An illustration of using multiple imputation versus listwise deletion analyses: the effect of Hanen's "More than words" on parenting stress.

The team looked at 62 parents who took Hanen’s “More Than Words” class.

Parents filled out stress surveys before and after the the study period.

Some surveys were blank. The authors ran the same numbers two ways: toss the blanks (listwise) or fill them in (multiple imputation).

Listwise deletion said the program hurts highly depressed parents. It made them look more stressed.

Multiple imputation said the program is neutral. No extra stress appeared.

Same kids, same parents, same class—opposite answers—just because of how missing surveys were handled.

McGarty et al. (2018) also used parent training, but they paid caregivers 50 cents a visit. Their tiny reward lifted both attendance and child skills. They kept every data point, so no deletion fight was needed.

Conine et al. (2025) coached parents to teach response-to-name. They saw gains fade when reinforcement stopped. Their single-case graphs kept all data; again, no missing-data headache.

Wolchik et al. (1982) ran a parent lottery and saw language jump. They used reversal designs that require full data sets. These older studies look cleaner, yet they never tested what happens when surveys walk away.

If you run parent groups, missing surveys are normal. Deleting them can make your good program look harmful. Use multiple imputation or another fill-in method before you sound the alarm. Check your stats software default—listwise is often the factory setting. Flip it so your data tell the truth, not the easy lie.