PSYC 217 Review Ch14 Research in lab and field experiment should show similar results Conceptual replications are more important than exact replications - not making same errors Solomon 4 group design assesses the pretest effect Meta-analysis - most important effect size Failure to replicate findings has same problems as non-significant findings because difficult to interpret results Ch 12 r^2 - variability, coefficient of determination Stdev = square root from variance Structural equation models= expected models of relationships Effect size - scale consistent for all the studies If no relationship - draw a scatterplot - e.g., curvilinear relationship Median - ordinal Mode - nominal Mean - interval and ratio Restrict the range - decrease the effect of r Histogram - quantitative; bar graph - categorical Criterion and predictor variables -> Regression equation Ch.13 Alpha = Type 1 Type 2 = 1 - power T-test - 1 tailed -> predicting in 1 direction -> iq 0 -> iq 100 - 2 tailed -> predicting in 2 directions Systematic variance = between-groups variance Chi-square - between categorical variables Analysis of variance - multiple variables Type 1 and 2 errors are inversely related