Power Analysis

Question 1

A statistical tests power is the probability that it will…

Answer 1

Avoid Type 2 error/correctly pick up on an effect

Question 2

This hypothesis states that the difference between the two variables is zero

Answer 2

Null

Question 3

One motivation for conducting a study is to provide evidence that the hypothesis is true. We do this by showing…

Answer 3

The null hypothesis is unlikely

Question 4

A Type 1 error is where

Answer 4

False positive

Question 5

A Type 2 error is a

Answer 5

False negative

Question 6

“Rejection of the null, when it is in fact true”

Answer 6

Type 1 error

Question 7

What is an effect?

Answer 7

What was found

Question 8

What is effect size

Answer 8

Measures strength of result - magnitude-based

Question 9

Difference between EFFECT SIZE and SIGNIFICANCE?

Answer 9

Effect size does not depend on sample size

Question 10

The p value indicates if what was found is g..

Answer 10

Generalisable to a population

Question 11

Effect size does not depend on

Answer 11

Sample size

Question 12

If we find a non-significant p-value despite a large effect, this is likely because

Answer 12

Small sample

Question 13

A prospective power analysis shows us that a certain ________ are needed to yield a significant result

Answer 13

Number of pps

Question 14

Sufficient power to find statistical significance minimises

Answer 14

Chance findings

Question 15

A power analysis is where one of several _______________ can be calculated given others

Answer 15

Statistical parameters

Question 16

There are ___ parameters involved in a power analysis

Answer 16

4

Question 17

4 parameters involved in a power analysis

Answer 17

(i) Alpha
(ii) Power
(iii) N
(iv) Effect size

Question 18

Coefficient alpha is the p…

Answer 18

Probability of finding significance where there is none - Making a Type 1 Error

Question 19

Coefficient alpha is usually set to

Answer 19

.05

Question 20

Power is usually set to

Answer 20

.80

Question 21

Power is the probability of finding

Answer 21

True significance

Question 22

Power is the probability of avoiding

Answer 22

A type 2 error

Question 23

Power analysis

The parameter usually solved for is…

Answer 23

N

Question 24

Expected effect size may be ascertained from… (4)

Answer 24

1. Pilot study
2. Findings from similar study
3. Field defined ‘meaningful effect’
4. Educated guess

Question 25

Alpha is usually set to ___ and power is usually set to ___

Answer 25

.05 | .80

Question 26

When trying to ascertain effect size, if there are no similar studies...

Answer 26

Start with pilot study

Question 27

There are FOUR TYPES OF POWER ANALYSIS

Answer 27

Priori Post-hoc Criterion Sensitivity

Question 28

For a priori, you'd compute ___. The rest is given

Answer 28

Sample size (N)

Question 29

For a post-hoc, you'd compute ____. The rest is given

Answer 29

Power

Question 30

For a criterion, you'd compute ___. The rest is given

Answer 30

Alpha

Question 31

Which type of power analysis is USUALLY DONE

Answer 31

Priori

Question 32

A 72% power indicates that you have a 72% chance of

Answer 32

Avoiding making a Type 2 error/Picking up on an effect

Question 33

Power is increased when a researcher increases

Answer 33

Sample size/Effect Levels/Significance Levels

Question 34

_______ is increased when a researcher increases sample size/effect levels/significance levels

Answer 34

Power

Question 35

How to calculate effect size

Answer 35

M1 + M2 / SD (collective)

Question 36

A power analysis is where

Answer 36

One of several statistical parameters can be calculated given others

Question 37

Reporting a prospective priori order

Answer 37

1. Pps 2. Effect size 3. Alpha 3. Power

Question 38

Reporting a sensitivity order

Answer 38

1. Pps 2. Effect size 3. Alpha 4. Power

Question 39

Reporting a post-hoc

Answer 39

1. Effect size 2. Pps 3. Alpha 4. Only achieved 73% power

Question 40

How do you begin reporting a post-hoc

Answer 40

The effect size for this study was calculated as

Question 41

How do you begin reporting a retrospective priori

Answer 41

The effect size for this study was calculated as

Question 42

Statistical power depends on (3)

Answer 42

1. Effect size 2. Sample size 3. Precision of measures

Question 43

Why does statistical power depend on precision of measures?

Answer 43

More reliable measures = smaller SE's/more precise variable estimates

Question 44

Effect size The ______ the better

Answer 44

Larger

Question 45

Sample size Larger sample size =

Answer 45

Smaller SE's

Question 46

Error bars Small samples will have

Answer 46

Large error bars

Question 47

Error bars Easier to estimate when

Answer 47

Error bars are smaller

Question 48

Large error bars = _____ sample size

Answer 48

Small

Question 49

Prospective power analysis We don't want to recruit too many pps

Answer 49

As this could result in significant result regardless of effect size

Question 50

Other than a Priori, you could calculate required sample size

Answer 50

Using Cohen's table

Question 51

Why does sample size affect power

Answer 51

Larger = smaller SE's | Small, non-overlapping error bars