Psychology Research Methods: Design, Hypotheses, and Statistical Tests

Between subjects design

Each subject receives only one level of the independent variable.

Random assignment

Each member of the research study has an equal chance of being placed into any of the treatment groups.

Variance due to individual differences

A large source of error variance in a study.

Central Limit Theorem

You need a minimum of 30 subjects in each treatment group for randomization to work.

Hypothesis test

A test of the null statistical hypothesis, not the research hypothesis.

Significance Level

The highest value that alpha can be set at is .05 (the standard significance level).

Rejecting the null hypothesis ($\text{H}_0$)

You can conclude that the results support the research hypothesis.

t-test of mean differences

You want to reject $\text{H}_0$ when you conduct this test.

Pretest - posttest design

You want to reject the null hypothesis when comparing the posttest mean scores.

Accepting the null hypothesis in pretest

You want to accept the null hypothesis when comparing the pretest mean scores to mean the groups are equivalent before the independent variable is given.

Type I Error

When you reject the null hypothesis ($\text{H}_0$) when the null hypothesis is true.

Type II Error

When you accept the null hypothesis ($\text{H}_0$) when the null hypothesis is false.

Power

When you reject the null hypothesis ($\text{H}_0$) when the null hypothesis is false (correctly rejecting a false $\text{H}_0$).

Within groups variance

Is due to error/chance, not the independent variable.

Homogeneity of variances

Assumes that the sample variances are equal.

Test statistic

A ratio of variances

Levene's test

You want to accept the null hypothesis and conclude that the variances are homogeneous (equal).

Effect size

The calculation needed to determine how much of the total variance in your study was due to the Independent Variable (IV).