2
$\begingroup$

In a sample of 250 cardiac patients who have had a myocardial infarction and just left hospital, I asked them how important it was to them to know several issues, such as "How important is it for you to know when to return to work, or when to revisit the hospital" etc. Answers were provided ranging from "not important at all" to "very important" (Likert-type scale with 5 answers, assuming all answers have "equal" distances between them).

I would like to know:

  1. How to treat the answers to the above questions (as categorical?).
  2. How to look at whether there is any association between the above type of questions and other questions that are categorical (for example: "Who do you prefer to give you information": a-physician, b-nurse, c-other; or "Level of education": a-primary level, b-secondary level, c-bachelor, or equivalent d-postgraduate degree).

This questionnaire hasn't got any cummulative score or subscale scores that could be used for testing associations.

$\endgroup$
  • $\begingroup$ "How important is it for you to know when to return to work, or when to revisit the hospital" this is an example of a double barreled question. If you asked it this way you wouldn't be certain if they were implying that knowing when to return to work, or revisiting the hospital is not at all important through very important. $\endgroup$ – Brandon Bertelsen Apr 1 '12 at 7:05
  • $\begingroup$ These are two distinct, seperate questions, I only presented them as an example. $\endgroup$ – Dimos Apr 1 '12 at 9:29
1
$\begingroup$

If you assume equal spaces between the points of the your Likert-type scale, as you write it, then you agree to treat the scale as metric (interval or ratio) rather than categorical. Interval-by-interval correlation such as Pearson r and Interval-by-nominal correlation such as eta are examples of statistics you can use, in response to your point 2 (go to Crosstabs procedure in SPSS). To check whether underlying association between your metric variables are reasonably linear you might compare Pearson r with Spearman rho or Brownian distance correlation.

$\endgroup$

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.