Hi,

I guess this is why questionnaires usually
consist of "multiple choice" questions rather than "essay" questions <g.

You hit it right on the nail. If only all questionnares were Closed-end
Multiple choice questions, things would have been so much easier.
Why closed end? Because even in a multiple choice question most of the times
(a very substantial amount) we have one option called "Other" (which is open
ended kind of question as a respondent can mention anything under the sun
within the other option). And *Other* is again to be dealt in the same way
like an Essay question.

For ex. if the question is --" Which of the following reasons, if any,
describe why you did not purchase anything today? (Please check all that
apply)"

a) You didnt find a style you liked
b) You couldn find your size
c) A product didnt fit right
d) The product was too expensive
e) You werent planning to buy/ just browsing today
f) Other (Please specify below)

In the above case option f) is a candidate for Essay kind of manipulative
analysis (Please note, there were many more options between e) and f) but
couldnt type anymore. Above is just for clarifcation/demonstration).

A query. Any idea whether "data file" used by dictionary program or
spell-checker would be availbale somewhere in the web.


Thanks a lot,
Hari
India

"Myrna Larson" wrote in message
...
Hmmm.... I see your problem. I guess this is why questionnaires usually
consist of "multiple choice" questions rather than "essay" questions <g.

If you can get into a data file used by a dictionary program or
spell-checker,
that might be a start.

To answer your question,
On Tue, 25 Jan 2005 20:21:21 +0530, "Hari Prasadh"

wrote:

Hi Myrna,

Just to give you an idea, I analyse MR data.

So if a question is asked -- "Please describe this product" -- and if one
respondent says "It's very Beautiful" and if another respondent says "It's
a
thing of beauty".
For me end meaning of both the sentences is same. So, before loading to
SPSS( statistics software) I would assign a code ( numerical equivalent to
a
particular verbal response) which is equal for both the responses. (I
would
like to perform analysis starting from simple like how many rated product
being a beauty etc. to more complex depending on client requirements).

So, what I meant was to have a dictionary (in soft format having nouns
adjectives of every word) against which I could check and programmatically
manipulate the same. For me in this case (programming has been fine
because
of Tim W's help) but now Iam on another barrier which is to get a list of
all words in let's say English language.

Please note the above question was just a sample, a respondent could use
multiple "adjectives" or "nouns" to describe the product. Like for Apple
I-pod they may say "It's a thing of beauty, very cool, so many
options...."
etc.

Thanks a lot,
Hari
India

"Myrna Larson" wrote in message
. ..
if a word"beautiful" and "beauty" appears in the
target array, then for me both are one and the same

You may not have much company in that opinion. Beautiful is an
adjective,
beauty is a noun. You can't take a sentence and that uses one of these
words
and replace it with the other and end up with something that is
gramatically
correct.