Pre-Analysis Plan for Social Norms, Values and Misperception

Last registered on September 12, 2024

View Trial History

Pre-Trial

Trial Information

General Information

Title

Pre-Analysis Plan for Social Norms, Values and Misperception

RCT ID

AEARCTR-0014167

Initial registration date

August 11, 2024

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

September 12, 2024, 4:03 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

Uganda

Region

Mbale

Primary Investigator

Name

Joshua Hill

Affiliation

University of East Anglia

Contact Primary Investigator

Other Primary Investigator(s)

Additional Trial Information

Status

In development

Start date

2024-08-23

End date

2024-09-20

Keywords

Behavior, Governance

Additional Keywords

Misperceptions, In-group bias, Social Norms, Motivated Reasoning, Cultural Identity, Curse of Knowledge, Projection Bias

JEL code(s)

C71, C91, C92

Secondary IDs

N/a

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

When measuring perceptions, previous research has shown that an individual’s own attitudes and beliefs are strongly, positively correlated with perceptions of the in-group, but negatively correlated with out-groups. Not only are there more positive correlations, perceptions are also more accurate for the in-group. While significant, these findings have been limited to a single ethnicity and political domain, and the underlying mechanisms remain unexplored. This study will replicate these finding across a range of domains and ethnicities. This study also investigates four potential mechanisms: Motivated Reasoning, Cultural Identity, Projection Bias and the Curse of Knowledge, both across perceptions of the in and out-groups. Lastly, this study will investigate the nature of inaccuracies within these perceptions, specifically whether these can be attributed to irrational inattention, identity bias or both.

External Link(s)

Registration Citation

Citation

Hill, Joshua. 2024. "Pre-Analysis Plan for Social Norms, Values and Misperception." AEA RCT Registry. September 12. https://doi.org/10.1257/rct.14167-1.0

Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Experimental Details

Interventions

Intervention(s)

This study does not contain any specific interventions, as it is not directly attempting to change behaviours or policy. The study is comprised of an incentivised survey in two phases. The first phase is to collect actual data on beliefs of five distinct tribal groups in Uganda. The second phase involves a lab-in-the-field experimental design, focusing on one specific group (Gisu), asking a number of incentivised questions related to phase 1. The second phase will also collect data to enable the identification of the underlying mechanisms at work.

Intervention Start Date

2024-08-23

Intervention End Date

2024-09-20

Primary Outcomes

Primary Outcomes (end points)

To replicate in-group/out-group differences in misperceptions, including the underlying mechanisms and sources of inaccuracy.

Primary Outcomes (explanation)

Previous work has shown in/out-group differences are substantial, but mainly in the political domain in the US. This study looks to replicate this across a number of different ethical (rather than political) groups, and across multiple domains in Sub-Saharan Africa.

Secondary Outcomes

Secondary Outcomes (end points)

The secondary outcomes are: 1) To investigate the underlying mechanisms at work and 2) to investigate the nature of inaccuracies.

Secondary Outcomes (explanation)

1) The literature identifies four potential mechanisms which could be driving the noted in/out-group differences: Cultural Identity, Motivated Reasoning, Curse of Knowledge and Projection Bias. I investigate each in turn and jointly. 2) Two potential causes of inaccuracy have been identified in the literature: Rational Inattention and Identity Bias. I investigate both, interacting accuracies with each of the mechanisms.

Experimental Design

Incentivised survey measuring own beliefs and perceptions of others, in addition to a number of supplemental questions, in a lab setting. Further details set out above.

Experimental Design Details

Full designs are included in the attached Pre-Analysis Plan or included in the above. Specifically, five domains will be investigated: Politics, religion, deception, violence and adultery. Five ethnic groups will also be considered: Gisu, Basoga, Baganda, Acholi and Teso. If, during the pilot, the number of experimental questions proves to be unrealistic, the Basoga and Teso will be excluded from the experimental design.

All participants will be paid a participation fee. For phase two, payments are contingent on how closely participants are (in their elicited beliefs) to actual beliefs from phase one. Other questions, such as those which are required to compute the underlying mechanisms, are not incentivised.

Randomization Method

Sampling frame to be adopted as set out page 8 of attached Pre Analysis Plan, though this will be refined during the pilot process. For phase one, which measures overall beliefs, participants will be randomly approached for participation at the street level. For phase two, a sampling frame will be identified (either from electoral register or university/college enrolment records) and participants will be randomly selected and approached for participation.

Randomization Unit

Individual level - no other clustering required.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

300 individual participants (lab phase) + 500 initial participants (to provide supplemental data). These numbers will change if two of the ethnic groups are dropped. Phase one will become 300, and phase two increase to at least 350.

Sample size: planned number of observations

800 participants; 500 x 5 questions + 300 x 126 questions = 40,300 data points

Sample size (or number of clusters) by treatment arms

800 participants

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

This depends on the specific hypothesis - please see power analysis in Appendix 2 of the attached pre-analysis plan. All specifications shown to be sufficiently powered with 300 participants (>80%).

Supporting Documents and Materials

IRB