Handling Missing Data: Listwise Versus Pairwise Deletion

Posted December 18, 2018

Missing data can pose issues to your statistical analyses, and your committee may offer multiple ways to address the issue of missing data. While some committee members may prefer imputation to replace missing data, other committee members may offer deletion to address missing data. Pairwise and listwise deletion may be implemented to remove cases with missing data from your final dataset. Prior to using deletion, it is important to note that pairwise and listwise deletion can be used when you are dealing with data that is missing at random. Non-random missing data may require other methods for correction.

Free Help Session: Quantitative Methodology

During these sessions, students can ask questions about research design, population and sampling, instrumentation, data collection, operationalizing variables, building research questions, planning data analysis, calculating sample size, study limitations, and validity.


Register Here

Researchers using listwise deletion will remove a case completely if it is missing a value for one of the variables included in the analysis. For example, say you are conducting analyses using cumulative high school GPA, hours of study for first semester, SAT score, and first semester grade in college algebra. Participant X is missing data for cumulative high school GPA, therefore, Participant X will be completely removed from the analyses because the participant does not have complete data for all the variables.

Researchers using pairwise deletion will not omit a case completely from the analyses. Pairwise deletion omits cases based on the variables included in the analysis. As a result, analyses may be completed on subsets of the data depending on where values are missing. For the example listed above, Participant X will be omitted from any analyses using cumulative high school GPA, but they will not be omitted from analyses for which the participant has complete data.

Researchers using listwise deletion may have to contend with losing large amounts of data due to missing cases. Researchers using pairwise deletion may have challenges with drawing inferences to the total sample. Results in analyses using pairwise deletion may be based on subsets of cases with complete data. The patterns identified in the analysis may not hold for the complete dataset because the analysis only includes a portion of the dataset.

For assistance with these issues or any other questions you have, you can always reach out to Statistics Solutions for assistance. Our mentors are here to help you move your project forward!

Want more tips from our dissertation experts? Fill out the form to the right to request our 7 Secrets to Completing Your Dissertation in One Year download.

Pin It on Pinterest

Share This