You could use a convenience sample of your existing students. They could either be randomly assigned to groups or divided by gender or placed into groups by socio-economic status. Choose what will give you honest results and the information that you desire.

If you only have students that are in coeducational instruction you will not be able to have a control group. I suggest that you change your study from an experimental design, which you can not do, to a more qualitative approach in which you investigate males and females' reaching to the instruction they are receiving. Are there differences? Do females and males react (cognitively and attitudinaly) differently? This could lead to a more refined experimental study, if it is ever possible.

I agree with the design of research explained to you by Dr Norman G Ledeman. It is the right way to do such research with the limitations you have. The other way, you could introduce an intervention and do measurements before and after introducing your intervention for each student (males and females), then you compare the impact on students learning. In this design the control will be the set of data collected prior to the intervention for each student. Good luck with your research. Prof S Azer

Thanks for all your suggestions have been very useful, can I use a demo sample with one work measurement before and after teaching? According to the semi-empirical approach?@@

This sounds like a simple pretest-posttest design. What question will this answer? Also, I am not sure what you mean by semi-empirical.

Norm