Now a days it is mandatory to report the reliability of the questionnaire if we are targeting a good journal.
First the individual item responses of each person completing the questionnaire must be entered with each item in a separate column. Then all the data must be checked (run frequencies for each item and check any anomolies). Then you need to make decisions about missing values. Will you impute missing values? Or will you leave them blank. Imputing missing values means replacing the missing value with the mean, median or mode either for that column (listwise) or for that person (casewise).
To calculate the Cronbach's alpha (i.e. all possible split half reliability combinations), in SPSS v21, you click on "Analyse" then "Scale" then "Reliability Analysis". In the dialogue box you transfer all the items for the scale you are testing, select "Alpha" and give the scale a name e.g. depression scale. Finally click on OK. The output will give you a Cronbach's alpha between 0 and 1.
Hope this helps.
- Badges
- Science topic
- Similar topics
- Medicine
- Public Health