Based on the original effect size (F = .13) we obtained in our meta-analysis (Blanken, Van de Ven, & Zeelenberg, 2013), we calculated with G*Power (Faul et al., 2009) that we would need at least 918 participants in our study to achieve 95% power. To achieve such a big sample size, we will run Study 2 on Amazon Mechanical-Turk. We will include at least 918 participants from the USA. Participants will receive $0.35 for participation. These participants will be randomly assigned to one of three conditions: A positive-trait condition, a neutral-trait condition and a control condition. We will stop data collection when 918 participants have correctly completed our study. To obtain a reliable sample, we will exclude participants with a rejection rate below 95%.</p>
The experiment exists of the original study materials of Sachdeva et al.’s Study 1 and Study 3 (Appendix 2). As explained below, we made to slight changes to these materials to increase the credibility of the online study (Appendix 4). The materials will be presented online in a Qualtrics survey program.
We adopted the cover story of the experiment; we will tell participants that the study is about general writing styles (instead of handwriting). Depending on assigned condition, participants will be exposed to nine positive trait words, negative trait words or neutral words and will be asked to re-type each word four times and to think about each word for 5-10 seconds. Subsequently, they will be instructed to type a short story about themselves including the words they just typed. Participants will be told that they have to do this seriously and if they do not seriously answer this question their HIT will be rejected. Subsequently, they will answer some neutral questions about the stories they just wrote and they will complete a short math-based filler-task. Subsequently, participants will complete both dependent variables from Sachdeva et al’s Study 1 (i.e., donating) and Study 3 (i.e., commons dilemma) in a counterbalanced order. In this way, we can also establish whether there are any order effects; we will test whether the licensing and compensation effects still occur after the first dependent variable.
The dependent variable from Study 1 consists of the amount of money (up to $10.00) participants are willing to donate to a worthy cause of their choice. Since we cannot use the cover story of the original experiment (‘Our lab usually asks participants if they would like to contribute to a worthy cause in an effort to increase social responsibility’), we made slight changes to the donation measure that Sachdeva et al. (2009) used. We will tell participants that 10 participants who will be randomly selected will win an additional $10 MTurk worker bonus next to their payment in return for their participation. They will then be asked that, in an effort to increase social responsibility, we would like to ask them that if they would be one of the winners, whether they are willing to donate a portion of the money to a cause of their choice. They will be asked to indicate the cause they would like to support and indicate the amount of money they would like to donate. They will be told that they have to pay this amount of money upon receiving a confirmation e-mail. The dependent variable from Study 3 is the commons dilemma, as described in the section of our Study 2. Finally, participants will complete the 7 self-presentation items from self-monitoring scale (Lennox & Wolfe, 1984), (Appendix 7). </p>
Participants will complete all study instructions and materials online. Prior to the experiment, participants will be asked to provide their informed consent. Next, participants will receive the instructions and complete the instruction check, the questionnaire and the demographic measures, respectively. At the end of the online experiment, participants will be debriefed and thanked for their participation.
Confirmatory Analyses plan
We will analyze the data in the same way as described in the confirmatory analysis sections of Study 1 and Study 2. To test whether there are any order effects for the dependent variables, we will include the order in which the dependent variables are presented as an extra predictor in our ANOVA models. In addition, we will request the descriptives and frequencies of the extra demographic variables level of education and family income and we will analyze whether gender or nationality have any effect on the amount of money participants are willing to donate or the amount of time participants are willing to run the filters and the secondary prosocial measures by means of a one-way ANOVA. If it turns out that one (or more) of these variables significantly (at the p = .05 level) affect willingness to donate or amount of time participants are willing to run the filters, these variables will be included as covariates in the main analyses of condition on willingness to donate. For syntax of the planned analyses, see Appendix 10.</p>
<p>Known differences from original study
• The original study was presented as a paper-and-pencil task in the lab, whereas the current study will be completed online without the presence of an experimenter. In both samples of participants, the donation is anonymous, and therefore we do not expect any differences between responses on the dependent variable.
• The original study was completed by a student sample from Northwestern University, whereas the current study will be completed by a more general population from the USA on Amazon Mechanical-Turk. Therefore, we included additional demographics (level of education and family income). As stated in the analyses section, we will control for the possibility that these variables affect the willingness to donate money dependent variable.
• Since we are planning to perform this study online, we made two slight changes to the original cover story. First, participants in the original study were told that the study is about their handwriting styles, whereas we will instruct the participants that the study is about general writing styles. Secondly, participants in the original study were told that the lab usually asks participants if they would like to contribute to a worthy cause in an effort to increase social responsibility, whereas we will ask participants that if they would be one of the $10 winners, whether they are willing to donate a portion of the money to a cause of their choice.
We do not expect any of these differences to influence the main results of our studies.</p>