Im having difficulty computing Cronbach's Alpha on a measure's subscales. The subscales are all likert ratings, of which the variables have been set up as ordinal string variables with 7 (0 to 6) values. When I use Analyze>Scale>Reliability and put all relevant subscales into ratings, set Model as Alpha and compute I get a Kappa output and not Alpha. Can anyone tell me what I am doing wrong? Thanks!
Asked
Active
Viewed 221 times
0
-
Please use `paste` to get the syntax of your command and add it to your post, that might help us understand what could be done. – eli-k Jul 06 '21 at 15:03
-
Hi! Here you go. It looks like I initiated a Kappa computation, but definitely did not - as far as I know... FLEISS MULTIRATER KAPPA EE1EmoDrain EE2UsedUp EE3Fatigued EE6WorkPPL EE8BurnedOut EE13Frustrated EE14WorkHard EE16Stress EE20End /CRITERIA IGNORE_CASE=FALSE ASYMPTOTIC_CILEVEL=95 /MISSING CLASSMISSING=EXCLUDE. – Clifford Lewis Jul 07 '21 at 16:59
-
It seems that for reliability analysis, if you add ratings SPSS goes to Kappa instead of whatever model you selected. See [here](https://www.ibm.com/docs/en/spss-statistics/SaaS?topic=kappa-overview-fleiss-multirater-command): "When at least two ratings variables are selected, the Fleiss' Multiple Rater Kappa syntax is pasted." – eli-k Jul 07 '21 at 17:55
-
Hi Eli. Thank you so much for your comments! So then am I understanding correctly that if I have a scale, made up of 9 subscales that are measured using a likert scale, that if I want to determine the reliability of these 9 subscales I can only do so with Kappa, not with Alpha? What are the most basic requirements then to be able to compute Alpha as a measure of reliability? Thank you! – Clifford Lewis Jul 07 '21 at 18:02
-
Actually my comment refers to what is available in SPSS, not the theoretical question (which I can't answer, and would be off topic in Stack-Overflow). To get an answer for that you might do better [HERE](https://stats.stackexchange.com/questions/ask) – eli-k Jul 08 '21 at 06:03
-
1Thanks Eli. Much appreciated! – Clifford Lewis Jul 08 '21 at 17:25