Thanks, this helps a lot.
It also led me to the -cialpha- package, of which I was unaware. It
seems very useful.
. alpha v1-v16
Test scale = mean(unstandardized items)
Average interitem covariance: .274784
Number of items in the scale: 15
Scale reliability coefficient: 0.9211
. cialpha
Cronbach's alpha one-sided confidence interval
--------------------------------------------------
Items | alpha [95% Conf.Interval]
---------+----------------------------------------
Test | .92106827 >= .81485882
--------------------------------------------------
If I'm interpreting this output correctly, with just the 8 cases on my
15-item scale, I can say with 95% confidence that the alpha is above 0.8?
--Chris
--
Christopher W. Ryan, MD
SUNY Upstate Medical University Clinical Campus at Binghamton
and Wilson Family Practice Residency, Johnson City, NY
cryanatbinghamtondotedu
GnuPG and PGP public keys available at http://pgp.mit.edu
"If you want to build a ship, don't drum up the men to gather wood,
divide the work and give orders. Instead, teach them to yearn for the
vast and endless sea." [Antoine de St. Exupery]
frone@ria.buffalo.edu wrote:
> Christopher,
>
> Regarding your second question, you might find the following article a
> useful summary:
>
> Duhachek, A., Coughlan, A.T., & Iacobucci, D. (2005). Results on the
> Standard Error of the Coefficient Alpha Index of Reliability. Marketing
> Science, Vol. 24, No. 2, pp. 294–301.
>
> They discuss the impact of scale length (p), average inter-item
> correlation (r), and sample size (n) on the point estimate of coefficient
> alphas and it standard error.
>
> Both p and r impact the point estimate, but n does not. So its not
> unreasonable to obtain a high alpha with a small sample.
>
> In contrast, p, r, and n affect its precision (i.e., standard error).
>
> In their conclusions they state:
>
> "Analysis 1 also proves analytically that p and r are substitutes in their
> beneficial (positive) effects on alpha, and that p, r, and n are
> substitutes in their beneficial (negative) effects on alpha’s standard
> error."
>
> Perhaps the best thing would be to report the point estimate with its
> standard error.
>
> Mike Frone
>
> ****************************************************************
> Michael R. Frone, Ph.D.
> Senior Research Scientist
> Research Institute on Addictions
> State University of New York at Buffalo
> 1021 Main Street
> Buffalo, New York 14203
>
> Office: 716-887-2519
> Fax: 716-887-2477
> E-mail: frone@ria.buffalo.edu
> Internet: http://www.ria.buffalo.edu/profiles/frone.html
> ****************************************************************
>
>
>
> "Christopher W. Ryan" <cryan@binghamton.edu>
> Sent by: owner-statalist@hsphsun2.harvard.edu
> 09/18/2006 06:26 PM
> Please respond to
> statalist@hsphsun2.harvard.edu
>
>
> To
> Statalist <statalist@hsphsun2.harvard.edu>
> cc
>
> Subject
> st: sample size for correlation, and for Cronbach's alpha
>
>
>
>
>
>
> I am attempting to create a summated rating scale to detect functional
> fecal retention (ffr) in children. The scale contains 15 items (v1-v16,
> but no v10 (which is an unrated distractor)). 12 Items are rated 0-2; 3
> items are rated 0-4. I had planned to sum the scores on the individual
> items; sum can range 0-36. Higher scores mean more constipated.
>
> I hope to correlate the score on the scale with visual assessments of
> constipation on plain radiographs.
>
> Two questions come to mind:
>
> Does Stata have a sample size routine for correlations?
>
> How does sample size relate to Cronbach's alpha? I've been told that
> roughly ten subjects per item would yield a reasonable sample size for
> reliability testing; this would be 150 in my case. I have little hope
> of recruiting more than 50-60 subjects for this initial study, unless I
> extend it to a longer duration, which I'd like to avoid.
>
> To my knowledge, no such survey instrument exists, so these are
> relatively uncharted waters. I've only tried out the survey on 8
> subjects, just to see if they could understand the questions:
>
> .slist v1-v16
>
> v1 v2 v3 v4 v5 v6 v7 v8 v9 v11 v12 v13 v14 v15 v16
> 0 0 0 0 0 0 0 0 0 0 0 0 1 2 0
> 1 0 0 0 0 0 0 0 0 0 0 0 1 1 1
> 0 0 0 0 0 0 1 0 0 0 0 0 1 3 .
> 1 0 0 0 0 1 0 0 0 0 0 0 2 2 1
> 2 1 1 2 0 1 2 1 2 1 0 2 3 2 2
> 1 0 0 1 0 0 1 0 2 0 1 1 3 1 0
> 2 2 2 2 1 1 2 2 0 0 0 2 1 . .
> 2 1 1 1 2 1 2 1 1 1 2 2 2 2 0
>
> I feel silly calculating it for just 8 cases, but alpha for the
> instrument as a whole was
>
>
> . alpha v1-v16
>
> Test scale = mean(unstandardized items)
>
> Average interitem covariance: .274784
> Number of items in the scale: 15
> Scale reliability coefficient: 0.9211
>
> What conclusions could I draw from this, about the likelihood of
> obtaining a respectable alpha with less than 150 subjects? Or is that a
> nonsensical question?
>
> Thanks.
>
> --Chris
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/