**Sample Size and Sample Demographics** Sample of approximately 800 participants will be recruited from a proprietary Internet panel of the opinion company STEM/MARK based in the Czech Republic. Note that sampling of participants for this and another study (available [here][1]) will be done jointly. Two third of participants accessing the web-based questionnaire will be randomly assigned to this study and one third of participants will be randomly assigned to the other study. As a result, the final number of participants in this study may differ from the target sample size (N = 800) due to random assignment. Sampling of participants The sample will be similar to the Czech adult population 18+ in terms of age, proportion of males and education groups but will not be representative. Participants will be invited to fill out an online questionnaire. Participants will be paid 28 CZK (equivalent of € 1.1) by STEM/MARK Company for completing the questionnaire. The questionnaire will not elicit personal information identifying individuals, and only participants’ identification numbers assigned by the opinion poll company will be known to researchers. True identity of participants will be unknown to researchers. Participation in the study will be completely voluntary. We will exclude as invalid responses from participants who will meet at least one of the following exclusion criteria: 1. Respondent does not correctly answer two attention checks (included in batteries of GEB items). 2. Respondent completes the questionnaire in extremely short, or in extremely long time (more than 3 SD from the average completion time measured separately for experimental and control groups). 3. Respondent drops out of the study either because s/he does not complete the questionnaire or because s/he explicitly expresses her/ his wish to drop out in the debriefing. 4. Respondent guesses the main hypothesis and/ or guesses that this study focuses on dishonest or immoral behavior. Blinded rating of hypotheses guesses will be used to identify such responses and the exclusion will be done before analysis of the data. Note that we will also run a robustness check to determine to what degree does failure of some participants to choose products (which effectively signals failure of the experimental manipulation) affect our results. We will conduct this check by including information about the purchase (either as an indicator of non-buyers or as price of chosen products) in interaction with store condition as additional terms in the regression model used to test our hypotheses. **Materials – Overview** *Initial moral behavior* This conceptual replication of Experiment 3 from Mazar and Zhong (2010) uses an online store similar to the one used in the original study to manipulate levels of initial moral behavior by involving participants in behavior usually perceived as moral (purchase of green products), or in less moral behavior (purchase of conventional products), whereas the control group does not purchase any goods in the store. Level of engagement in the moral behavior is manipulated indirectly by changing the share of ecological and conventional products offered in the store. Participants in the green store condition (high likelihood of moral behavior) select products they would wish to obtain from a list of products featuring predominantly green products (9 out of 12 products listed are green products), whereas participants in the conventional store condition select from a list of predominantly conventional products (9 out of 12 products are conventional). All items featured on the list of products are real products available on the Czech market and resemble the products used in the original study by Mazar and Zhong (2010). In addition, we selected items so that each green product and its conventional alternative have the same package size and could be found for a similar price on the Czech market. The list of products features a small picture of each product, a short description, and its price (see [green store][2] and [conventional store][3] screengrabs). Participants are invited to click on products to see a one-paragraph long description of each product (see example [here][4]). Environmental attributes of green products are mentioned on the list of products (e.g., "green", "ecological", "organic"), and also in the detailed descriptions. Stores in the two conditions are not labelled as ecological or conventional. Participants can choose one product from each item in the total value up to 450 CZK (equivalent of € 18) and they informed that "several participants" will be randomly drawn at the end of the study and will actually receive their chosen products. The online store used in this study differs from the one used in Mazar and Zhong (2010) in the following aspects: 1. Whereas both stores are computer-based, this study uses online setting (Mazar and Zhong's experiment was carried out in the lab). 2. Participants are told that "several respondents will be randomly drawn at the end of the study and will receive products they have chosen". In contrast, 1 in 25 participants received their selected products in the original study. 3. Participants who do not choose any products in our store are not reminded that they have not chosen their products before proceeding to the next task (we only show them the total value of products selected). Participants in Mazar and Zhong's (2010) study who do not select any products are reminded that they have not selected products before proceeding to the next task. 4. Key features of the store are same in the two studies but the visual presentation of the store is slightly different. 5. Products and their prices are different in the two studies although we tried to use similar items and prices to those used in Mazar and Zhong's study. *Cheating task (die casting task)* The die casting task used in this study to measure honesty is adapted from Jiang (2013; see also Mann et al., 2016). Participants are told that their goal in the task is to foresee what numbers will be rolled on a die. Specifically, they are told that they will choose in their mind either the upper (facing up) or the bottom side (facing down) of the die prior to throw. The virtual die is cast then and the throw determines their point gain (point gain equals the number on the side of the die they have chosen). However, participants indicate their choice of the side only after the die has been cast which gives them the opportunity to cheat and indicate the side of the die which bears the higher number even if they chose the other side initially). Participants are explained in the introduction to the task that "several participants" will be drawn at the end of the study who will receive monetary prize, with each point having value of 4 CZK (equivalent of 0.16 €). This gives them opportunity to win between 80 and 480 CZK (3.2 and 19.2 €). We also explain the payoff function in the introduction using an example of a die throw. The task has 20 trials, each of which consists of the following steps: (i) Participant is asked to choose either the upper or the bottom side of the die in his mind. (ii) Virtual die is cast. (iii) Numbers rolled on the upper and bottom side of the die are displayed and participants indicate which side they have chosen prior to die cast. (iv) Cumulative gain is displayed and the number of remaining trials. This task differs from the one used in Jiang's study (Jiang, 2013) particularly in the following features: (a) We do not test whether participants understood the payoff function. (b) We display both the upper and the bottom value rolled on the die (whereas the original study displays only the upper side value). (c) Cumulative point gain and the number of remaining trials is displayed after each trial (whereas the original study displayed only the total gain at the end of the study). (d) Only selected "several" participants (3 participants are selected at the end of the study) can win their prize in our study whereas all participants received their prizes in the original study. **Design** The study has mixed design. The between-subject component of the design consists of participants being randomly assigned to one of the experimental conditions: green store condition (participants select products from a list with majority of green products), conventional store condition (participants select products from a list with majority of conventional products), and control condition (participants are not shown a list of products and do not select any products). The within-subject component of the design consists of a random sequence of 20 trials of the die casting task (same for all participants). **Procedure** Participants are recruited from the proprietary internet panel and are invited to participate in a study that ostensibly focuses on opinion formation and reasoning. After accessing the online questionnaire and providing their implicit informed consent, participants are randomly assigned to one of the three experimental conditions. Participants in the green store and conventional store conditions choose products from the list of items. All participants then proceed to the die casting task. After completing the die casting task, participants in the green store and conventional store conditions answer a battery consisting of 8 items which assess their perception of products they have selected and another battery of 4 items adopted from PANAS which assess their affective state at the time of choosing the products. All participants then proceed to two batteries of 14 items (including two catch items) which assess their moral values of care and fairness (Graham, Haidt, & Nosek, 2008). After these two batteries, participants proceed to an unrelated 4-minut long experiment on costly signaling of altruistic behavior through choice products with green and socially-conscious labels (this experiment is described [here][5]). All participants then proceed to a 50-item GEB battery (Byrka, Kaiser, & Olko, 2017; Kaiser & Wilson, 2004) which assesses their level of environmental attitude. Next, participants provide their socio-demographics (gender, age, education level) and proceed to debriefing. The debriefing provides participants with the contact information and information about how the data from the study will be used. Participants can then decide to drop out from the study, in which case all their responses are discarded (except for the socio-demographics that are used to identify differences between those who dropped and those who did not). All participants are then thanked and those who agreed to be included in the study are redirected to the web page of the opinion poll company for remuneration processing. [1]: https://osf.io/4jn9h/ [2]: https://osf.io/jx4qs/ [3]: https://osf.io/9b4rd/ [4]: https://osf.io/2evzx/ [5]: https://osf.io/p9czs/