I have also done that for multi-item constructs that used Likert scales.
Now, I need to ask if I can use Cronbach alpha to measure the reliability of a multiple-choice question biology test to assess students' knowledge of Science Process Skills. Each test question has four responses A, B, C and D. I hope to get some responses. Will the alpha that I obtain be meaningful?
Chronbach's alpha meaures internal consistency reliability, that is it meaures how well the items that reflect the same construct yield similar results; I this since you can use it inyour case (it can be considred as the average correlation between all of the test items) . If you get high alpha, there is evidence that the items (questions) are measuring the same underlying construct (science process skils). To find the alpah you need to code A, B, C, D into numbers 1, 2, 3, 4.
Perhaps Kuder-Richardson and test-retest are more applicable than Cronbach's alpha for your case. You can re-code all wrong responses as 0 and right ones as 1.
Thanks, both of U. I am hesitant, because SPSS seems so powerful and IS SO powerful. Able to provide whatever we ask, JUST ASK AND U SHALL RECEIVE...
So I need to be sure the info I get is meaningful.
- Badges
- Science topic
- Similar topics
- Education