Request for Multidimensional Preference AnalysisMultidimensional Preference Analysis (MDPREF): This capability is currently not available in JMP. This method is useful for representing 2-way data where the rows and columns represent different things, like a summary table. The rows are represented as vectors and columns as points and the length of projection of the points on the vectors represent the preference. This will be useful for various types of data including sensory data. This will also be useful for maxdiff data where the rows are people and columns are items and if you can represent segment information with color coding.
GraphicsStatistics and AnalyticsUser InterfaceRe: Request for Multidimensional Preference AnalysisHi @mjoner,
It is on our list for consideration. It won't be in JMP 16 but we'll keep it in mind as we start JMP 17 planning.
Re: Request for Multidimensional Preference AnalysisAny news on whether this could be developed?
Although number of uses for this are still currently low we are seeing increased interest in this. One solution will be to have people export their JMP tables in order to enable us to run the analysis using intranet servers. We would like something better, however.
Re: Request for Multidimensional Preference AnalysisBiplot via Principal Components is one way of handling it. However, this model is unique to preference data. Available in SAS.
Re: Request for Multidimensional Preference Analysis - Status changed to: AcknowledgedRe: Request for Multidimensional Preference AnalysisWe have in the meantime pursued avenues to accomplish this outside of JMP. We should start gathering some statistics soon on how much we are really using this capability.
Re: Request for Multidimensional Preference Analysis - Status changed to: ArchivedArchiving based on discussion with P&G. Please let me know if you would like this reopened.
