Lecture 10: Experimental Design and Analysis I

Question 1

Confound

Answer 1

An uncontrolled extraneous variable or flaw in the experiment. If a study is confounded, we do not know if the changes in the DV were caused by the IV

Question 2

Internal Validity

Answer 2

The extent to which results can be attributed to manipulation of the IV rather than some confounding variable.

Question 3

3 major confounding variables that are threats to internal validity

Answer 3

Nonequivalent control group

History effect

Maturation effect

Question 4

Nonequivalent control group description and how to overcome

Answer 4

Problems in subject selection or assignment may lead to important differences between the subjects assigned to the experimental and control groups.

Overcoming: Use random sampling and random assignment

Question 5

History effect description and how to overcome

Answer 5

Changes in the DV may be due to outside events that take place during the course of the study

Overcoming: Use an equivalent control group

Question 6

Maturation effect description and how to overcome

Answer 6

Changes in the DV may be due to the subjects maturing during the study

Overcoming: Use an equivalent control group

Question 7

External Validity

Answer 7

The extent the results can be generalised beyond the experiment

Question 8

2 threats to external validity

Answer 8

Generalisation to populations. Accomplished by randomly selecting subjects from population

Generalisation from laboratory settings (artificial environment)

Question 9

Two-Groups Between Subjects Designs

Answer 9

In a between-subjects design, the subjects in each group are different. Participants are assigned to different groups. There are therefore two groups to compare, either two experimental groups or an experimental group and a control group.

Question 10

Independent variable:

Answer 10

The variable the experimenter manipulates

Question 11

Dependent variable:

Answer 11

The variable the experimenter measures. The dependent variable depends on the independent variable that is manipulated.

Question 12

Experimentation involves control

Answer 12

To minimise the effect of individual differences, and to best generalise our sample to the population, we want to randomly allocate participants to conditions

Question 13

Independent measures t-tests

Answer 13

A parametric statistical test that compares the means of two different samples of participants

Question 14

Independent measures t-tests concepts

Answer 14

Categorical IV
(One IV with two levels, participants are only in one level)

Continuous DV

Question 15

Independent Measures t-test: Example

Answer 15

You’ll have to look at the slides for examples, diagrams and explanations of the examples.

Question 16

Independent Measures t-tests: Effect Size

Answer 16

Effect size is a calculation of the relationship between
the IV and the DV, independent of sample size. This is good to know, as increasing sample size could eventually just “push” a obtained statistic into the critical region

Question 17

Independent Measures t-tests confidence intervals

Answer 17

Confidence interval represents 95% certainty the mean difference will fall between those values.

Question 18

Assumptions and appropriate use of Independent Measures t-test

Answer 18

Normality of distributions

Homogeneity of variance (The variance of each distribution is similar)