i think there are two issues here:
1. does the treatment work? (which is unspecified in this example ...)
2. are the two measures equivalent?
the answer to 1 is based on good experimentation
the answer to 2 is in some relationship investigation between the two measures
there is no good way in this study to make any inference about emotional
CHANGE ... and therefore, no good way to infer that scores on the emotional
measures have the same "import" as a change agent as possibly for the
physical ... the DESIGN just will not allow it
but you can correlate the two measure scores (which is a simple thing to
do) ... and see what you get but, of course, there are about 2 times as
many items on physical than emotional ... thus, the correlation between the
two will be impacted somewhat by differential reliabilities ... and of
course, how much variance there might be on either or both measures
At 08:32 AM 3/30/01 +0000, [EMAIL PROTECTED] wrote:
>dennis roberts writes:
>
> > i think it is too early to be worrying if FA should be used ...
> >
> > i am not quite sure what you are trying to do here ... you have measure A
> > that you do pre and post ... and measure B that you do only post ...
> >
> > why this difference?
>
>Most researchers use pre & post when they want to find out how well a
>certain treatment works, but this is not always possible. For instance when
>a patient is too ill to fill in a questionnaire before treatment or when he
>or she is brought in in an emergency. Only doing a posttest would save a
>lot of time, effort and money. By reformulating the questions of an
>existing scale into 'transition'-items, we have to make sure that we're
>still measuring the same constructs.
>
> > by convergent validity ... we are usually interested in finding several
> > sources of information that all point in the same direction ... to some
> > effect occurring ...
> >
> > in your case, you have physical and emotional ... is your reason for
> > inquiring basically ... are the data from the two measures consistent?
>
>Not yet. First we want to know if we can reasonably assume that the two
>measures measure the same constructs (physical and emotional wellbeing).
>After that we will check if the 2 measures are consistent. I realize that
>these questions are related.
>
> > so ... if you could provide a bit more of what you are trying to do with
> > your data ... AND, a little more of WHAT the data really are ... what
> kinds
> > of items? how many? etc. etc. ... would help
>
>There are 16 items, of which 11 intent to measure physical wellbeing and 5
>emotional wellbeing. They are all part of an existing scale. N=218.
>
>I hope this extra information makes it clearer for you. Thanks for all the
>reactions I had up to know.
>
>Heike
_________________________________________________________
dennis roberts, educational psychology, penn state university
208 cedar, AC 8148632401, mailto:[EMAIL PROTECTED]
http://roberts.ed.psu.edu/users/droberts/drober~1.htm
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
