Log in

View Full Version : Likert scales and statistics


kerry1113
Mar 28, 2006, 02:41 PM
Hello,
I have done a survey that is trying to find the opinions about various aspects of HR and how these opinions differ in the public and private sector. The survey I have used involves a Likert scale. Any recommendations as to what sort of statistical analysis I should try?
Thanks

joseph2000
Mar 29, 2006, 02:00 AM
You can do a chi square stat to test the difference in perceptions. You can also add all the scores and used t-test for the difference in means

kerry1113
Apr 6, 2006, 01:35 PM
I have done a survey using a Likert scale and have tried to do a Mann whitney test for the following statement. “I consider the reputation of an organisation before deciding whether I want to work for them.” Both private and public sector employees were asked. The results I have calculated are

Sample a ( private sector) mean rank is 24.1
Sample b (public sector) mean rank is 22.4
Z=0.4
P1=0.3446
P2=0.6892

Z value of 0.4 means the chance probability Q is 0.344578 or about 1 in 3

What does this actually tell me though?

Many thanks
Kerry