Module 3: Magnitude of Effects

Question 1

Bayesian advocates state that just because the null goes poorly does not mean…

Answer 1

The alternative will go well

Question 2

Bayes factor

Answer 2

Ratio of the likelihood of the alternative hypothesis relative to the likelihood of the null hypothesis

Value of 1 = equal likelihood of alternative relative to null

Value less than 1 = null more likely

Value greater than 1 = alternative more likely

Question 3

Imagine two experiments. Experiment 1 yields a p-value of .001 and experiment 2 yields a p-value of .01. The chance explanation is LESS viable in experiment 1than experiment 2 but can we say that experiment 1 has produced a larger more systematic effect than experiment 2?

Answer 3

No. We would need to know the sample sizes of these experiments were identical. Concluding the null is unlikely is not the same asconcludingthe magnitude of effects is large.

Question 4

Standardized effect size indicescan…

Answer 4

be applied to measures of different metrics and express magnitude of effects in common metric

Question 5

What is a small, medium, and large effect size associated with Cohen’s d

Answer 5

Small effect size – 0.2, medium effect size – 0.5, large effect size – 0.8

Question 6

Cohen’s dsand Cohen’sdav

Answer 6

Standardized effect sizeindices


Used when therearetwo means to compare

Question 7

Pearson’s r coefficient

Answer 7

Standardized effect sizeindices
Canbecalculatedtoexpressthestrengthanddirectionofassociationbetweentwocontinuousvariables(rangesfrom–1to+1)

Question 8

Wheninterpretingrasaneffectsize,whatisconsideredsmall,medium,andlarge?

Answer 8

Smalleffectsize-.10,mediumeffectsize-.30,largeeffectsize-.50


Question 9

What is an important consideration when interpreting whether an effect size is significant or not?

Answer 9

Researchers must consider other extraneous factors that may influence the practical significance (I.e., durability, cost/benefit analysis)rather than looking merely at thresholds

Question 10

Two contexts in which small effects may be considered impressive…

Answer 10

1. Minimal manipulations of the independent variable

2. The dependent variable is difficult to influence

Question 11

Small standard errors produce __________ confidence intervals

Answer 11

Narrow


Question 12

How do we interpret a traditional 95% confidence interval?

Answer 12

There is a 95% chance that the interval calculated contains the population value
OR
If we ran this study 100 times, 95 of our studies would generate confidence intervals that contain the population value

Question 13

Consider a confidence interval of-.10 to .7. This confidence interval contains 0 which indicates what?

Answer 13

There is a 95% chance that the interval contains the effect size and one of these values is 0 (meaning no effect at all)

Question 14

Cohen’s d bias

Answer 14

Due to high sampling error in small samples, cohen’s d overestimates the effect size.

Question 15

What corrects for Cohen’s d?

Answer 15

Hedges g