Scribd
Upload
11 views30 pages

Factorial ANOVA

Uploaded by

zehra.ozen
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
11 views30 pages

Factorial ANOVA

Uploaded by

zehra.ozen
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 30

Factorial ANOVA

(Independent Measures)
Bahar Tanyaş
29 April 2020
• In most research situations, the goal is to
examine the relationship between two variables.
• A typical experiment focuses on one independent
variable and one dependent variable.
• In real life, however, variables rarely exist in
isolation. Behaviour usually is influenced by a a
variety of different variables acting and
interacting simultaneously.
• Factorial ANOVA allows us to explore the
effect of more than one independent variable
(two or more) on one dependent variable.
• When a research study involves more than
one factor, it is called factorial design.
Factorial
ANOVA

Two or more
independent variables
In this lecture, we will review the simplest version of a
factorial design. Thus, we will use Two-Factor ANOVA

IV
DV
IV
TWO –FACTOR ANOVA
IVs: AROUSAL LEVEL & EXAM DIFFICULTY /DV: NUMBER OF MISTAKES

FACTOR B: AROUSAL LEVEL


FACTOR A: EXAM DIFFICULTY

LOW MEDIUM HIGH


DIFFICULT

3 1 6

2 1 4
EASY
MAIN EFFECTS AND INTERACTIONS
• The mean differences among the levels of one factor are
referred as the main effect.

• Effect of arousal level on the number of mistakes is a main effect


– Factor: arousal level
– Levels of the factor : low, medium and high

Do we perform differently under different levels of arousal?

Do we make more mistakes in high arousal level in comparison to


medium and low arousal levels.
• Effect of exam difficulty on the number of
mistakes we made.
– Factor: Exam Difficulty
– Level of Factor: Easy vs. Difficult

Is there any difference in our performance when the


exam difficulty changes?
Do we perform better in easy exams than we do in
difficult exams.
• An interaction between two factors occurs whenever the
mean differences between individual conditions are
different from what would be predicted from the overall
main effects of the factors.
• For example, we know that we perform better (do less
mistakes) when the exam is easy . Assume that this
difference in our performance in easy and difficult exams
disappears when we have medium level of arousal.
• Thus effect of one factor (exam difficulty)has changed
depending on the levels of the other factor (arousal level)
• Examples for intercation effect:
• For example, there is a considerable effect of
anti-depressant medication on depression.
But this effect depends on the level of
depression. If the client’s level of depression is
mild, the effect is relatively low.
• Smoking is very dangerous for our health. Its
detrimental effects rises when you consume
alcohol while you smoke.
Interaction- Examples
• People with schizophrenia benefit a lot from
psychotherapy; these benefits are even more if
they use their medications regularly. If they do not
comply with the medical treatment, the effect of
psychotherapy remains limited though overall
psychotherapy has an effect on psyhotic
symptoms.
• If you use some medications with alcohol, they
interact and you get drunk with e.g. One glass of
wine
Conditions # of relapses in 5 years period (e.g.
Significant increase in auditory
hallusinations )

Use of medication only 4


Psyhotherapy only 6
Medication+ Psychotherapy 2
Main Effects and Interactions
• Data showing a main effect for Exam Difficulty no main
effect of Arousal Level and no interaction between
factors

Low Arousal High Arousal


Easy Task 10 10 Mean # of

10-point difference
mistakes for
easy task: 10

Difficult 20 20 Mean # of
mistakes for
difficult task :
20

No difference
Main Effects and Interactions
• Data showing a main effect for both Exam Difficulty and
Arousal Level but no interaction between factors

Low Arousal High Arousal


Easy Task 10 30 Mean # of

10-point difference
mistakes for
easy task: 20

Difficult 20 40 Mean # of
mistakes for
difficult task :
30

20-point difference
Main Effects and Interactions
• Data showing no main effect for either factor but an
interaction between factors

Low Arousal High Arousal


Easy Task 10 20 Mean # of

No difference
mistakes for
easy task: 15

Difficult 20 10 Mean # of
mistakes for
difficult task :
15

No difference
Psychotherapy No psychotherapy
Medication 10 20
No Medication 30 40

• Describe the main effect for medication


• Descirbe the main effect for psychotherapy
• Does there appear to be an interaction between the two factors?
ANSWERS
• There is a main effect for medication:
• Medication group average is 20 points lower than no medication group
• There is a main effect for psychotherapy:
• Psychotherapy group average is 10 points lower than no psyhotherapy
group
• There is not interaction; 20 point difference between
medication and no medication group is constant; it does not
change depending on the levels of psyhotherapy.
Graph Example: No Interaction

Parallel lines indicate that there is no interaction


Grap: Interaction
HYPOTHESES
• Hypotheses for Main Effects
– Ho= There is no significant effect of Factor A on
dependent variable
– H1= There is a significant effect of Factor A on
dependent variable

– Ho= There is no significant effect of Factor B on


dependent variable
– H1= There is a significant effect of Factor A on
dependent variable
HYPOTHESES
• Hypotheses for Interaction Effect
– Ho= There is no significant interaction between
Factor A and Factor B
– H1= There is a significant interaction between
Factor A and Factor B
Assumptions
• Dependent variable has to be numeric (ratio or interval) and the
independent variables has to be categorical (nominal or ordinal)
• Dependent variable should be normally distributed.
• Variances across samples have to be homogenous
• observations to be mutually independent from each other (e.g., no
repeated measurements)
• Since the factorial ANOVA includes two or more independent
variables it is important that the factorial ANOVA model contains
little or no multicollinearity.

Multicollinearity occurs when the independent variables are


intercorrelated and not independent from each other.
Which ANOVA ?
• What is the difference in average pain levels among post-surgical patients
given three different painkillers?

• We conduct a study examining how one’s anxiety level prior to giving a


speech is affected by two factors: their self-confidence prior to the speech,
and the size of the group to which they’re speaking. Subjects are
classified as having either high or low self-confidence, and the group is
classified as either small, medium, or large

• A researcher is interested in the effects of time of day on performance of a


memory recall task. He has 8 subjects perform the task at 4 times during
the day: 8 am, noon, 4 pm, and 8 pm; each subject performs the task at
each of the four times
ANSWERS
• Independent Measures ANOVA
• Factorial (Two-Factor) ANOVA
• Repeated Measures ANOVA
SPSS OUTPUT
IV: Gender DV:Anxiety
IV: Education DV:Anxiety
IVs: Gender & Education DV: Anxiety
Reporting Results in APA
• There was a significant effect of gender on anxiety, F(1,167)
=42.281, p< 0.01. Female participants are significantly more
anxious than male participants. There was a significant effect
of education on anxiety, F(2,167) =240.74, p< 0.01. As the
level of education increases the level of anxiety also increases.
The interaction between gender and education was
significant, F(2, 167) =34.24, p< 0.01. The significant
difference between male and female participants disappear
when participants had university or upper level education.

You might also like