I have selected 3 public and two private universities. Now I have to make comparison in public and private universities appraisal system and employees satisfaction with this level. I have to test hypotheses. Population of teachers in public universities is large while in private is comparatively small. How to determine sample size in such case? Say if it's 1000 in public universities and population of private universities is 150 only. Then how to select sample and what proportion
Are the employees the teachers in the Universities you are studying? You have mentioned these two respondents in your question.
You will have to use stratified sampling using type of University (public or private) as one of the strata. Obviously, the sample selected in public will be higher.
Are you also interested in whether male or female? If yes, again use stratified sampling after determining the population in each strata.
are going to do a quasiexperiment the sample size should be the same. Have you done a trial for the study with the sample? If You've done one then you can look at the pearson correlations "r's" of each scale with the trial sample, if the average ir your correlations are less than .40 you must increment the size for the final sample, this indicates low power and correlations between .40 and .60 show a moderate level, .60 and more you can keep a similar sample, I recommend you Cohen (1978) literature about power analysis. A power primer
You need to define your question more specifically - what is the outcome ? is it just binary (yes/no), categorical (varying levels) or continuous ? how many potential modifying factors you are interested in - sex, age, etc ...
Then there are many text books and software packages that can provide you the mechanics for estimating samples size.
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publication