PSY 2174 Lecture Notes - Lecture 6: Operational Definition, Null Hypothesis, Total Correlation

• School: University of Ottawa
• Department: Psychology
• Course: PSY 2174
• Professor: Heather Poole

Document Summary

We use measures to assess all kinds of construction. Often heard to observes directly, so we count on our measures. Valid ( measure what its meant to measure) Statistically measure using correlation between instance 1 and instance 2. Degree to which all items of a test measure the same thing. Average of correlations between all split halves. Delete the one that would increase alpha most. Continue until alpha >0. 75 item - total correlation. Delete any that show an opposite trend from the overall score. Degree of agreement between two independent raters. Factors influencing reliability: clarity of operational definition, ex how much of the course material do you know, extend to which operations definition is followed, number of observations on which overall score is based. The range of scores that we need to cover to get a good idea of performance on our variable of interest. When considering range, we must consider range, we must consider.