Replication of “On the motivational power of pride: a field study with participants in a road–running event”

We conduct a field experiment with participants in three UK road-running races: the Great Scottish Run Half Marathon (29th September 2019), the Great Birmingham Run (13th October 2019) and the Oxford Half (13th October). We start recruitment from runners registered for the Scottish run (15,000 expected participants) and if necessary we will complete recruitment with the Birmingham event (10,000 expected participants) and Oxford event (7,000 expected participants).

We administer 3 online surveys: the Registration form, the Pre race survey and the Post race survey.

2019-09-26

2019-11-30

- Performance measures: (1) an indicator for whether the participant finished at or under the goal time entered in the pre-race survey and (2) the deviation of chip time from goal time in percentage terms.

- Expected performance: (1) a participant's belief entropy about the likelihood of meeting their goal and (2) their expected finish time.

- Self reported measures of pride: (1) directly reported "pride in the race" and (2) constructed variables "positive pride" and "negative pride" using the Tracy and Robins psychological pride scale (2007).

- Donation decision: an indicator for whether a participant who was in one of the Incentive treatments chose to donate their winnings to charity.

- Beliefs in the Pre race survey: participants' beliefs about their chances of meeting their goal before and after the treatment manipulation.

- Beliefs in the Post race survey: runners guesses of the percentage of study participants who met their goal in (i) the treatment they were in and (ii) one other treatment.

The experiment has a two by two design, with variations on the message runners receive about their goal time, and the incentive levels they are offered to meet their goal time.

Further details will be available on completion of the study (and are fully outlined in the PAP).

In the Pre race survey, runners are randomized into one of 4 different treatments, which differ in whether (i) we express doubt about their ability to meet their time goal in the race; (ii) we offer them monetary incentives for meeting their goal.

The groups are Control (C), Incentives (I), Pride Message (M) and Pride Message & Incentives (MI). The monetary incentives in the I and PI treatments are cross randomized to be either £10, £20 or £40 Amazon vouchers.

- C: "Good luck with the race! Don't forget to pace yourself. Let's see how things go on Sunday!"

- M: "Good luck with the race! Don't forget to pace yourself. We have to say, we would be surprised if you met your goal of running under hh:mm. Although we're certain you can make it in your dreams ;-) Let's see how things go on Sunday!"

- MI: "Good luck with the race! Don't forget to pace yourself. We have to say, we would be surprised if you met your goal of running under hh:mm. Although we're certain you can make it in your dreams ;-) In fact, we are willing to bet money on your race performance: if we are wrong and you meet your goal, we will send you a £X Amazon gift card by email once you complete the Post race survey. Let's see how things go on Sunday!"

- I: "Good luck with the race! Don't forget to pace yourself. If you meet your goal of running under hh:mm, we will send you a £X Amazon gift card by email once you complete the Post race survey. Let's see how things go on Sunday!"

£X is randomized to be £10, £20 or £40.

In the Pre race survey, randomization into C or one of the 3 treatment groups is done in Qualtrics; an equal number of survey participants are allocated into each group.

Likewise, the incentive level for the Money and Pride & Money Treatments are also randomized into equal groups offered £10, £20 or £40 by Qualtrics.

In the Post race survey, 50% of participants are informed that if they make correct guesses in some questions, they will be eligible for a £40 prize; this randomisation is also done by Qualtrics.

Individual

No

We want a total of 800 individuals (200 in each treatment group) to complete the full study. We anticipate some attrition between the Registration phase, and the Pre and Post race surveys, hence we aim to over-recruit for the first survey (900).

800 individuals.

200 in C, and 200 in each of the three treatment groups.

With a sample size of 800, we anticipate we will be able to detect a minimum of a 10% difference in those attaining their goal time between treatments (assuming a 45% rate of goal attainment in C). This is at a power of 0.8.