Hi, I'm doing master dissertation and since there are some items measuring one scale in negative statement, I did reverse coding. But the Cronbach's Alpha is much lower after reversing.
Could anyone give me some advice on what the problem is and how could I do the correct recoding? Thanks a lot!
Jun Yin. Hello: I think that getting a higher alpha could indicate problems in your data. Therefore, I recommend that you look carefully at the nature of your data.
Certainly, if you need to reverse-code items, that needs to occur regardless of the Cronbach's alpha values that you obtain.
How are the bivariate correlations and scatter plots? If there is a problem with an individual item, whether in need of reverse scoring or not, this is a good place to look.
Daniel Wright Thank you for your help. And the attachment is the bivariate correlations. There are two groups(two questons first group including 11 items and second including 5 items) both of them aim to measure unlearning. The items I did reverse are the ones had problems of sig.
You have a lot of variables. Are you proposing they are all measuring the same construct?
Also, and this is tangential to your question, but given the criticisms about Cronbach's alpha that are easily found with google, I assume you have a specific reason for using it.
I hope this is helpful: My impression is that there is a reasonable number of variables being analysed (11 in one set and only 5 in the other set) - and I suspect that they are intended to measure two different constructs.
I would look very carefully at the correlation matrix for any "problem" items, or low or weird correlations, however. For example, I noticed one correlation of -.028.
Apart from that, as Daniel has recommended, I suggest you check out the criticisms of alpha. There is quite a lot of literature criticising it. Here is a reasonably recent, and nicely written, article:
Cho, E., & Kim, S. (201 Cho, E., & Kim, S. (2014). Cronbach's coefficient alpha: Well known but poorly understood. Organizational Research Methods, 18(2), 207-230. https://doi.org/10.1177/1094428114555994.
- Badges
- Science topic
- Similar topics
- Correction