LECTURE 11: MANOVA Questions and
Answers Grade A+ 2023
What, When, And Why do we use a MANOVA? What are some examples?
What is one of the main differences between ANOVA and MANOVA?? - -•To test for differences
between groups
•When we have several dependent variables (outcomes)
•Tests for 'multivariate' effect
•Common in clinical research
•Example:
•assess gender (male/female) differences in wellbeing:
•Life
•Job
•Relationship
Mental, etc
That a MANOVA can handle more than one DV (multivariate), whereas ANOVAS (univariate
tests) can only do one DV
-Example research questions (help) - -- Does a program designed to improve students' study
habits actually work? (this would be the IV)
- the multiple DV's of interest could be:
- attitudes towards study
- time spent studying
- grades
- study satisfaction
- study engagement, etc.
- Do everyday orientations towards people with mental illness differ between males and
females?
- the multiple DV's of interest could be:
- attitudes towards mental illness
- activities towards helping people with mental illness
- feelings towards people with mental illness
The independent variable is the cause. Its value is independent of other variables in your study.
The dependent variable is the effect. Its value depends on changes in the independent variable
-What, When, And Why can a MANOVA tell us? and what can it do for us?
, What do multivariate tests control for and why? How would we have to control for this?
What would happen if we tried running a MANOVA through ANOVA? - -•MANOVA can tell us
whether groups differ along a combination of dimensions (various DV's)
•Potentially more powerful than ANOVA
- MANOAVA has much greater power to detect an effect if it really is there and this is because
it takes into account the relationships between DVs - in other words it maximises the ability to
find a significant between groups effect
•Potentially better than running separate ANOVAs
- Mutivariate tests control familywise error rate (Type I error
we would have to lower our alpha value.
if you run a series of ANOVA tests each of those DV's would be treated as independent of all
other effects and this would inflate familywise error. we would have to control for this by
lowering our alpha level, that's what we call the bonferoni adjustment, especially if the DV's are
related with one another and so the main reason for conducting MANOVA is to include several
outcomes in a single analysis where the DV's are not treated as independent of one another. It
gives us this level of control analysis which then in turns controls for this type 1 error, that
inflated error rate if we run those analyses separately for each DV and it is therefore considered
better then running several ANOVAS on the same set of data and so in sum when we run a
MANOVA we are protected against this familywise error
-Discriminant function
- What is MANOVAS key aim and why?
- What do we call this?
- How can we achieve this?
- What has this also been known to be called as?
- What does this do?
- What does this say about the overall effect?
What is a discriminate function? - -•MANOVA's key aim is to determine if the groups differ on a
set of DVs
•We call this the overall - multivariate effect (takes into account all the DV's in the analysis)
•To achieve this, discriminant function is formed
•This is also called a variate
•This maximises the differences between groups for the full set of DV
•It maximally discriminates between 'groups'
•The overall effect reflects the differences between group centroids
To achieve a multivariate effect a discriminat funciton is created. This function maximises the
differences between groups for the full sets of the DV's and in the process, it minimizes the
Answers Grade A+ 2023
What, When, And Why do we use a MANOVA? What are some examples?
What is one of the main differences between ANOVA and MANOVA?? - -•To test for differences
between groups
•When we have several dependent variables (outcomes)
•Tests for 'multivariate' effect
•Common in clinical research
•Example:
•assess gender (male/female) differences in wellbeing:
•Life
•Job
•Relationship
Mental, etc
That a MANOVA can handle more than one DV (multivariate), whereas ANOVAS (univariate
tests) can only do one DV
-Example research questions (help) - -- Does a program designed to improve students' study
habits actually work? (this would be the IV)
- the multiple DV's of interest could be:
- attitudes towards study
- time spent studying
- grades
- study satisfaction
- study engagement, etc.
- Do everyday orientations towards people with mental illness differ between males and
females?
- the multiple DV's of interest could be:
- attitudes towards mental illness
- activities towards helping people with mental illness
- feelings towards people with mental illness
The independent variable is the cause. Its value is independent of other variables in your study.
The dependent variable is the effect. Its value depends on changes in the independent variable
-What, When, And Why can a MANOVA tell us? and what can it do for us?
, What do multivariate tests control for and why? How would we have to control for this?
What would happen if we tried running a MANOVA through ANOVA? - -•MANOVA can tell us
whether groups differ along a combination of dimensions (various DV's)
•Potentially more powerful than ANOVA
- MANOAVA has much greater power to detect an effect if it really is there and this is because
it takes into account the relationships between DVs - in other words it maximises the ability to
find a significant between groups effect
•Potentially better than running separate ANOVAs
- Mutivariate tests control familywise error rate (Type I error
we would have to lower our alpha value.
if you run a series of ANOVA tests each of those DV's would be treated as independent of all
other effects and this would inflate familywise error. we would have to control for this by
lowering our alpha level, that's what we call the bonferoni adjustment, especially if the DV's are
related with one another and so the main reason for conducting MANOVA is to include several
outcomes in a single analysis where the DV's are not treated as independent of one another. It
gives us this level of control analysis which then in turns controls for this type 1 error, that
inflated error rate if we run those analyses separately for each DV and it is therefore considered
better then running several ANOVAS on the same set of data and so in sum when we run a
MANOVA we are protected against this familywise error
-Discriminant function
- What is MANOVAS key aim and why?
- What do we call this?
- How can we achieve this?
- What has this also been known to be called as?
- What does this do?
- What does this say about the overall effect?
What is a discriminate function? - -•MANOVA's key aim is to determine if the groups differ on a
set of DVs
•We call this the overall - multivariate effect (takes into account all the DV's in the analysis)
•To achieve this, discriminant function is formed
•This is also called a variate
•This maximises the differences between groups for the full set of DV
•It maximally discriminates between 'groups'
•The overall effect reflects the differences between group centroids
To achieve a multivariate effect a discriminat funciton is created. This function maximises the
differences between groups for the full sets of the DV's and in the process, it minimizes the
