Sparse Estimation and Uncertainty with Application to Subgroup Analysis

Ratkovic, Marc; Tingley, Dustin

doi:10.1017/pan.2016.14

Cited by 53 publications

(66 citation statements)

References 78 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is particularly the case for subgroup analyses of conjoint experiments. Such exercises are an increasingly common feature of experimental analysis (Green and Kern, 2012;Ratkovic and Tingley, 2017;Grimmer, Messing, and Westwood, 2017;Egami and Imai, 2018). For example, the Hainmueller, Hopkins, and Yamamoto (2014) study of immigration attitudes splits the sample in two using a measure of ethnocentrism and then compares AMCEs for the two subgroups.…”

mentioning

confidence: 99%

Measuring Subgroup Preferences in Conjoint Experiments

Leeper¹,

Hobolt²,

Tilley³

2019

Polit. Anal.

475

326

View full text Add to dashboard Cite

Conjoint analysis is a common tool for studying political preferences. The method disentangles patterns in respondents’ favorability toward complex, multidimensional objects, such as candidates or policies. Most conjoints rely upon a fully randomized design to generate average marginal component effects (AMCEs). They measure the degree to which a given value of a conjoint profile feature increases, or decreases, respondents’ support for the overall profile relative to a baseline, averaging across all respondents and other features. While the AMCE has a clear causal interpretation (about the effect of features), most published conjoint analyses also use AMCEs to describe levels of favorability. This often means comparing AMCEs among respondent subgroups. We show that using conditional AMCEs to describe the degree of subgroup agreement can be misleading as regression interactions are sensitive to the reference category used in the analysis. This leads to inferences about subgroup differences in preferences that have arbitrary sign, size, and significance. We demonstrate the problem using examples drawn from published articles and provide suggestions for improved reporting and interpretation using marginal means and an omnibus F-test. Given the accelerating use of these designs in political science, we offer advice for best practice in analysis and presentation of results.

show abstract

mentioning

confidence: 99%

Measuring Subgroup Preferences in Conjoint Experiments

Leeper¹,

Hobolt²,

Tilley³

2019

Polit. Anal.

475

326

View full text Add to dashboard Cite

show abstract

“…(2) principled model selection. One of the most significant challenges faced by researchers using (Bloniarz et al, 2016;Ratkovic and Tingley, 2017).…”

Section: Machine Learning and "Big Data"mentioning

confidence: 99%

Machine Learning for Public Administration Research, With Application to Organizational Reputation

Anastasopoulos¹,

Whitford²

2018

Journal of Public Administration Research and Theory

View full text Add to dashboard Cite

Machine learning methods have gained a great deal of popularity in recent years among public administration scholars and practitioners. These techniques open the door to the analysis of text, image and other types of data that allow us to test foundational theories of public administration and to develop new theories. Despite the excitement surrounding machine learning methods, clarity regarding their proper use and potential pitfalls is lacking. This paper attempts to fill this gap in the literature through providing a machine learning "guide to practice" for public administration scholars and practitioners. Here, we take a foundational view of machine learning and describe how these methods can enrich public administration research and practice through their ability develop new measures, tap into new sources of data and conduct statistical inference and causal inference in a principled manner. We then turn our attention to the pitfalls of using these methods such as unvalidated measures and lack of interpretability. Finally, we demonstrate how machine learning techniques can help us learn about organizational reputation in federal agencies through an illustrated example using tweets from 13 executive federal agencies.

show abstract

“…While some hypotheses can be formed regarding treatment effect heterogeneity (see below), repeatedly subsetting the data leads to multiple inference problems and implicitly acknowledges that the current model is misspecified. I will therefore use LASSOplus, a Bayesian method to identify relevant subgroups (Ratkovic and Tingley 2017). The estimator both estimates and selects relevant treatment effects.…”

Section: Treatment Effect Heterogeneitymentioning

confidence: 99%

“…It considers all lower order and interaction terms and returns the ones estimated to be different from zero. As shown by a simulation experiment in Ratkovic and Tingley (2017), LASSOplus estimates are conservative but powerful in identifying nonzero effects. As this estimation is post-inferential, it describes effects that are apparent in my data and can therewith point to interesting questions to be considered in future research.…”

Section: Treatment Effect Heterogeneitymentioning

confidence: 99%

Mental Accounting and Savings Decisions of the Poor

Spantig¹

2017

AEA Randomized Controlled Trials

View full text Add to dashboard Cite

This documents describes the planned analyses for the lab-in-the-field experiment assessing the impact of cash-in-hand on savings decisions. Between February and March 2017, 300 microfinance clients took part in paid, individual interviews. All clients were informed one week in advance that at the next weekly (borrower) center meeting, a randomly drawn subsample would be invited to take part in interviews that would pay P 300 ($6). During the interview, clients were asked whether they wanted to save some of their payment in their existing savings account. The study uses a 2x2 design with two levels of randomization. First, cash-in-hand was randomized at the individual level: 50 percent of the participants received their show-up fee at the very beginning of the survey and before making the savings decisions whereas the other half received the remainder of the cash they did not save directly after the decision. Second, at the day of the interviews, in half of the 31 centers a participation fee of P 500 ($10) was announced after the recruitment of volunteers but before the start of the interviews. The present document describes the outcome variables and econometric methods that I will employ to assess the effect of cash possessions on savings decisions.

show abstract

Sparse Estimation and Uncertainty with Application to Subgroup Analysis

Cited by 53 publications

References 78 publications

Measuring Subgroup Preferences in Conjoint Experiments

Measuring Subgroup Preferences in Conjoint Experiments

Machine Learning for Public Administration Research, With Application to Organizational Reputation

Mental Accounting and Savings Decisions of the Poor

Contact Info

Product

Resources

About