[afro-nets] RFI: ASE Constructs (2)

[Ayankola Oluseun <johnayank@hotmail.com>]

RFI: ASE Constructs (2)
-----------------------

Good enough, you have recoded your data in a categorical order.
Note that this data follows a non-parametric distribution, and
that if these responses are suspected to be naturally at vari-
ance, you may need to assume a log normal distribution.

By experience, Epi-info both old and new version (6.02/6.04-
2000, 2002) and the latest version 3.01 cannot get you Cronbach
alpha statistics, and am not sure Stata will do it. Use SAS,
BMDP or SPSS.

I do not know whether you are conversant with these software or
not, but if you are using SPSS, it's a bit easier and straight
forward.

Just export the data into D-base format, and open it using SPSS.
Then go to the analyse menu. Click the pull down menu, there you
will find the sub menu 'Scale', click on it, then on reliability
analysis. A dialog box display and just pick the variables you
want to get the inter corr. Click on statistics and then the ap-
propriate ones you will need. Ok it and the output will be out
there for you.

You can use this same step to get Split Gutman, Brown etc.

You can contact me for further clarification or explanation if
you find using any of the software difficult:
mailto:significanceltd@hotmail.com
or
mailto:johnayank@hotmail.com


J.O. Ayankola
Tel.: 08035178455
mailto:johnayank@hotmail.com