Scribd
Upload
8 views

Lesson 32

The document discusses the results of a two-way ANOVA examining the effects of gender and treatment on depression scores. It finds a significant interaction effect between gender and treatment, with treatment II increasing depression scores more for females than males, while the main effect for gender is non-significant. ANOVA is highlighted as a robust test that allows for multiple comparisons without inflating the Type I error rate.

Uploaded by

misslaghari2
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
8 views

Lesson 32

The document discusses the results of a two-way ANOVA examining the effects of gender and treatment on depression scores. It finds a significant interaction effect between gender and treatment, with treatment II increasing depression scores more for females than males, while the main effect for gender is non-significant. ANOVA is highlighted as a robust test that allows for multiple comparisons without inflating the Type I error rate.

Uploaded by

misslaghari2
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Lesson 32

Two-Way ANOVA-II
Interpreting the Plots

The output gives a plot of the depression scores for males and females, across the treatments. This
plot is very useful for allowing us to visually inspect the relationship among variables. Although
presented last, the plots are often useful to inspect first to help us better understand the impact of
two independent variables. The main effect for gender is non-significant and the plot indicates that
the mean score of depression for both males and females is same. The main effect for treatment is
based on the tendency for the depression scores to increase from treatment I to treatment II as
shown in graph.

The plots show the interdependence of both factors. Treatment I is likely to produce a decrease in
depression scores for females than males. On the other hand, treatment II influences females more
than males. That is, females are more likely to have increase in depression scores after treatment
II.

©copyright Virtual University of Pakistan


ANOVA IS ROBUST TEST

Each time we do a hypothesis test, we select an alpha level that determines the risk of a Type I
error. With alpha.05, for example, there is a 5%, or a 1-in-20, risk of a Type I error.

Often a single experiment requires several hypothesis tests to evaluate all the mean differences.
However, each test has a risk of a Type I error, and the more tests we do, the more risk there is.

For this reason, researchers often make a distinction between the test wise alpha level and the
experiment wise alpha level. The test wise alpha level is simply the alpha level that is selected
select for each individual hypothesis test. The experiment wise alpha level is the total probability
of a Type I error accumulated from all of the separate tests in the experiment.

For example, an experiment involving three treatments would require three separate t tests to
compare all of the mean differences:

 Test 1 compares treatment I with treatment II.


 Test 2 compares treatment I with treatment III.
 Test 3 compares treatment II with treatment III.
The advantage of ANOVA is that it performs all three comparisons simultaneously in one
hypothesis test. Thus, no matter how many different means are being compared, ANOVA uses
one test with one alpha level to evaluate the mean differences, and thereby avoids the problem of
an inflated experiment wise alpha level. ANOVA or F-test is also robust to moderate departures
from normality when sample sizes are reasonably large and are equal.

The ANOVA, therefore, can tolerate data that is non-normal (skewed or kurtotic distributions)
with only a small effect on the Type I error rate.

Reporting Results

A two-way between-groups analysis of variance was conducted to explore the impact of gender
and treatment on depression. The interaction effect between gender and treatment was statistically
significant, F (1, 12) = 7.53, p < .05. There was a statistically significant main effect for treatment
(Treatment I, M=4, SD=3.42; Treatment II, M=8, SD=3.46), F (1, 12) = 7.53, p < .05, 𝜂2 = 0.37.

©copyright Virtual University of Pakistan


The main effect for gender was not significant (Male, M=6, SD=2.61; Female, M=6, SD=5.09), F
(1, 12) =.00, p =1.00.

©copyright Virtual University of Pakistan

You might also like