The term median split refers to
A) including only the scores above the median of the dependent variable in a multiple regression analysis.
B) classifying those who score above the median on a measure as high on the construct being measured and those who score below the median as low on the construct.
C) reducing multicollinearity by using sample medians rather than sample means when conducting multiple regression analyses.
D) splitting a data set into multiple subsamples, running regression analyses in each subsample, and using the median value of each regression coefficient as the best estimate of that coefficient.

Question

Quizplus · Accepted Answer

Median split refers to the practice of dividing a group of participants into two halves based on their scores on a particular measure, with those who score above the median being classified as "high" and those who score below the median as "low." This is often done in research studies to create groups for comparison or analysis. Option A is incorrect because median split does not refer to only including scores above the median in a regression analysis. Option C is also incorrect because median split does not address multicollinearity in regression analysis. Option D is incorrect because it describes a different statistical technique.

The Term Median Split Refers to