cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Try the Materials Informatics Toolkit, which is designed to easily handle SMILES data. This and other helpful add-ins are available in the JMP® Marketplace
Choose Language Hide Translation Bar
pdawgy13
pdawgy13
Level II

Help with determining most frequently chosen combination of answers?

Jan 4, 2016 03:19 PM (10399 views)

Hello all.  I am new to JMP and this user community.  I am currently conducting an analysis on data collected across my organization, and am "stuck" on how to obtain the information I need from the data.

86 respondents (managers of conservation programs) were asked which conservation strategies of a list of 24 (labeled a through x) that they were using in their programs.  They were not given a limit on how many they could choose.  We are looking to determine the most frequently chosen combination of strategies, for use in determining whether there are certain "models" or "typologies" under which these projects can be categorized. 

The average number of strategies chosen by each respondent is about 10.  We are looking to tease out 5-10 "models" or "typologies" based on the most frequent combinations of strategies.  I have the data set up in a matrix with each strategy given its own column (a, b, c, ...) and each respondent given its own row.  If a strategy was chosen, it's letter is in the cell, if it is not chosen, the cell is left blank.  I also have another version where choice is denoted by a 1 and non-choice by a 0.

Any ideas on how we might crack this nut would be very welcome!!

Warm Regards,

Nikki

10 REPLIES 10
pdawgy13
pdawgy13
Level II

Re: Help with determining most frequently chosen combination of answers?

Jan 5, 2016 03:06 PM (743 views)  |  Posted in reply to message from stan_koprowski 01-05-2016

Thank you Stanley, both for the suggestion and for going to far as to actually calculate as an example.  If I am understanding correctly, what this analysis does is show the likelihood of one answer being chosen if another answer is chosen.  For example, if I know "a" was chosen, the likelihood that "b" would also be chosen, etc.  The cell plot aids in visualization of the numbers.  However, does this plot say anything about the likelihood of "b" AND "c" being chosen if I know "a" was chosen, etc.?

Warmest Regards,

Nikki

0 Kudos