Author(s) This pre-registration is currently anonymous to enable blind peer-review. It has 4 authors.
Pre-registered on 2024/01/02 - 12:36 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? We have five main questions:
q1. Do emotional appeals and/or social pressure increase donations?
q2. Is there avoidance of emotional appeals and/or social pressure?
q3. Does social pressure generate more donations and more avoidance than emotional appeals?
q4. How does the justification of the donation choice influence the effect of social pressure on donations and avoidance?
3) Describe the key dependent variable(s) specifying how they will be measured. Participants face a real effort task to earn an endowment of 5 pounds. Afterwards, they are informed that they will have to decide between option A: keep the endowment and donate nothing to the charity Save the Children or option B: keep 1 pound and donate 8 pounds. Subsequently, they are informed that before making a decision between A or B they will watch a video. They may watch a described video (Treatment video) or they may choose to watch an unrelated video (Alternative video) of similar length. The Treatment video and its description vary by treatment (see below).
Individuals are also informed that the computer can override their preferred video and that they may thus end up watching the non-preferred video with a small (unrevealed) chance. After the video choice, and the random implementation, individuals make a donation choice.
Our key outcome variables are:
1. The proportion of donations (individuals who choose option B).
2. The proportion of individuals selecting the Alternative video.
4) How many and which conditions will participants be assigned to? Our experiment consists of 4x2 between-subject treatments. First, individuals are randomized into one of four video treatments. The alternative video for each of these treatments is a video on how grass grows.
-Emotional video Treatment: Individuals see an emotion inducing charity video from Save the Children.
- Neutral video Treatment: individuals see a neutral emotion video on how waves work.
-Social Pressure video Treatment: individuals see a recording from the Director of Fundraising from Save the Children UK requesting for a donation and are also asked to provide a justification for their donation decision. This justification is then sent to the Director of Fundraising.
- Social Pressure video Treatment no justification: as the Social Pressure video Treatment but individuals do not need to provide a justification for their donation decision.
In each of those video treatments subjects are randomized into two further conditions:
-Non-preferred video: 40% of the time participants receive the non-preferred video.
-Preferred video: 60% of the time individuals receive their preferred video.
5) Specify exactly which analyses you will conduct to examine the main question/hypothesis. Donations:
q1. The video increases donations in the Emotional and in the Social Pressure Treatment versus the Neutral Treatment.
Test: We estimate the following regression model:
Model1: Donation = at+ bt(Preferred Treatment video) + ct(Received Treatment video)
Where the outcome variable Donation is a dummy which is 1 if option B was selected, t={N,E,P} depending on the treatment for which the model is estimated. The independent variables are also dummies with self-explanatory names. We then perform Wald tests to compare the model coefficients across treatments. We expect that cE>cN for the Emotional versus Neutral Treatment comparison and cP>cN for the Social Pressure Treatment.
Avoidance:
q2. We expect more individuals choosing the Alternative video in the Emotional Treatment and Social Pressure Treatment versus the Neutral.
Test: We compare the proportion of participants choosing the alternative video in each treatment and do a Chi-square proportions test for the comparison between each charity related treatment and the Neutral.
Social Pressure versus Emotional appeals:
q.3. To evaluate whether Social Pressure has a stronger effect in terms of donations and avoidance than Emotional appeals we compare the overall proportion of donations and proportion that prefers the alternative video in each treatment using a Chi square test. In addition, we evaluate whether, controlling for differences in selection, the exposure to the social pressure video has a higher effect on donations than the exposure to the emotional video. We thus compare coefficients cE and cP from Model 1 (see above) using a Wald test. In secondary analysis, we will investigate the effect of overruling the video preferences in both treatments (see below).
Justification and social pressure:
q.4 We do the same three comparisons as in q3. but for the Social Pressure no justification versus the Social Pressure Treatment comparison.
6) Describe exactly how outliers will be defined and handled, and your precise rule(s) for excluding observations. Individuals who do not complete the experiment or fail to earn the endowment in the first stage of the experiment (which also acts as an attention check) will be excluded from our analysis. In addition, we have a 3-question attention check after the video visualization. With a probability of 1/100 we exclude participants with 2 or more wrong answers (this is known to the participants).
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. Sample size is determined using power calculations. We will collect 400 observations for the Emotional, Social Pressure and Neutral Treatment and 200 observations for the Social Pressure without justification Treatment.
8) Anything else you would like to pre-register? (e.g., secondary analyses, variables collected for exploratory purposes, unusual analyses planned?) We will also explore if the identified avoidance is strategic. For this, we estimate the following regression model for each treatment t:
Model 2: Donation = at + bt(Preferred Treatment video) + ct(Received Treatment video) + dt(Preferred Treatment X Received Treatment video)
We expect ct>0 and that dt<0 for t={E,P}. This means that the exposure effect of the video is stronger for those who prefer not to see it. In addition, we expect that cN and dN are approximately zero, that is, that there is no strategic avoidance in the Neutral Treatment (Wald test).
We will explore further evidence for sophistication:
-We will evaluate data on beliefs: do individuals correctly estimate average donations in different conditions?
-We will evaluate the answers to an emotional questionnaire and to questions on anticipated pressure and temptation. For example: do individuals anticipate higher pressure associated with the Social Pressure Treatment video? Is this more so when a justification is required? Are individuals who anticipate more pressure more likely to choose the Alternative video?
