Class Notes (1,100,000)

CA (630,000)

McGill (30,000)

PSYC (4,000)

PSYC 305 (50)

Heungsun Hwang (50)

Lecture 12

# PSYC 305 Lecture Notes - Lecture 12: Repeated Measures Design, Migraine, Analysis Of Variance

This

**preview**shows pages 1-3. to view the full**9 pages of the document.**February 21st

Repeated measures design

We often use an experimental design in which measurements on a single DV are

repeated a number of times within the same subject.

Such designs in which subjects are crossed with at least one experimental factor are

called repeated-measures designs.

The simplest experiment of this kind may be a before- and after-treatment

design (two conditions)

One-way repeated measures design

One-way means the number of independent variables is one

N subjects are measured on a single DV under K conditions or levels of a

single IV or factor.

Subjects are repeatedly measured across all levels of independent

variables

One-way ANOVA example

IV: there is one independent variable

This factor has multiple levels

k is the amount of levels

Example: class row—there are three levels of class row

This is the same independent factor, but separated into several groups

This affects our continuous dependent variable

DV: for example, this could be exam scores

We assume that all groups have different subjects in regular one-way ANOVA

k different groups have different subjects

But in one-way repeated measures ANOVA, we have the same participants go

to all of the levels

k different groups always have the same subjects

This is the main difference between one-way ANOVA and one-way

repeated measures ANOVA

Example 1 (simplest)

We have two subjects and both have an acne problem, and we observe them before

and after using an acne product

Subject 1: before—has 10 pimples after—5 pimples on the left

Subject 2: before—has 10 pimples after—5 pimples on the right

So, we can tell that both subjects reacted to the acne product positively, but

somewhat differently

Example 2

National college entrance exam application essay test interview

You take the college entrance exam, and if you do well enough on it, you fill in your

university application. If you do well enough in the application, you do the essay

test. If this goes well enough, you do the interview.

You are the participant, and you, the same participant, go through all the tests

find more resources at oneclass.com

find more resources at oneclass.com

Only pages 1-3 are available for preview. Some parts have been intentionally blurred.

Four separate professors interview you, one after the other

The independent variable is the interview—there are four levels (four

professors)

The dependent score would be the score you receive after your interview

The subject also changes throughout all levels

The subject could be getting more and more comfortable as the interview

goes on, or there could be other factors (they could be getting more and more

tired)

So, when the mean changes across professors, you cannot be certain whether

the professor or the subject initiated this change

Example: one-way repeated measures design

Four migraine sufferers were asked to record the duration of their migraine

headaches. After the baseline recording, during which no training was given, each

subject had a six-week period of relaxation training. We would like to know if the

relaxation therapy can help to relieve the migraine. The dependent variable is the

duration (hours/week) of headaches recorded at the third and sixth week of the

therapy.

We may be interested to find out if the relaxation training is effective in reducing the

duration of migraine headaches.

DV = Duration of headache

IV = Migraine therapy

The four same patients are repeatedly measured across three levels of

migraine therapy

The scores are the dependent variable (duration of headache)

There is only one independent variable, which is migraine therapy

Group means plot

find more resources at oneclass.com

find more resources at oneclass.com

Only pages 1-3 are available for preview. Some parts have been intentionally blurred.

Subject scores plot

The subjects may change over the different conditions

This is one source of change we may want to look at

Superimposed

Possible questions

Overall, are there significant differences in the mean scores of DV across groups?

Between-group effect of treatment

Overall, are there significant differences across subjects?

Subject-level variability (between-subject effect)

Technically, we can test both of these things, but in reality we dont care very much

about subject-level variability

One-way repeated measures ANOVA

It can examine the effect of treatment (IV) (between-group effects)

H0T: μ1 = μ2 … = μk

It can also test the effect of subjects (between-subject effect)

As the levels of the subject factor are individual subjects, this effect

represents the variance of subjects.

If variance is large, subjects are more different from each other

H0s: Vs = 0

Meaning that all subjects are the same

find more resources at oneclass.com

find more resources at oneclass.com

###### You're Reading a Preview

Unlock to view full version