Multiple regression: R square from ANOVA TABLE 1st line divided by regression line Coefficient table DV=a + b1=b2 ..... Df(1st and second line no total) EFFECT SIZE provides a quantitative measure of the strength of a phenomena Eta squared the preferred effect size for single factor ANOVA SSB/SSTOTAL F statisfics is calculated variability between group/variability within groups (In stat variability is a matter of how spread out the data is!) A power analysis tells us how large a sample we need in order to be reasonably confident of detecting a difference, if it exists in the population Post hoc test is an additional hypothesis test that are done after an ANOVA to determine exactly which mean differences are sign and which are not for not specific predictions One way ANOVA df( between, within). Effect size/ eta squared: between divided by total Metric dv, independent normal and equal (like t test) Table df resid= within group df at a level 0.06 which is eta medium effect ( 0.80 power!) table for group the number represent each group Between subject for equal assumption met is Lavene's within subject sphericity/Maucly's BONFERRONI'S ADJUSTMENT we adjust the sign level based on the level 0.05/(n of contrast) for contrast interpretation Single factor ANOVA: ass2 ass 2 A study was conducted to ..... it was hyphothesised that ... table/figure bellow provides the descriptive stat for this scenario ... As can be seen in T or F describe in order of M and SD ... A single factor ANOVA showed that F(btwn/under)=,p,eta squared ... a planned contrast/ A-S-H-K test revealed that ... as expected/contrarily to expectation Between no interaction: ass 2 A study was conducted to explore the differences in maths enjoyment for children in different school sectors. It was hypothesised that children from Catholic schools would enjoy maths more than children from either independent or government schools. They also suggested that the difference in enjoyment across sectors would be greater for boys than for girls. Means and standard deviations for maths enjoyment are presented in Table 1. A three (sector) by two (sex) analysis of variance indicated that there was no significant interaction between school sector and sex of student, F(2,324) = . 12, p = . 89. There was a significant difference in mean levels of math enjoyment across different school sectors, F(2,324) = 6.86, p = . 001, partial n 2 = . 04. Planned contrasts showed that Catholic schools, on average, produced significantly higher levels of math enjoyment than government schools, t(327) = 3.38, p = . 001 and independent schools, t(327) = 3.02, p = . 003. There was no significant difference in mean levels of math enjoyment between males and females, F(1,324) = 1.69, p = . 195. The hypotheses were only partially supported. As expected, Catholic schools produce higher levels of math enjoyment than government and independent schools. However, there was insufficient evidence to suggest that the difference in maths enjoyment across sectors would be greater for boys than for girls.