If you want to use these as case studies I would suggest writing them as a formal case series or, if your data is appropriate for it you could look at single case experimental design series (although if you have only two measurements pre and post this is unlikely to be appropriate).  A couple of sources of reading I would suggest are:

Kazdin, A. E. (2011). Single-case research designs: Methods for clinical and applied settings . Oxford University Press. . . . . . . . to see if your design would lend itself to a single case experimental design series

Yin, R. K. (2013). Case study research: Design and methods. Sage publications. . . . . . as a starting point for developing and reporting case study research.

There are lots of other texts and online resources on these topics - many of which are very good.

I hope this helps,

Jason

Thank you for your kind answer. To clarify the issue, this is a research that I did. I planned to do it through pretest-posttest one group. A group of student were chosen purposfully. They were pretested firstly then they got an intervention then they were posttested ...when i used t-test to analyze the data and to explore if there are any differences. They told me that the sample is not enough and i used a wrong mehtod ...and better to chane it totally into case study ...now what suggestions you give to me to solve this problem as some researchers suggeted to me to use nonparametric tests like wilconxon and Mann-whitenny u test ...thanks again

Dear Amr

I do not think you could turn your data toward a case study. The case study, although combines quantitative and qualitative data, implies to study in depth a unit of analysis (eg, individual, group, organization, etc.). In the best case you would have to take a lot of additional qualitative data. Additionally, experimental data are not usually presented in the case studies.

On the other hand, you consider that the size of the sample is related to the size of the effect you want to detect. In general, the smaller the effect you want to detect, the larger your sample should be.

Regards,

Jorge