Actually, I already established that the wording made no difference in mean responses on the scale. This question about possible significant differences in reliability coefficients is a secondary issue. Cheers.
Lise
At 04:19 PM 11/19/2003 -0600, bolson wrote:
I think perhaps you are doing the wrong test. If you are interested in the effect of wording change on responses - you would examine only respondent scores on the items having different wordings. This may mean that you compare mean scores on those items (t-test) or examine the proportional difference between responses if you have say a 5-category Likert type scale. Then you could establish any significant difference in responses due to item wording for a given item.
Cronbach's alpha assesses scale reliability and that is a bit of a different issue. Unless you are interested only in scale-level effects. If this is the case there is no (that I know of) statistical comparison of two Cronbach scores. Rather there is a standard rule of thumb such that alphas equal to or greater than 0.70 are considered strong while those less than 0.60 are considered weak.
Brittawni L. Olson, Ph.D. Candidate Project Director Drug Free Communities Support Project Lincoln Council on Alcoholism & Drugs, Inc. 914 L Street Lincoln NE 68508 P: 402-475-2695 F: 402-475-2699 E: [EMAIL PROTECTED]
"Save a Tree ....... Use Email!!"
-----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Dennis Roberts Sent: Wednesday, November 19, 2003 3:52 PM To: [EMAIL PROTECTED] Subject: Re: [edstat] question re comparing Cronbach's alphas
well, are the MEANS different? maybe that's all that is important but, ONE way you might ... assuming that alphas can be interpreted as correlations ... is to use a test of the difference between 2 correlations however, i am not sure what n would be ... #items? #examinees?
At 04:41 PM 11/19/03, Lise DeShea wrote: >List mates: > >I have a study where I randomly assigned participants to completing one of >two versions of the same scale so that I could determine whether a change >in wording made any difference in scores. I computed separate Cronbach's >alphas for each version of the scale. > >My question is, how would I test whether the two alphas are significantly >different? > >Thanks for your help. >Lise > > > >~~~ >Lise DeShea, Ph.D. >Assistant Professor >Educational and Counseling Psychology Department >University of Kentucky >237 Dickey Hall >Lexington KY 40506 >Email: [EMAIL PROTECTED] >Phone: (859) 257-9884 > >"... for only by varied iteration can alien conceptions be forced on >reluctant minds." > -- Herbert Spencer, in the preface to The Data of Ethics, > 1881. > > >. >. >================================================================= >Instructions for joining and leaving this list, remarks about the >problem of INAPPROPRIATE MESSAGES, and archives are available at: >. http://jse.stat.ncsu.edu/ . >=================================================================
---------------------------------------------------------- Dennis Roberts Email: [EMAIL PROTECTED] Web: http://www.personal.psu.edu/users/d/m/dmr/droberts.htm
. . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
. . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
~~~ Lise DeShea, Ph.D. Assistant Professor Educational and Counseling Psychology Department University of Kentucky 237 Dickey Hall Lexington KY 40506 Email: [EMAIL PROTECTED] Phone: (859) 257-9884
"... for only by varied iteration can alien conceptions be forced on reluctant minds."
-- Herbert Spencer, in the preface to The Data of Ethics, 1881.
. . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
