Response to variation in pro-social incentives

When starting the task, participants will be informed about the company, what a slogan is and the task. The task is then described as: We have in total 21 products. Your task is to create at least three slogans. One slogan for each of three different products. But of course, we appreciate if you decide to create slogans for more than three products.
The products will be shown to you in sequence. The next product will be shown to you only after you created the slogan for the previous product. For each product you will see a detailed description and a picture. Underneath the product description you will find a blank box. In that blank box you can write the slogan(s). Once you are sure about the slogan please select “submit”.
Participants will then get randomized into one of 9 treatments:
1. Baseline treatment: The task is considered completed if you create 3 slogans. Upon completion of the task, you will be paid the pre-announced wage of $1.50. However, we would really appreciate if you could create at least 6 slogans.
2. Unconditional donation treatment: The task is considered completed if you create 3 slogans. Upon completion of the task, you will be paid the pre-announced wage of $1.50. Furthermore, we will make a donation of 75 cents to Doctors without Borders (US). This donation is made unconditionally on whether you create any extra slogans. However, we would really appreciate if you could create at least 6 slogans. We will post the receipt of the donation online for you to verify 14 days after we collected all the slogans (and donations). Totally anonymous.
3. Unconditional donation treatment (with explanation): The task is considered completed if you create 3 slogans. Upon completion of the task, you will be paid the pre-announced wage of $1.50. Furthermore, we will make a donation of 75 cents to Doctors without Borders (US). This donation is made unconditionally on whether you create any extra slogans. However, we would really appreciate if you could create at least 6 slogans. We will post the receipt of the donation online for you to verify 14 days after we collected all the slogans (and donations). Totally anonymous. Why do we make a donation? We are strongly committed to be a socially responsible company (e.g. helping the larger community), even if this implies sacrificing some profit.
4. Conditional donation treatment: The task is considered completed if you create 3 slogans. Upon completion of the task, you will be paid the pre-announced wage of $1.50. Furthermore, if you create (at least) 6 slogans instead of 3, we will make a donation of 75 cents to Doctors without Borders (US). This donation is made conditionally on you creating at least three extra slogans. Indeed, we would really appreciate if you could create at least 6 slogans. We will post the receipt of the donation online for you to verify 14 days after we collected all the slogans (and donations). Totally anonymous.
5. Conditional donation treatment (with explanation): The task is considered completed if you create 3 slogans. Upon completion of the task, you will be paid the pre-announced wage of $1.50. Furthermore, if you create (at least) 6 slogans instead of 3, we will make a donation of 75 cents to Doctors without Borders (US). This donation is made conditionally on you creating at least three extra slogans. Indeed, we would really appreciate if you could create at least 6 slogans. We will post the receipt of the donation online for you to verify 14 days after we collected all the slogans (and donations). Totally anonymous. Why do we make the donation? The donation is profitable for us: you doing (at least) three extra slogans for a donation of 75 cents is less costly for us than hiring another worker to do (at least) three slogans for a wage of $1.50.
6. Unconditional bonus treatment: The task is considered completed if you create 3 slogans. Upon completion of the task, you will be paid the pre-announced wage of $1.50. Furthermore, you will receive a bonus of 75 cents. This bonus is given unconditionally on whether you create any extra slogans. However, we would really appreciate if you could create at least 6 slogans.
7. Unconditional bonus treatment (with explanation): The task is considered completed if you create 3 slogans. Upon completion of the task, you will be paid the pre-announced wage of $1.50. Furthermore, you will receive a bonus of 75 cents. This bonus is given unconditionally on whether you create any extra slogans. However, we would really appreciate if you could create at least 6 slogans. Why do we give you the bonus? We are strongly committed to be an employee-friendly company, even if this implies sacrificing some profit.
8. Conditional bonus treatment: The task is considered completed if you create 3 slogans. Upon completion of the task, you will be paid the pre-announced wage of $1.50. Furthermore, if you create (at least) 6 slogans instead of 3, you will receive a bonus of 75 cents. This bonus is given conditionally on you creating at least three extra slogans. Indeed, we would really appreciate if you could create at least 6 slogans.
9. Conditional bonus treatment (with explanation): The task is considered completed if you create 3 slogans. Upon completion of the task, you will be paid the pre-announced wage of $1.50. Furthermore, if you create (at least) 6 slogans instead of 3, you will receive a bonus of 75 cents. This bonus is given conditionally on you creating at least two extra slogans. Indeed, we would really appreciate if you could create at least 6 slogans. Why do we give you the bonus? The bonus is profitable for us: you doing (at least) three extra slogans for a wage of 75 cents is less costly for us than hiring another worker to do (at least) three slogans for a wage of $1.50.

We then ask the workers: How many slogans do you want to create? 3 Slogans OR 6 Slogans

Workers are shown the products in sequence. Once they create three slogans, one for each of three different products, they get to the next page.

We thank them for completing the task. Depending on the treatment, we summary the payments (plus bonus or donation). Then we ask everybody: “Would you like to do any more slogans? We would really appreciate if you could create at least 3 more slogans. Yes OR No”

If they clicked “yes” in the previous page, they are shown again the slogans in sequence. After each new slogan created, they are shown a screen with sentence: “Thank you for creating this extra slogan. Would you like to create one more? [Yes/No]” Once they click “No” the experiment is over.

We end with a survey in which we ask:

1. Gender
2. How attractive would our company be as a potential employee? (Likert scale: 0-Not attractive at all - 10-Very attractive)
3. Please rate PharmaGIC on a scale from 0 "Not at all socially responsible" to 10 "Very socially responsible": [0-Not at all socially responsible  -  10-Very socially responsible]
4. How satisfied were you with the incentives that we provided for this task? [0-Not at all satisfied-  10-Very satisfied]
5. "What do you think:
- How many other MTurkers did more than 3 slogans? (0-100%)
- How many other MTurkers did more than 6 slogans? (0-100%)
- How many slogans did the others do on average? (3-21)
- How motivating are the incentives for other MTurkers?[0-Not at all satisfied-  10-Very satisfied]"
6. "How calculated do you think was our choice of incentives?" [0-Not at all calculated-  10-Very calculated]
7. How often do you donate money to a charitable organization? Never, Rarely, Sometimes, Often, Regularly
8. How often do you volunteer for a good cause? Never, Rarely, Sometimes, Often, Regularly
9. Let us know your opinion about our task and our incentives: (leave empty box to insert text)

2017-01-30

2017-02-14

The key outcome variable is the number of slogans created by each subject. In particular, we will look at i) the average number of slogans per treatment; ii) the percentage of subjects who created more than 3 slogans in each treatment (3 slogans was the minimum number to consider the task completed and thus to be paid); iii) the percentage of subjects who created more than 6 slogans in each treatment (6 slogans was the minimum number required to make the donation in the strategic CSR treatments). Additionally, for each subject, we store data on how much time they took to create each slogan. We could also analyze the ‘quality’ of the slogans. Finally, we will look at responses from a survey regarding their perception about the company (e.g., how attractive that company is for a potential employee)

We vary whether the donation is made unconditionally on the performance of workers (non-strategic CSR), or whether it is made conditionally on workers’ creating more slogans (strategic CSR). We also vary the statements that explain why the donation is made. Finally, we implement the same treatment variations for standard incentives that take the form of a monetary bonus given to the workers, which can be conditional or unconditional to their performance.

Subjects will choose to participate in this study by selecting it on Amazon's Mechanical Turk service. Before the participants choose to participate, they will be provided with a very short brief description of the task to be completed; which also tells a guaranteed flat-pay of $ 1.50 for successful submission and a time estimate for completion of 10 minutes. After this, participants will be directed to complete a task with randomly-selected incentive structures. The task is to create slogans for the products promoted by PharmaGIC. There are 9 different randomly-assigned treatments with varying types of incentives. Each participant only faces one type of incentive (between-subjects design) All participants are clearly informed of their incentive structure before playing their task.
Upon completion of the task, participants will be thanked for their contribution and the flat payment of $1.50, along with any additional bonus, will be given within the designated time period for their treatment (typically within 48 hours). Those participants in the relevant treatments, will also receive a link in which we will upload the receipt of the donations generated during the intervention. Finally, all participants will be asked to answer a short survey concerning their perceptions about the company (e.g., how attractive is the company for potential employee, to what extent it is perceived as social responsible company, and so on…) and about some personal characteristics (e.g., gender, whether they regularly donate…). Please see the “intervention” section above for more details.

Number generation in Qualtrics will randomize participants into each treatment group subject to the guarantee of an equal number of participants in each group.

Individual subject

No

The number of clusters is the same as the number of observations, see below.

The number of clusters is the same as the number of observations. 2700 people is the ideal number of subjects planned for the study. The hope is to obtain at least 1800 subjects. The task is open from 9am to 9pm EST each weekday. The task will be kept open on Amazon Mechanical Turk until either (i) two weeks have passed or (ii) 2700 subjects have completed the study, whichever comes first. If two weeks pass without 1800 subjects completing the task, then the task will be kept open (up to six weeks) until 1800 subjects are obtained.

It was determined that we wanted to reach a sample of at least 200 subjects per treatment for a total of 200*9 = 1800 subjects so as to attain sufficiently precise estimates of the productivity per treatment. The ideal sample that we hope to achieve is approximately 1.5 times as large as that.

We ran a small pilot to ensure that our protocol was working and that we were able to generate variation in the creation of taglines. The pilot also allowed us to get an estimate for what the standard deviation will be in our task. Based on 230 pilot participants, the standard deviation of created slogans was around 2.23. Assuming that this is approximately the standard deviation of each treatment in the experiment and assuming a sample size of 1800 (200 per treatment), there is thus an 80% power to reject the null hypothesis of zero difference in average slogans between two treatments when the actual difference between the two treatments is 0.64 slogans. Assuming instead a sample size of 2700 (300 per treatment), there is then an 80% power to reject the null hypothesis of zero difference when the actual difference is 0.51 slogans. Based on our pilot, different treatments can create differences in average number of slogans by as much as 0.7-1.88 slogans, a difference of which can easily be detected statistically given the preceding calculations.