0

I am currently working on a research proposal in which I am testing a new concussion recovery method's effects on personality change from concussions. So right now, my IV is either New Recovery or Traditional Recovery which is nominal. My DV is whether or not personality changed, so I feel this would be nominal since it seems to be mutually exclusive. With that being said, unless I am incorrect with labeling the DV as nominal, how would I plot this data? I am having difficulty determining how to best represent this data.

Thank you

Brandt
  • 1

1 Answers1

0

With only four numbers (one per combination of IV and DV values), you're better served with a confusion matrix (i.e., contingency table) than a plot.

Kodiologist
  • 19,063
  • 2
  • 36
  • 68
  • The contingency table of the data wouldn't be quite the same as a confusion matrix. – gung - Reinstate Monica Apr 20 '18 at 19:57
  • @gung Why not? Isn't a confusion matrix just a square contingency table where you regard one dimension as the prediction and the other as the real outcome? – Kodiologist Apr 20 '18 at 21:41
  • A (2x2) contingency table and a confusion matrix are both tables / matrices w/ the same dimensions & filled with counts. That makes them seem very similar. But what the numbers represent is ontologically different. Moreover, if you had contingency table data & fit, say, a logistic regression model, then converted the predicted probabilities into predicted categories to make a confusion matrix, you would end up w/ a matrix w/ 2 0's & 2 non-0 values. Ie, it wouldn't be the same numbers as the original data. – gung - Reinstate Monica Apr 21 '18 at 00:45
  • @gung "But what the numbers represent is ontologically different." — How? I mean, if you gave me a 2 × 2 confusion matrix and told me it was a contingency table, how would I tell that you were wrong? What's the difference? – Kodiologist Apr 21 '18 at 07:20
  • You could ask a new question about this, if you wanted. If you were simply given an unlabeled 2x2 matrix, w/ no information, you couldn't tell whether it was a contingency table, a confusion matrix, or something else. They still mean different things. – gung - Reinstate Monica Apr 21 '18 at 12:19