Field
Abstract
|
Before
This document describes the analysis plan for a randomized experiment examining the psychological effects of poverty on cognitive function. We will recruit 1012 respondents from Amazon Mechanical Turk and expose our treatment group to a prime that triggers feelings of poverty (Mani et al., 2013). Then, participants complete ten Raven progressive matrices and 75 items of a Stroop task. The design of the study is a replication of the study by (Mani et al., 2013). This plan outlines the design of the experiments, the outcomes of interest, the econometric approach and the dimensions of heterogeneity we intend to explore.
|
After
This document describes the analysis plan for a randomized experiment examining the psychological effects of poverty on cognitive function. We will recruit 2000 respondents from Amazon Mechanical Turk and expose our treatment group to a prime that triggers feelings of poverty (Mani et al., 2013). Then, participants complete ten Raven progressive matrices and 75 items of a Stroop task. The design of the study is a replication of the study by (Mani et al., 2013). This plan outlines the design of the experiments, the outcomes of interest, the econometric approach and the dimensions of heterogeneity we intend to explore.
|
Field
Last Published
|
Before
August 21, 2015 08:50 AM
|
After
August 21, 2015 09:24 AM
|
Field
Planned Number of Clusters
|
Before
1012 respondents on Amazon Mechanical Turk.
|
After
2000 respondents on Amazon Mechanical Turk.
|
Field
Planned Number of Observations
|
Before
1012 respondents on Amazon Mechanical Turk.
|
After
2000 respondents on Amazon Mechanical Turk.
|
Field
Sample size (or number of clusters) by treatment arms
|
Before
506 treatment individuals;
506 control individuals.
|
After
1000 treatment individuals;
1000 control individuals.
|
Field
Power calculation: Minimum Detectable Effect Size for Main Outcomes
|
Before
The chosen sample size of 1012 participants for the experiment ensures that we can detect an effect size of about 0.18 at a significance level of 0.05 with a power of about 0.8. Given that the effect sizes reported by Mani et al. (2013) are between 0.8 of a standard deviation and 1 standard deviation, we can be confident that our sample is sufficiently large to provide us with sufficient statistical power to detect effects.
|
After
The chosen sample size of 2000 participants for the experiment ensures that we can detect an effect size of about 0.125 at a significance level of 0.05 with a power of about 0.8. Given that the effect sizes reported by Mani et al. (2013) are between 0.8 of a standard deviation and 1 standard deviation, we can be confident that our sample is sufficiently large to provide us with sufficient statistical power to detect effects.
|