chapter 3

Question 1

publication bias

Answer 1

• the bias that journals mainly (90%) publish studies with significant differences only

Question 2

p hacking and harking

Answer 2

• cherry picking data that is shared/selective reporting
• Hypothesizing after results are known
• not controlling type 1 error rate

Question 3

researcher degrees of freedom

Answer 3

ways a researcher can influence the p value

Question 4

acronym for addressing problems w NHST

Answer 4

• Effect size
• Meta analysis
• Bayesien Estimation
• Registration
• Sense

Question 5

wilcoxon rank sum test

Answer 5

does not assume normality and is used for comparing two independent groups.

Question 6

what is the reason for a post hoc test

Answer 6

• performed after ANOVA
• used to give more results that ANOVA does not cover
• ex. fishers LSD, tukey, etc.

Question 7

mean square formula

Answer 7

SS/df

Question 8

what do you typically want to be smaller, between or within variance

Answer 8

between should be larger than within (makes samples more independent of one another)

Question 9

bonferroni

Answer 9

• one of the most conservative tests
• partitioning the p-value acorss all tests
• ie if you have more than 2 samples, split the p value between then so it is still 0.05 when all tests are combined
• used as an alternative to ANOVA if you dont want to look at all the variables

Question 10

what do r=0.1, r=0.3, and r=0.5 mean

Answer 10

• 0.1= small effect, it accounts for 1% of variance
• 0.3= med effect, it accounts for 9% of variance
  -0.5= large effect, it accounts for 25% of variance

Question 11

t or f: in a pearsons r the r scores are measured on a linear scale

Answer 11

false

Question 12

odds ratio formula

Answer 12

odds= p(event)/p(noevent)