'shame broadcasts social norms_study4' (AsPredicted #74589)
Author(s) This pre-registration is currently anonymous to enable blind peer-review. It has 2 authors.
Pre-registered on 2021/09/14 - 07:18 AM (PT)
1) Have any data been collected for this study already? No, no data have been collected for this study yet.
2) What's the main question being asked or hypothesis being tested in this study? The main question being asked: Does seeing someone express shame in response to a behavior make people less likely to engage in the behavior compared to seeing someone have a neutral reaction to the behavior?
Participants (Ps) see someone express shame or have a neutral reaction to using a Rhyme Booster in a supposed group task in which the group member who generates the most rhymes gets a larger bonus.
We predict that Ps will be less likely to use a Rhyme Booster when they see this other person express shame about using a Rhyme Booster compared to when they see this other person react neutrally to using a Rhyme Booster.
3) Describe the key dependent variable(s) specifying how they will be measured. The main dependent variable is whether Ps use a Rhyme Booster. A Rhyme Booster gives Ps an advantage in generating rhymes. All Ps see the target use a Rhyme Booster. That is, using the Rhyme Booster is the choice that is congruent with target's behavior. Foregoing the option to use a Rhyme Booster is doing the opposite of the target's behavior.
DV1 is P's binary choice to not use a Rhyme Booster or to use a Rhyme Booster.
4) How many and which conditions will participants be assigned to? This is a between-subjects design. There are two conditions: the target expressed shame, the target expressed a neutral emotion.
5) Specify exactly which analyses you will conduct to examine the main question/hypothesis. We will analyze the dependent variable using a test of independence:
Stata Syntax: tab condition DV1, chi
6) Describe exactly how outliers will be defined and handled, and your precise rule(s) for excluding observations. We will exclude all observations for any P who appears in the dataset more than once, determined by both duplicate P IDs and duplicate IP Addresses (if IP addresses are available). We will also exclude any observation that does not provide a valid P ID.
Ps must pass an AV check to access the study. Ps who fail this check twice are excluded from participating.
Ps answer six questions about the instructions. For our primary analysis, we will include only Ps who answered all six questions correctly. If a single question excludes more than 20% of all Ps who would otherwise be included in the analysis, we will drop the question as an inclusion criterion. In this scenario, Ps must answer the remaining five comprehension check questions correctly.
A P answers a comprehension check question correctly if they answer it correctly the first time they are asked. If Ps answer the question correctly when asked a second time, they are not included in the primary analysis.
We will exclude Ps who do not complete the study in full, as determined by not having a study progress score of 100.
At the end of the study, Ps are asked whether they had any technical issues. If Ps answer "yes," Ps are then asked whether these technical issues prevented them from watching the videos. We will exclude Ps who answer "Yes" to this latter question.
Ps answer a question about whether the target used a Rhyme Booster. Ps who do not report that the target used a Rhyme Booster will be excluded.
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 powered this study to detect a 10-percentage point difference between conditions with 90% power using the following command:
Syntax: power twoprop .5 .6, power(.9)
We found that we need 519 Ps per condition to detect an effect where alpha=.05. The study will be posted for 1050 Ps (525 per condition) to Amazon Mechanical Turk. Data collection will stop once the request is filled on the platform (which sometimes leads to slightly fewer or more Ps).
8) Anything else you would like to pre-register? (e.g., secondary analyses, variables collected for exploratory purposes, unusual analyses planned?) Additional Measures:
Ps are asked three questions prior to making their incentivized choice related to perceived injunctive norms and perceived descriptive norms.
Injunctive norms against using a Rhyme Booster
Q1. "In this group, is it more acceptable to use a Rhyme Booster or not to use a Rhyme Booster? (More acceptable to use a Rhyme Booster; More acceptable not to use a Rhyme Booster; Similarly acceptable to use and not to use a Rhyme Booster).
Q2 "In this group, is it more wrong/inappropriate to use a Rhyme Booster or not to use a Rhyme Booster? (More wrong/inappropriate to use a rhyme Booster; More wrong/inappropriate not to use a Rhyme Booster; Similarly wrong/inappropriate to use or not to use a Rhyme Booster).
We will use a multinomial logit on Q1 and Q2 separately predicted by condition assignment. Syntax: mlogit DV1a i.condition,
DV2: We also will aggregate Q1 and Q2 in the following way: We will score Ps as 1=selected "More wrong/inappropriate to use a Rhyme Booster" & selected "More acceptable not to use a Rhyme Booster;" 0 = all other choices. We will treat a score of 1 as seeing stronger injunctive norms against the target's choice. We will compare the percentage of 1s between condition using a chi-square test of independence: Syntax: tab condition DV2, chi
Descriptive norms about using a Rhyme Booster
Q3. Ps see pictures of the three non-target group members. For each non-target group member, Ps indicate whether they think the person used a Rhyme Booster or did not use a Rhyme Booster.
For each target, we will score the answer as 1=Did not use a Thyme Booster; 0=Used a Rhyme Booster." In DV3, we will sum the scores for each target to create an overall measure of descriptive norms about using a Rhyme Booster. We will compare mean differences in the scores between conditions using an independent samples t test or OLS regression: Syntax: reg DV3 i.condition, vce(robust)
We also will explore whether perceived injunctive norms and/or perceived descriptive norms relates to Ps' choice of using a Rhyme Booster using mediation analysis. Our mediation analysis will follow the procedure outlined here in Example 1: https://stats.idre.ucla.edu/stata/faq/how-can-i-analyze-multiple-mediators-in-stata/
Stimuli sampling and additional exploratory analyses
We vary the video that Ps watch across three different actors (one male and two female actors) for purposes of stimuli sampling. In our primary analysis, we will collapse across these videos.
After Ps answer DV1, they proceed to the task. We will test whether those who used a Rhyme Booster submitted more rhymes than those who did not use a Rhyme Booster.
