Hi

I am preparing a systematic review of some studies in education. I need help in interpreting some stats reported in one of the studies.

The study uses multiple baseline interrupted time series design to compare reading scores before and after the introduction of a teaching intervention. There are eight participants.

The author reports the slope value for each participant individually, before and after the introduction of the intervention. However she does not combine slope values in any way to report any measures of central tendency. She does report Mean scores (not slopes) for each participant for each time period. She does not perform any tests of significance on these data.

I want to summarise the findings of this study. Is it legal for me to calculate the mean slope value for each time period by adding the slope values for each participant during each time period, then dividing by 8 (the number of participants)? Alternatively, can I just take the mean scores and compare them (forgetting the slope values)? If the latter, can I combine the Means for each participant to calculate a Mean for the whole group. If so, how do I calculate the SD for the group as a whole. I have the SDs for individual participants but not raw data.

I guess it might be possible that the author has chosen not to report Means of the group as a whole because the sample is so tiny and she feels it makes more sense to analyse the data on a case by case basis. but I'm not sure.

Any help most gratefully received.

Thank you