How would you show respondent ideal points for a related question in the same survey? Specifically, how would you chart each respondents' ideal attribute groupings along the same dimensions created in this question on the same chart?

Dataset is structured same as the one in that linked question. But it includes an additional question that gets at the attributes respondents believe to be most desirable when generally deciding upon a purchase. Output is the same: 0 = "No, I do not associate this attribute with what is desirable when choosing this product" and 1 = "Yes, I associate this attribute with what is desirable when choosing this product".

My first thought was to somehow create one column per respondent so that I could deploy the same technique suggested in the above solution. But I don't know how to transform the row data. Another thought was that SPSS may have the ability to only crunch data for the question mentioned on a respondent-by-respondent basis--so pretty much the program would somehow automatically cut data at the respondent level and perform the computation I need for this question (just like the brand question above).

Is it possible to show the code that will permit the presence of ideal points per respondent in addition to the brand points identified in the above solution?

  • $\begingroup$ Because the question you link to can be open and observed there is no need to copy that thread to here, so I removed it. $\endgroup$ – ttnphns Sep 3 '15 at 18:47
  • $\begingroup$ Regarding ideal point - it is a good idea to include the data for it as a "supplementary category/ies". Such point(s) is plotted on the biplot but do not influence the configuration of other points on it. Ideal profiles are routinely processed this way in CA. (For math behind supplementary category, if you need, please look here.) $\endgroup$ – ttnphns Sep 3 '15 at 18:51
  • $\begingroup$ I know that ideal points do not influence the configuration of other points. I only seek to understand how to show ideal points on the perceptual map. Separately, I don't know what this refers to: "supplementary category/ies". Regarding the math piece, thanks. $\endgroup$ – user87618 Sep 3 '15 at 19:16
  • $\begingroup$ It refers to an option in SPSS correspondence analysis, I thought you know. $\endgroup$ – ttnphns Sep 3 '15 at 21:06
  • $\begingroup$ If I seek to cross Attribute with Brand AND include a third variable (id) as a "supplementary category/ies" variable, does this require multiple correspondence analysis? If so, how would it be executed against the first example (stats.stackexchange.com/questions/68938/…)? $\endgroup$ – user87618 Sep 3 '15 at 22:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.