Flashcards in Psychology 202 - Test 2 Deck (28):

1

## hypothesis testing

### a statistical method that uses sample data to evaluate a hypothesis about a population

2

## hypothesis

### the prediction about the relationship between two variables

3

## critical value

### set cut-off sample score

4

## directional hypothesis

###
make a prediction regarding direction

i.e. increase...or decrease...

5

## one-tail test

###
looking at one tail/extreme of the distribution

- 5% > 1.64

- 1% > 2.33

6

## non-directional hypothesis

###
no prediction regarding direction

- just know there is a change, don't know in/decrease

7

## two-tail test

###
need to look/think about both extremes

- 5% > 1.96

- 1% > 2.57

8

## distribution of means

### set of sample means from a given population

9

## rule #1

### the mean of a distribution of means (μm) is the same as the mean of the population of individuals

10

## rule #2a

### The variance of the distribution of means is the variance of the population of individuals divided by the number of individuals in each sample

11

## rule #2b

### the standard deviation of the distribution of means is the square root of the variance of the distribution of means

12

## rule #3

###
the shape of the distribution of means is approximately normal if at least one of the conditions is met

- sample size is 30 or more

- the distribution of the population of individual scores is normally distributed

13

## Variance & SD formulas for Distribution of Means

###
variance --> δ²m = δ² / N

SD --> √δ²m = δ² / N or √δ²m

14

## z-test

###
hypothesis testing procedure using the mean of the sample when the population variance is known

- comparing sample mean to distribution of means

15

## z-test formula

### Z = (M - μm) / δ²

16

## statistical significance

### the number is so extreme it is unlikely to have gotten it by chance

17

## Alpha (α) - Type I Error

###
- the null is true and the data tells us to reject

i.e. jury finding innocent man guilty

18

## Beta (β) - Type II Error

###
- the null is false and the data tells us to retain it

i.e. jury lets guilty man go free

19

## comparing studies

### - as long as you have statistical significance, neither score is more than the other

20

## practical significance

###
difference meaningful in real-world context

- one leads to more improvement over other

21

## effect sizes

### the extent to which population means differ and distributions overlap

22

## large effect size

### little overlap with vastly different means

23

## small effect size

### a lot of overlap with different but close means

24

## Cohen's D

###
measure of effect size

- mean and how spread out the distribution is (SD/SE)

- allows us to examine practical significance (how different the groups are) and compare studies

25

## effect size cut-offs

###
small = .2

medium = .5

large = .8

26

## statistical power

###
the probability that a study will yield a statistically significant result if the research hypothesis is really true

- opposite is beta/type II error

27

## what affects power

###
- effect size

- sample size

- significance level (alpha)

- one vs. two tailed tests

- statistical test

28