Last updated on November 18th, 2026 at 08:04 am

You’ve finally collected 300 questionnaires for your SPSS dissertation, but panic sets in when you open the dataset. Several respondents skipped key questions. Your first instinct during SPSS data cleaning is to delete the incomplete cases and move on. But should you remove them, or impute them? How you handle missing values decides the integrity of your findings.
Quick answer
Use listwise deletion only when the amount of missing data is small and the data are Missing Completely At Random (MCAR). When data are Missing At Random (MAR), multiple imputation is usually the better choice because it keeps your sample size and reduces bias. Whatever you choose, inspect the missingness pattern first and justify the method in your methodology chapter.
Why missing data matters
Unaddressed missing data in SPSS silently damages quantitative research. It shrinks your sample size, reduces statistical power, inflates standard errors and skews parameter estimates. Ultimately, mishandling missing observations threatens the validity of your conclusions. The amount and pattern of missingness should decide your solution.
Listwise deletion
Listwise deletion removes any case with missing data on the variables in an analysis. If 30 of your 300 respondents skipped a variable used in your regression model, listwise deletion drops them entirely and your usable sample falls to 270.
It is simple and easy to interpret. However, it sacrifices statistical power and introduces bias if the deleted cases differ systematically from the complete cases.
Multiple imputation
Multiple imputation in SPSS is a more robust alternative. Rather than substituting a simple mean, it uses information from other observed variables to predict plausible values. It creates several imputed datasets, analyses each one and pools the results. This preserves your sample size while reflecting the uncertainty that missing data creates.
In SPSS, start with Analyze > Multiple Imputation > Analyze Patterns, then use Impute Missing Data Values to create the imputed datasets.
Listwise deletion vs multiple imputation at a glance
| Listwise deletion | Multiple imputation | |
|---|---|---|
| Sample size | Reduced | Preserved |
| Statistical power | Lower | Retained |
| Bias risk | High unless data are MCAR | Lower when data are MAR |
| Ease of use | Very simple | Needs more steps and reporting |
| Best for | Small, random missingness | Moderate missingness with useful predictors |
The missingness mechanism
Your choice depends on why the data are missing:
- Missing Completely At Random (MCAR): missingness is unrelated to any variable. Listwise deletion may only cost you power.
- Missing At Random (MAR): missingness is related to other observed variables. Deleting cases is likely to bias results, so multiple imputation is usually preferred.
- Missing Not At Random (MNAR): missingness depends on the missing value itself. No standard method fully fixes this, so report it as a limitation and consider a sensitivity analysis.
Making the right choice
There is no universally correct method. Do not choose multiple imputation just to look sophisticated, and do not default to listwise deletion just because it needs fewer clicks. Before deciding how to handle missing data in SPSS:
- Inspect the exact amount and pattern of missing values.
- Check whether specific variables or groups are disproportionately affected.
- Document and justify your chosen treatment clearly.
The takeaway
The best approach is rarely the most convenient one. Choose a defensible method, justify it, apply it consistently and report its implications transparently in your methodology chapter. Missing data is one part of good preparation; see our full guide to SPSS data cleaning before analysis and the 15 SPSS mistakes supervisors reject.
Not sure how to treat your missing data?
Our analysts check your missingness pattern, apply the right method and help you report it.
Get SPSS helpFrequently asked questions
What is listwise deletion in SPSS?
Listwise deletion removes every case that has a missing value on any variable in the analysis. It is the default in many SPSS procedures.
When should I use multiple imputation?
Use multiple imputation when a meaningful share of data is missing and the data are plausibly Missing At Random, so that other variables can help predict the missing values.
Is mean substitution acceptable?
Mean substitution is generally discouraged because it shrinks variance and distorts relationships between variables. Multiple imputation is usually a better option.
How do I report missing data in my dissertation?
Report how much data was missing, the pattern you found, the method you used and why, and any effect it may have on your results.










Comments are closed