I am developing a questionnaire and am looking at issues like validity, reliablity and power calculations.
For example I understand that a check for internal consistancy is Cronbach a. Anyone got any formulas / software or can recommend specific (easy to read)books regarding how to make this calculation. Is it on Microsoft Excel? (I found out how to calculate confidence intervals via Excel)
Also does anyone know how to power calculation? I have assumed 20% "abnormal answers" on a 5 point Likert scale, which would have 0.32 confidence intervals in case of 80 participants. Anyone know what assumptions would be normal? Is that documented anywhere?

In which field are you? It depends from the field what software you use for statistics (I use SPSS and STATA). But in any case, all softwares have functions for calculation of Cronbachs. Regarding calculating the power- there is a special equation for it almost in any stats book, I really dont rememebr it by heart. Good luck!

The formula forpower calculation has to be in any dictionary of statistics, btw

Thanks Verdy,

I am in the medical field, although the questionnaire will cover psychological issues related to medical doctors.

I presume I will have to use SPSS, yet have not used that myself yet. I am trying to calculate thinks myself as I want to understand what I am doing instead of just using software.

Thanks for help

That is a very intelligent approach, indeed:)