PSYC 2740 Lecture Notes - Lecture 4: Discriminant, Lee Cronbach, Face Validity
Document Summary
Most personality research is in the form of a correlational investigation in which two or more personality scales are administered to a sample and these scales are compared to determine if they are significantly associated with each other. Positive correlation: as one variable tends to increase, the other increases to a corresponding degree. Negative correlation: as one variable tends to increase, the other decreased to a corresponding degree. Cohe(cid:374)"s (cid:373)ag(cid:374)itude of (cid:272)orrelatio(cid:374)al effe(cid:272)ts: . (cid:1005)(cid:1004) = s(cid:373)all, . (cid:1007)(cid:1004) = medium, . 50 large. Important to note that statistically small correlations can have practical significance ie. taking aspirin and preventing a heart attack (r = . 03) If two variables are related, all that can be stated is that they are associated to each other; it cannot be assumed that the variable associated to the other caused or contributed to the variability of the other variable. Cannot tell us which variable comes first when two variables are significantly correlated.