine girl. There is one key dependent variable: children's rankings (1-4) of each character.
4) How many and which conditions will participants be assigned to? There are no between-subjects conditions. However, there are two within-subject variables: character gender label and character gendered appearance.
5) Specify exactly which analyses you will conduct to examine the main question/hypothesis. We will assess if children's rankings differ based on a character's gender label (i.e., "boy" vs. "girl") and gendered appearance (i.e., masculine vs. feminine). To do this we will conduct two linear regressions, one for each type of attribute (masculine and feminine). For both regressions, participants' character rankings will be the response/dependent variable. We will include four predictors and all interactions: character gender label, character gendered appearance, participant gender (i.e., boy vs. girl), and participant age (as a continuous variable). To account for the fact that each participant provides multiple responses (i.e., rankings), we will cluster the standard errors at the participant level. All predictors will be mean-centered.
As a robustness test, we will also run an ordinal logistic model with the same predictors and with standard errors clustered at the participant level.
If the model finds interaction effects, we will conduct the relevant marginal tests to interpret these interactions.
At the end of the sessions, children will rate the gendered appearance of all characters (i.e., whether each character looks more like a girl or more like a boy). We will use these ratings to verify our description of the characters as feminine vs. masculine in their gendered appearance.
6) Describe exactly how outliers will be defined and handled, and your precise rule(s) for excluding observations. We will use the following exclusion criteria:
- Parent or sibling interference
- Insufficient attention to the study
- Technical difficulties leading to loss of data or data integrity
- Experimenter error
- Non-binary participants will only be included in analyses that do not involve participant gender
7) How many observations will be collected or what will determine sample size? No need to justify decision, but be precise about exactly how the number will be determined. We plan to recruit a final sample of 46 children ages 4 and 5 years. Assuming some exclusions, we anticipate collecting data from approximately 70 participants.
8) Anything else you would like to pre-register? (e.g., secondary analyses, variables collected for exploratory purposes, unusual analyses planned?) We have so far collected data from 3 out of 46 participants to ensure that the procedure runs smoothly and that children can understand our instructions.
If we find significant main effects of both gender label and gendered appearance in our models above, we may compare the coefficients of these two variables to draw conclusions about which one has a more powerful influence on children's stereotyped attributions.
To obtain effect size estimates, we may standardize the dependent variable and any continuous predictors and re-run the regression models described above.
We may compare the data from this study with those from a similar study conducted on an older sample.
