Hello Ravisha,

If cases are randomly assigned to treatment condition, there's no reason that post-only design results should be considered uninformative.

Designs with pre-post measures can offer the added benefits of: (a) allowing for estimation of change (though unless scores are completely reliable, the change scores will be less reliable than either the pre- or post- score by itself); or (b) pre-scores can be used as a covariate, to adjust for randomly occurring differences across groups.

One noted threat to pre-post designs is that if the interval separating them is too short, the post-results, and therefore group comparisons, can be biased, especially with measures of affect.

Ultimately, the answer depends on what your target ES might be: If it is post-treatment differences across groups/conditions, then either design can contribute. You could estimate ES separately by study type to see whether inclusion of pre-test appears to account for differences.

If it is strictly pre-post change, then post-only designs can't contribute (again, though, note the caveats above).

Good luck with your work.

David Morse Thank you so much! This helps a lot!