PSYC2203 Lecture Notes - Lecture 7: Null Hypothesis, Scatter Plot
Document Summary
Reliability: test/retest, alternate forms (2 di erent forms of the same test), split-half (correlations of peoples performance). The extent to which there is a clear relationship between distributions of 2 variables. Positive linear: high scores on one go with highs on the other. Negative linear: high scores one go with lows on the other. High correlation/strong relationship = data points on scatterplot fall very close to straight line sloping up or down. Link deviation scores from the mean for one variable to deviations from the mean on the other variable. Multiply deviation scores for both variables for each person. Must then standardise these relationships bc using di erent types of values and di erent numbers of values in di erent data sets. Standardise scores rst by transforming them into equivalent z-scores (relationships stay the same, and now everything is on the same scale - thus can compare correlation numbers directly).