2022
DOI: 10.1037/met0000361
|View full text |Cite
|
Sign up to set email alerts
|

Missing data in experiments: Challenges and solutions.

Abstract: Missing data is a common feature of experimental datasets. Standard methods used by psychology researchers to handle missingness rely on unrealistic assumptions, invalidate random assignment procedures, and bias estimates of effect sizes. We describe different classes of missing data typically encountered in experimental datasets, and we discuss how each of them impacts researchers' causal inferences. In this tutorial, we provide concrete guidelines for handling each class of missingness, focusing on two metho… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
11
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
7
1

Relationship

0
8

Authors

Journals

citations
Cited by 23 publications
(11 citation statements)
references
References 24 publications
(52 reference statements)
0
11
0
Order By: Relevance
“…The next items ask researchers to outline clear data cleaning and screening strategies (M7) together with their plans for handling missing data (M8). More information on critical considerations and practical approaches for the treatment of missing data can be found in Gomila and Clark (2020), Graham (2009), and Newman (2014).…”
Section: Structure Of the Preregistration Templatementioning
confidence: 99%
“…The next items ask researchers to outline clear data cleaning and screening strategies (M7) together with their plans for handling missing data (M8). More information on critical considerations and practical approaches for the treatment of missing data can be found in Gomila and Clark (2020), Graham (2009), and Newman (2014).…”
Section: Structure Of the Preregistration Templatementioning
confidence: 99%
“…Given these results and the lack of other demographic or socioeconomic variables, no variables were included in the models as covariates. Using R, the authors conducted two inverse probability weighted (IPW) multinomial logistic regression (MLR) analyses (Gomila & Clark, 2022) to determine whether the independent variables would increase the probability that a call would result in one of four outcomes (Monitored, Intervention, Adverse Outcome, Unknown). MLR analyses are robust against abnormality of data and variance within categories (Sturdivant et al, 2013) making it uniquely qualified for these data.…”
Section: Discussionmentioning
confidence: 99%
“…For the belief items, we relied on linear regressions. We accounted for participant attrition by weighting on the inverse propensity weight for predicting missing observations (Gomila & Clark, 2020), relying on full information maximum likelihood for the linear regressions (Li & Stuart, 2019), and using covariates that were observed on all respondents (i.e., gender, age, race, education, partisanship, and moral traditionalism) 5 . Full model results are provided in Appendix B.…”
Section: Discussionmentioning
confidence: 99%