MANOVA

Presented by/
‫االء سمير السيد‬
‫ايه على سعيد‬
‫كريمة صالح محمد‬
‫حاتم يوسف الزم‬
‫محمود صالح أبورية‬
‫محمد اقبال‬

Presented to/ Dr. Magda Gaied

November 2021

Doctoral degree 3rd semester

1
Definition:

Multivariate analysis of variance (MANOVA) is a statistical analysis used

when a researcher wants to examine the effects of one or more independent
variables (IVs) on multiple dependent variables (DVs). This method is an
extension of the analysis of variance (ANOVA) model and is the most commonly
used multivariate analysis in the social sciences.

For example: you might want to test the hypothesis that three different
methods of teaching writing result in significant differences in ratings of student
creativity, student acquisition of grammar, and assessments of writing quality by
an independent panel of judges.

MANOVA tests, whether they are statistically significant or not, produce

differences among levels of the IVs for multiple DVs. MANOVA tests belong to a
larger family of statistical techniques known as the general linear model, which
include analyses such as ANOVA, multiple types of regression, and repeated-
measures designs.

MANOVA is an inferential statistical analysis. Communication researchers

use this analysis to deduce a causal relationship between IVs and DVs.

What is the difference between ANOVA and MANOVA?

ANOVA stands for “Analysis of Variance” is a statistical tool used to detect

differences between experimental group means. In it we examine for statistical

2
differences on one continuous dependent variable by an independent grouping
variable. Can be applied to independent groups or repeated measures designs.

The MANOVA extends this analysis by taking into account multiple

continuous dependent variables, and bundles them together into a weighted linear
combination or composite variable. The MANOVA will compare whether or not
the newly created combination differs by the different groups, or levels, of the
independent variable. In this way, the MANOVA essentially tests whether or not
the independent grouping variable simultaneously explains a statistically
significant amount of variance in the dependent variable.

Why to use MANOVA?

MANOVA should be done instead of a series of one-at-atime ANOVAs for

two main reasons:

1. To reduce the experiment-wise level of Type I error (8 F tests at .05 each

means the experiment-wise probability of making a Type I error (rejecting
the null hypothesis when it is in fact true) is 40%! The so-called overall test
or omnibus test protects against this inflated error probability only when the
null hypothesis is true.
2. None of the individual ANOVAs may produce a significant main effect on
the DV, but in combination they might, which suggests that the variables are
more meaningful taken together than considered separately.
3. MANOVA takes into account the inter-correlations among the dependent
variables

3
Uses of MANOVA

This is used in studies where more than one factors affect the dependent
variable. When the various factors affecting the dependent variable, including all
their combinations at different levels, are studied and tested, these are called
Factorial experiments in MANOVA. No matter which type of MANOVA is being
used, the dependent variables are always linearly combined.

Benefits of MANOVA

Using MANOVA has a lot of advantages. It is a very robust analysis method

that has the ability to weed out a wide degree of effects across related variables, as
opposed to other methods like ANOVA (analysis of variables), which has a more
limited scope.

1. Saves time: Since MANOVA techniques can test multiple variables at once,

it’s a time saver compared to other forms of analysis, which test only two
variables, then another two variables, and so forth.
2. Powerful statistics: MANOVA is a strong statistical method that can
identify the smaller impact of correlated dependent variables, which typical
ANOVA techniques can’t discover.
3. Evaluate patterns: MANOVA can evaluate patterns among multiple
dependent variables, uncovering how the elements within a model can
impact the connection between dependent variables rather than influencing
an individual dependent variable.
4. Reduced likelihood of errors: When you have multiple dependent
variables, a series of ANOVA analysis can increase the likelihood of

4
 rejecting a null hypothesis. But in a single MANOVA analysis, error rates
are equal to the significance levels.

Assumptions of MANOVA

1- Two or more dependent variables should be measured at the interval or ratio

level (i.e., they are continuous). Examples of variables that meet this criterion
include revision time (measured in hours), intelligence (measured using IQ score),
exam performance (measured from 0 to 100), weight (measured in kg), and so
forth.

2- Your independent variable should consist of two or more categorical,

independent groups. Example independent variables that meet this criterion include
ethnicity (e.g., 3 groups: Caucasian, African American and Hispanic), physical
activity level (e.g., 4 groups: sedentary, low, moderate and high), profession (e.g.,
5 groups: surgeon, doctor, nurse, dentist, therapist), and so forth.

3- Independence of observations, which means that there is no relationship

between the observations in each group or between the groups themselves. For
example, there must be different participants in each group with no participant
being in more than one group.

4- You should have an adequate sample size. Although the larger your sample
size, the better; for MANOVA, you need to have more cases in each group than the
number of dependent variables you are analyzing.

5- There are no univariate or multivariate outliers. First, there can be no

(univariate) outliers in each group of the independent variable for any of the
dependent variables. This is a similar assumption to the one-way ANOVA, but for
each dependent variable that you have in your MANOVA analysis. Univariate
5
outliers are often just called outliers and are the same type of outliers you will have
come across if you have conducted t-tests or ANOVAs. We refer to them as
univariate in this guide to distinguish them from multivariate outliers. Multivariate
outliers are cases which have an unusual combination of scores on the dependent
variables.

6- There is multivariate normality. Unfortunately, multivariate normality is a

particularly tricky assumption to test for and cannot be directly tested in SPSS
Statistics. Instead, normality of each of the dependent variables for each of the
groups of the independent variable is often used in its place as a best 'guess' as to
whether there is multivariate normality. You can test for this using the Shapiro-
Wilk test of normality, which is easily tested for using SPSS Statistics.

7- There is a linear relationship between each pair of dependent variables for

each group of the independent variable. If the variables are not linearly related, the
power of the test is reduced.

8- There is homogeneity of variance-covariance matrices. You can test this

assumption in SPSS Statistics using Box's M test of equality of covariance. If your
data fails this assumption, you may also need to use SPSS Statistics to carry out
Levene's test of homogeneity of variance to determine where the problem may lie.

9- There is no multicollinearity. Ideally, you want your dependent variables to

be moderately correlated with each other. If the correlations are low, you might be
better off running separate one-way ANOVAs, and if the correlation(s) are too
high (greater than 0.9), you could have multicollinearity. This is problematic for
MANOVA and needs to be screened out.

6
 Running a MANOVA in SPSS

Go to Analyze/General Linear Model/ Multivariate

Multivariate dialogue box will appearMove independent variables (IV) into
the FixedFactor(s) box.
Move dependent variables (DV) into the DependentVariables box.
Then click on Plots button
Multivariate:ProfilePlots dialogue box will appear, Move independent
variables (IV) into the HorizontalAxis box.
Click on Add button to add the IV in the Plots box
Click continue button to return to Multivariate dialogue box
Click on the PostHoc button
Multivariate: Post Hoc Multiple Comparison for Observed Means
dialogue box will appear
Move IV into Post Hoc Test for: box and check tukey in the Equal
Variances Assumedarea
Click on Continue button to return to Multivariate dialogue box
Click on the EM Means button.
Multivariate: Estimated Marginal Means dialogue box will appear
Move IV into display means for box and check Compare main effects
Click on Continue button to return to Multivariate dialogue box
Then click on Optionsbotton.
Multivariate: Option Dialogue Boxwill appear

7
 In the display table, check Descriptive statistics, Estimates of effect size
observed power and Homogeneity tests.
Set the confidence level to .05
Click on Continue button to return to Multivariate dialogue box then ok.

Interpretation of SPSS MANOVA output

The steps for interpreting the SPSS output for MANOVA:

steps interpretations
1. Look in the Box's Test of Equality If it is LESS THAN .05, then
of Covariance Matrices, in the Sig. researchers have violated the
row. This is the p-value that is assumption of homogeneity of
interpreted. covariance and should not interpret the
outputs further.

If it is MORE THAN .05, then

researchers can continue with the
analysis and have met the assumption
of homogeneity of covariance
2. Look in the Levene's Test of If it is LESS THAN .05, then
Equality of Error Variances table, researchers have violated the
under the Sig. column, for each assumption and should not continue
outcome variable that has a p-value with the analysis.
testing the assumption of homogeneity
of variance. If it is MORE THAN .05, then
researchers can continue with the
8
 analysis and interpretation.
3. Look in the Multivariate Tests If it is LESS THAN .05, then
table, under the Sig. column, for the researchers have evidence of a
row that has the categorical predictor significant main effect.
variable's name and Pillai's Trace.
This is the p-value that is interpreted. If it is MORE THAN .05, then
researchers do not have evidence of a
significant effect. No further
interpretation is needed. Report the p-
value.
4. Look in the Tests of Between- If a p-value for one of the outcome
Subjects Effects table, under the Sig. variables is LESS THAN .05, then
column, for the row denoting the there was a significant main effect
categorical predictor variable. These among the independent groups or
are the p-values that are interpreted for levels of that outcome. If researchers
each individual outcome variable. find this significant main effect, they
These should be interpreted if there need to further run post hoc analyses to
was a significant main effect in the explain the effect.
Multivariate Tests table.
If a p-value for one of the outcome
variables is MORE THAN .05, then
there was NOT significant main effect
among the independent groups or
levels of that outcom
5. Scroll down to the Estimated These are the means and standard
Marginal Means section of the output. errors of the outcome for each group or
Look in the Estimates table. level of the categorical variable.

9
6. Look in the Pairwise Comparisons If a p-value is LESS THAN .05, then
table, under the Sig. column. These are there was a significant difference
the post hoc p-values that are between the independent groups or
interpreted.  levels of the categorical predictor
variable.

If a p-value is MORE THAN .05, then

there was NOT a significant difference
between the independent groups or
levels of the categorical predictor
variable.

Example

The pupils at a high school come from three different primary schools. The head
teacher wanted to know whether there were academic differences between the
pupils from the three different primary schools. As such, she randomly selected 20
pupils from School A, 20 pupils from School B and 20 pupils from School C, and
measured their academic performance as assessed by the marks they received for
their end-of-year English and Maths exams. Therefore, the two dependent
variables were "English score" and "Maths score", whilst the independent variable
was "School", which consisted of three categories: "School A", "School B" and
"School C".

Descriptive Statistics

10
The first important one is the Descriptive Statistics table shown below. This table
is very useful as it provides the mean and standard deviation for the two different
dependent variables, which have been split by the independent variable. In

addition, the table provides "Total" rows, which allows means and standard
deviations for groups only split by the dependent variable to be known.

Multivariate Tests

The Multivariate Tests table is where we find the actual result of the one-way
MANOVA. You need to look at the second Effect, labelled "School", and the
Wilks' Lambda row (highlighted in red). To determine whether the one-way
MANOVA was statistically significant you need to look at the "Sig." column. We
can see from the table that we have a "Sig." value of .000, which means p < .0005.
Therefore, we can conclude that this school's pupils academic performance was
significantly dependent on which prior school they had attended (p < .0005).

11
Reporting the previous results would be

There was a statistically significant difference in academic performance based on a

pupil's prior school , F (4, 112) = 13.74, p < .0005; Wilk's Λ = 0.450, partial η2 = .
33.

References

1- https://www.researchoptimus.com/article/what-is-manova.php

12
2- https://www.researchgate.net/post/What-is-the-difference-between-
ANOVA-MANOVA
3- https://methods.sagepub.com/reference/the-sage-encyclopedia-of-
communication-research-methods/i9318.xml
4- https://statistics.laerd.com/spss-tutorials/one-way-manova-using-spss-
statistics.php
5- https://www.statisticssolutions.com/free-resources/directory-of-statistical-
analyses/manova/

13