Lecture 2 Exam 2
UIC BioS 101 Nyberg
9
Categorical Data
-
Discrete (non-overlapping) categories that collectively cover all possibilities
-
If one has a model that generates probabilities of being in each possible category, one can use the Chi-squared test to compare the observed outcomes (=numbers) to the expected numbers once the total number of observations is known.
Speaker Notes:
9
It is important that the categories are mutually exclusive (non-overlapping) and cover all possibilities (100%).