I am determining my minimum needed sample size for my study. My study included an initial pilot survey first which encompassed 70 respondents followed by a primary study, after Cronbach's alpha was performed for validity of the survey questions on the pilot survey, which encompassed 550 respondents. When G*Power software is used, you have to select effect size conventions. Effect size conventionsshown in the pic below are small f2 = 02,medium f2 = .05, and large f2 = .35. How do youdetermine whether to pick small, medium, and large effect size conventions?? Also should I include only one G*Power pic since I guess this minimum suggested respondent number would work for both the smaller (respondent wise) pilot study and the larger (respondent wise) primary study??Again, my respondent number for my pilot study survey was 70 and for my primary study survey was 550.Everything else I know is correct in the pic below inserted. Multiple regression with 2 independent variables and 1 dependent variable is being utilized and I have 2 inserted for number of predictors.

3 G*Power 3.1.9.7 le View Tests Calculator Help and noncentral distributions Protocol of power analyses X critical F = 3.13814 0.8 0.6 0.4 0.2 CO 10 12 Test family Statistical test F tests o rooraction. Fixed model, Rz deviation from zero Effect size conventions Type of power f 2 . 02 - small A priori: Com . 15 medium power, and effect size f2 = .35 large Input Parameters Output Parameters Determine =>> Effect size f 0.1 1 Noncentrality parameter 10.2000000 & err prob 0.05 Critical F 3.1381419 Power (1 -B err prob) 0.8 Numerator df 2 Number of predictors 2 Denominator df 65 Total sample size 68 Actual power 0.8044183 X-Y plot for a range of values Calculate Material