I have collected responses for questions asking about satisfaction of a educational program.I have checked the satisfaction asking 6 question with 5 likert scale responses.(Strongly agree to strongly disagree).
1.I want to analyse data to find out satisfaction difference between gender,age group and times participated for the program.(Simply I want to know whether male are more satisfied than female or wise versa).
How could I do these type of analysis using spss?What are the statistical method that I want to use?
QUANTITATIVE. Assume that you treat you data is quantitative, you may entertain 2 options (1) normally distributed and (2) not normally distributed.
NORMALLY DISTRIBUTION. Provided that you have adequate sample size and the data is normally distributed, the test statistic for the group difference may be for n1 and n2 where n1 = male and n2 = female:
T = A / B
A = (mean1 - mu1) - (mean2 - mu2)
B = sqrt (1/n1 + 1/n2)
NOT NORMALLY DISTRIBUTED. Since we have two groups, group differences may be tested by goodness of fit by overlapping the two variances using the F test:
F = var1 / var2
Compare the F calculated with the critical value of F by degree of freedom of the two group. Use the larger variance as numerator. F > F* means significant difference where F* is critical value read from the F table.
Make use of t-test (independently) just to see if there is a difference or not..but to see the difference curve, make use of Wilcoxon Signed Rank Test
You should consider combining your separate Likert-scored items into a scale, which you can use a dependent variable in t-Tests, correlations, etc. The first step in this process is to investigate the reliability of such a scale, using coefficient alpha.
- Badges
- Science topic
- Similar topics
- Design
- Architecture