Hi

It depends on what you are trying to determine with the data. If you are just 
interested in preferences of the 10 types, you could think about some count for 
each of the 10 objects (e.g., # times chosen first, # times chosen at all, or 
some other categorization). A chi2 could determine if there is significant 
variation in preference for the 10 types. If you had predictions, it is 
possible to partition the overall chi2, much as you would partition a main 
effect in anova with planned contrasts.

The trick in translating the individual trials to a numerical score is just 
using the rank, as you mentioned (e.g., 0, 1, 2, 3  for not chosen, 3rd, 2nd, 
and 1st) or determining the appropriate weighting for each rank to properly 
reflect the underlying preference dimension (e.g., 1, 2, 4, 8 if each rank 
doubles the preference). Such scores might make it easier to analyze group 
differences in preferences (e.g., by gender), although that could be done with 
the count data as well.

Good luck

Jim

Jim Clark
Professor & Chair of Psychology
University of Winnipeg
204-786-9757
Room 4L41A (4th Floor Lockhart)
www.uwinnipeg.ca/~clark


-----Original Message-----
From: Annette Taylor [mailto:[email protected]] 
Sent: December-27-15 12:01 PM
To: Teaching in the Psychological Sciences (TIPS)
Subject: [tips] back to the stats well

I need to know if there even is an adequate statistical test, other than using 
descriptives for this situation:

We asked people to order their number 1, 2, 3 choices from a list of 10 
options. So 7 options were essentially all tied at 0. So this would be ranked 
and ordinal data to the best of my understanding. I wonder if we should have 
made them rank all 10? But we really weren't interested in anything less than 
the top three.

We'd like to see whether there is a systematic, not attributable to chance way 
to characterize the choices that people made. 

Any ideas? I have been directed to a website that offers what seem to me to be 
partial solutions but I'd like to see if any of you have any other suggestions 
that are not biased by the other suggestion.

Thanks!

Annette


Annette Kujawski Taylor, Ph. D.
Visiting Professor,
Ashoka University, Delhi, India
[email protected]
Professor, Psychological Sciences
University of San Diego
[email protected]
---
You are currently subscribed to tips as: [email protected].
To unsubscribe click here: 
http://fsulist.frostburg.edu/u?id=13251.645f86b5cec4da0a56ffea7a891720c9&n=T&l=tips&o=47685
or send a blank email to 
leave-47685-13251.645f86b5cec4da0a56ffea7a89172...@fsulist.frostburg.edu

---
You are currently subscribed to tips as: [email protected].
To unsubscribe click here: 
http://fsulist.frostburg.edu/u?id=13090.68da6e6e5325aa33287ff385b70df5d5&n=T&l=tips&o=47686
or send a blank email to 
leave-47686-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu

Reply via email to