Sample size for repeated measures or dependent-sample t

Hi John:
I have another question. I collected these data from parents and their children about children's chronic condition. I wanted 100 responses, but since I requested the documents from their pharmacy, the response rate was low. I only got 50 parents and 50 children. Since they are related, I suppose I don't need as many as I would for independent samples??? Also, I collected the same variables (not too many) 2 years ago. So, I also want to do a longitudinal analysis. In both cases, would you say that I don't need to have many (well, more the better, but...)? If I did not make myself understood, I am sorry. Please let me know. Thank you.


TS Contributor
Yes, in general, if you have dependent samples, you can get away with smaller sample sizes because of the higher power of the test design.

50 should be plenty....