PSYC 3301 Lecture Notes - Lecture 3: Pooled Variance, Repeated Measures Design, Null Hypothesis
Document Summary
Two groups of data groups are independent (between groups design) Reliable and appropriate manipulation of an independent variable at two levels. Reliable and valid measures of the dependent variable. Control of extraneous variables: hold constant, elimination, random assignment, balance across groups, herb. Evaluate the mean differences between the two treatment conditions. High internal validity so we know that. High external validity so we know that. Pooled variance: is obtained by averaging the two estimated sample variances. It makes sense to average these two things only if they both are estimating the same population variance: both groups are represented equally, proportionate representation. If they are estimating different population variances, then averaging them is meaningless. It"s the sa(cid:373)ple (cid:373)ea(cid:374) differe(cid:374)(cid:272)e divided (cid:271)y the esti(cid:373)ated sta(cid:374)dard error. Observed difference between groups divided by the difference expected by chance: t = observed difference difference expected by chance, observed difference, m1 m2, denominator: s (m1-m2, formula, m1 m2.
