PSY3213

TEST 3

Chapter 10

Analyzing Experimental Data

I. Types of Statistics

a. Descriptive statistics: summarize and describe the behavior of participants in a study. Example: mean and standard deviation

b. Inferential statistics: help researchers infer how the results from their sample might generalize to the population. Example: t-tests and F-tests

1. Help researchers draw conclusions about their data

2. Tell us if the differences between groups is larger than would be expected due to chance or error variance

II. Types of Hypotheses

a. Null Hypothesis: says there will be no differences between groups: the independent variable had no effect on the dependent variable

b. Experimental Hypothesis: says there will be differences between groups: the independent variable had an effect on the dependent variable

c. The main purpose of our experiment is to see whether or not we reject the null hypothesis:

1. If we reject the null, we’re saying the independent variable had a significant effect on the dependent variable

2. If we fail to reject the null, we’re saying that the independent variable had no significant effect on the dependent variable

d. Inferential statistics tell us whether or not to reject the null

III. Types of Errors

a. Type I error: we decide to reject the null when we shouldn’t have done so

1. The null is true

2. Changes were due to chance

3. Finding something that isn’t there

b. Type II error: we fail to reject the null when we should have rejected it

1. The null is false

2. Missing something that is there

c. Alpha level: the probability of making a Type I error

1. We decide to reject the null by using alpha level = .05

2. There are 5 chances out of 100 of having a Type I error

d. Beta level: the probability of making a Type II error

1. Power = 1-beta

2. We want a study with high power because that means we aren’t likely to make Type II errors

IV. Power

a. Power: the probability that you’ll reject the null hypothesis when the null is false

b. Ways to increase power:

1. Have a lot of participants

2. Use formulas that tell you how to get good amounts of power

c. Effect size: how much of the variability in the dependent variable is due to the independent variable

1. Ranges from .00 to 1.00

2. Effect size of .30 = 30% of the variability in the dependent variable is a direct result of our manipulation of the independent variable

V. One-tailed and Two-tailed Tests

a. One-tailed test: the experimental hypothesis specifies a direction

1. Example: higher or lower than…

b. Two-tailed test: the experimental hypothesis does not specify a direction

1. Example: different from…

VI. T-tests

a. T-test: inferential statistic that tells if there is a difference between the means of two groups

b. T-tests are used in a one way design with two groups

c. Computed t-value: represents how different the means of the 2 groups are

1. Accounts for error

2. Given to researcher by SPSS

d. Critical t-value: looked up in chart by using degrees of freedom (= number of participants - 2), alpha level (= .05), and knowing if it’s one or two-tailed

e. Comparing t-computed and t-critical:

1. If t-comp is greater than t-crit: we reject the null: there are group differences

2. If t-comp is lower than t-crit: we fail to reject the null: no group differences

f. If there are fewer participants it is harder to reject the null because it is difficult to find differences and be sure that they don’t occur by chance

g. Different types of two groups designs that require t-tests:

1. Two independent groups

2. Two matched groups

3. Repeated measures / within subjects

h. Paired t-test: used in matched-subjects or within-subjects designs

1. It is more powerful because it reduces error variance that comes from individual differences



	PSY3213 PSY3213 TEST 3 Chapter 10 Analyzing Experimental Data I. Types of Statistics a. Descriptive statistics: summarize and describe the behavior of participants in a study. Example: mean and standard deviation b. Inferential statistics: help researchers infer how the results from their sample might generalize to the population. Example: t-tests and F-tests 1. Help researchers draw conclusions about their data 2. Tell us if the differences between groups is larger than would be expected due to chance or error variance II. Types of Hypotheses a. Null Hypothesis: says there will be no differences between groups: the independent variable had no effect on the dependent variable b. Experimental Hypothesis: says there will be differences between groups: the independent variable had an effect on the dependent variable c. The main purpose of our experiment is to see whether or not we reject the null hypothesis: 1. If we reject the null, we’re saying the independent variable had a significant effect on the dependent variable 2. If we fail to reject the null, we’re saying that the independent variable had no significant effect on the dependent variable d. Inferential statistics tell us whether or not to reject the null III. Types of Errors a. Type I error: we decide to reject the null when we shouldn’t have done so 1. The null is true 2. Changes were due to chance 3. Finding something that isn’t there b. Type II error: we fail to reject the null when we should have rejected it 1. The null is false 2. Missing something that is there c. Alpha level: the probability of making a Type I error 1. We decide to reject the null by using alpha level = .05 2. There are 5 chances out of 100 of having a Type I error d. Beta level: the probability of making a Type II error 1. Power = 1-beta 2. We want a study with high power because that means we aren’t likely to make Type II errors IV. Power a. Power: the probability that you’ll reject the null hypothesis when the null is false b. Ways to increase power: 1. Have a lot of participants 2. Use formulas that tell you how to get good amounts of power c. Effect size: how much of the variability in the dependent variable is due to the independent variable 1. Ranges from .00 to 1.00 2. Effect size of .30 = 30% of the variability in the dependent variable is a direct result of our manipulation of the independent variable V. One-tailed and Two-tailed Tests a. One-tailed test: the experimental hypothesis specifies a direction 1. Example: higher or lower than… b. Two-tailed test: the experimental hypothesis does not specify a direction 1. Example: different from… VI. T-tests a. T-test: inferential statistic that tells if there is a difference between the means of two groups b. T-tests are used in a one way design with two groups c. Computed t-value: represents how different the means of the 2 groups are 1. Accounts for error 2. Given to researcher by SPSS d. Critical t-value: looked up in chart by using degrees of freedom (= number of participants - 2), alpha level (= .05), and knowing if it’s one or two-tailed e. Comparing t-computed and t-critical: 1. If t-comp is greater than t-crit: we reject the null: there are group differences 2. If t-comp is lower than t-crit: we fail to reject the null: no group differences f. If there are fewer participants it is harder to reject the null because it is difficult to find differences and be sure that they don’t occur by chance g. Different types of two groups designs that require t-tests: 1. Two independent groups 2. Two matched groups 3. Repeated measures / within subjects h. Paired t-test: used in matched-subjects or within-subjects designs 1. It is more powerful because it reduces error variance that comes from individual differences