Can we extract the answered questions from each one?
If one of the question that may answered by the respondent, may consider as his opinion in that only a question?
Witch statistic basics prove that?
Dear Chems,
In methodology section, "Response rate" is given which is the ratio of received to floated questionnaires. Further, out of the received questionnaires some may not be completely filled, such questionnaires suppose to be omitted as you are getting partly response. For example, you might get the response for independent variables but not for the dependent variable or vice versa. then it is difficult to run the model.
Either you do it manually while enetering the data or SPSS shows it as excluded no. ie in case of 100 respondent, if 12 are incomplete then SPSS will consider 88 observations.
Hope it helps.
Dear The Evaluation Center,
This will be entirely New research, as you have different objective "who did not complete the questionnaire"
Chems -
What you have described is called "item nonresponse," and in practice, it often occurs. In such a case, data are often either imputed by an accepted method (I like regression, for which we can get variance estimates), or the survey weights are expanded to account for the missing data. If you have a census, you can still use "response propensity groups" to weight to cover nonresponse.
The response above that you received from "The Evaluation Center" can also lead you to important stratifications. You should break your data into strata based on similar characteristics.
I think it might be helpful for you to do an Internet search on "item nonresponse," and perhaps also on "imputation," "weighting for nonresponse," and "response propensity."
Cheers - Jim
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics