Evaluating Longshot Preference in the Field

How to cite

Zhong, Songfa. 2019. "Evaluating Longshot Preference in the Field." AEA RCT Registry. AEARCTR-0004790. https://doi.org/10.1257/rct.4790-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Songfa Zhong*

* Primary investigator

Registry ID
AEARCTR-0004790
DOI
https://doi.org/10.1257/rct.4790-1.1
First registered
Last updated
Status
Completed
Trial dates
2019-10-01 – 2019-10-16
Countries
Singapore (Singapore)
Registry record
http://www.socialscienceregistry.org/trials/4790

Abstract

We collaborate with a support team in an Asian University, which provides information technology services to the university community. The support team aims at motivating participation in a crossword game, through which users will master the use of some self-services provided by them. To rise participation rate, each participant will receive a lottery as reward. We design six lotteries and randomly divide the users into six groups. The lottery incentive varies among groups. By comparing the response rate in each group, we can evaluate the longshot preference under the setting of motivating behavior.

Experimental design

We design six lotteries, which differ in: (1) sources of uncertainty (familiar vs. unfamiliar); (2) skewness of probability; (3) number of prizes; (4) method of probability realization (“randomly draw one out of 100 participants” vs. “bet on the last two digits of some indexes”); (5) vagueness of information (“lucky draw” vs. “win with probability 1%”). We randomly divide all users in six groups. Each group receives one lottery as reward for participation.

Design details

Randomization method
The randomization is done by a computer. For each subject, we generate a random number following standard normal distribution using STATA. After that, we sort the sample according to the value of this random variable. The first one sixth of subjects consist of our first group, and so on.
Randomization unit
Individual.
Clusters
This study has no cluster.
Observations
60,000 people in university, including staff and students.
Arms
10,000 people
Primary outcomes
The participation rates in six groups.

Cited by

No citation data available for this registration yet.

Keywords