scale reliability

[Helga Walz]

To the experts with thanks ahead of time:

For a community outreach project I collected data for 20 items reflecting stereotypes of aging. Because I could not find an existing scale, I developed items using a 5-point Likert type response format. Responses for the steretoype items were collected separately from young and older adults. The sample size is approximately 25 for young and 25 for older adults. It was suggested that I do a Repeat Measures ANOVA for item analysis or scale analysis (cannot do Factor Analysis, sample too small).
I would appreciate assistance in determing whether 1) the items do, indeed, indicate stereotypes of aging 2) the validity of the combined items, and 3) how to arrive at the answers using SPSS - so that I may explore stereotypical behaviors and their relationship with other variables.

Thank you so much,
Helga

[Matthias Spörrle]

Helga,

just some thoughts concerining your questions:

Alreck (Consumer age role norms, then and now, 2000) has developed a scale
you might find interesting, because it measures age role norms (example
item: "Sex and dating should be a thing of the past for people in their
sixties or older") which "can readily be seen as dimensions of age
stereotypes" (p. 907).
You can use this scale for the development of a new version of your scale or
for comparing the content of the items of the two scales (to examine
[content] validity). See her references for further papers concerning
stereotypes which might also be helpful for your validity question.

See Charter (Study samples are too small to produce [...], 2003) for a
discussion of adequate sample size concerning reliability estimates.

HTH
Regards
Matthias




On 5/15/07, Helga Walz <[hidden email]> wrote:

>
> To the experts with thanks ahead of time:
>
> For a community outreach project I collected data for 20 items reflecting
> stereotypes of aging. Because I could not find an existing scale, I
> developed items using a 5-point Likert type response format. Responses for
> the steretoype items were collected separately from young and older adults.
> The sample size is approximately 25 for young and 25 for older adults. It
> was suggested that I do a Repeat Measures ANOVA for item analysis or scale
> analysis (cannot do Factor Analysis, sample too small).
> I would appreciate assistance in determing whether 1) the items do,
> indeed, indicate stereotypes of aging 2) the validity of the combined items,
> and 3) how to arrive at the answers using SPSS - so that I may explore
> stereotypical behaviors and their relationship with other variables.
>
> Thank you so much,
> Helga
>

[Matthias Spörrle]

Helga,

yes, your sample is very small.

You could merge your samples and perform a correlation analysis to
investigate the relationship between the variables. Since all items are
supposed to measure one construct one would expect all (i x (i-1)) / 2
correlations to be positive [after recoding] which might serve as a very
first and simple indicator of  unidimensionality and reliability. (I want to
note that research indicates that age stereotypes are multi-dimensional.)

Moreover, as Alreck (2000) and others found out you might find higher levels
of stereotypes (i.e., higher values of your [recoded] scales) for older
people than for younger ones and an inverse relationship between levels of
stereotypes and education. These results would be in line with your
assumption that you have reliably measured age stereotypes. However, they
are no prove of it.

I would suggest that you try to validate the content of your items by
comparing them with other instruments and existing research. Then you should
increase your sample size to perform more accurate analyses. You should
include other stereotype scales and/or criteria when collecting your new
data, since they can then be used for your validity question.

HTH
Matthias


On 5/15/07, Helga Walz <[hidden email]> wrote:

>
> Matthias,
>
> Tank you so much for your quick reply. I plan to look into the
> references you suggested. I gather from your response that
> the sample size limits me in doing anything with the items/scale?
>
> Again, thank you,
>
> Helga
>
>
> -----Original Message-----
> From: SPSSX(r) Discussion on behalf of Matthias Spörrle
> Sent: Mon 5/14/2007 7:24 PM
> To: [hidden email]
> Cc:
> Subject: Re: scale reliability
>
>
>
>         Helga,
>
>         just some thoughts concerining your questions:
>
>         Alreck (Consumer age role norms, then and now, 2000) has developed
> a scale
>         you might find interesting, because it measures age role norms
> (example
>         item: "Sex and dating should be a thing of the past for people in
> their
>         sixties or older") which "can readily be seen as dimensions of age
>         stereotypes" (p. 907).
>         You can use this scale for the development of a new version of
> your scale or
>         for comparing the content of the items of the two scales (to
> examine
>         [content] validity). See her references for further papers
> concerning
>         stereotypes which might also be helpful for your validity
> question.
>
>         See Charter (Study samples are too small to produce [...], 2003)
> for a
>         discussion of adequate sample size concerning reliability
> estimates.
>
>         HTH
>         Regards
>         Matthias
>
>
>
>
>         On 5/15/07, Helga Walz <[hidden email]> wrote:
>         >
>         > To the experts with thanks ahead of time:
>         >
>         > For a community outreach project I collected data for 20 items
> reflecting
>         > stereotypes of aging. Because I could not find an existing
> scale, I
>         > developed items using a 5-point Likert type response format.
> Responses for
>         > the steretoype items were collected separately from young and
> older adults.
>         > The sample size is approximately 25 for young and 25 for older
> adults. It
>         > was suggested that I do a Repeat Measures ANOVA for item
> analysis or scale
>         > analysis (cannot do Factor Analysis, sample too small).
>         > I would appreciate assistance in determing whether 1) the items
> do,
>         > indeed, indicate stereotypes of aging 2) the validity of the
> combined items,
>         > and 3) how to arrive at the answers using SPSS - so that I may
> explore
>         > stereotypical behaviors and their relationship with other
> variables.
>         >
>         > Thank you so much,
>         > Helga
>         >
>
>
>
>

[Melissa Ives]

Why not use the Reliability command?  When you get the output, use the
'Results Coach' to help interpret.

Melissa

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Helga Walz
Sent: Monday, May 14, 2007 5:18 PM
To: [hidden email]
Subject: [SPSSX-L] scale reliability

To the experts with thanks ahead of time:

For a community outreach project I collected data for 20 items
reflecting stereotypes of aging. Because I could not find an existing
scale, I developed items using a 5-point Likert type response format.
Responses for the steretoype items were collected separately from young
and older adults. The sample size is approximately 25 for young and 25
for older adults. It was suggested that I do a Repeat Measures ANOVA for
item analysis or scale analysis (cannot do Factor Analysis, sample too
small).
I would appreciate assistance in determing whether 1) the items do,
indeed, indicate stereotypes of aging 2) the validity of the combined
items, and 3) how to arrive at the answers using SPSS - so that I may
explore stereotypical behaviors and their relationship with other
variables.

Thank you so much,
Helga


PRIVILEGED AND CONFIDENTIAL INFORMATION
This transmittal and any attachments may contain PRIVILEGED AND
CONFIDENTIAL information and is intended only for the use of the
addressee. If you are not the designated recipient, or an employee
or agent authorized to deliver such transmittals to the designated
recipient, you are hereby notified that any dissemination,
copying or publication of this transmittal is strictly prohibited. If
you have received this transmittal in error, please notify us
immediately by replying to the sender and delete this copy from your
system. You may also call us at [phone] for assistance.