1-Way Anova Flashcards
Definition of Anova
Anova examines if there is a statistically significant difference in the means of two or more groups or conditions.
What are the types of designs used in an anova?
Between-subjects anova (independent) = different participants are used in each condition
within-subjects anova (related) = the same participants are used in each condition.
What is parametric?
Parametric test means it makes assumptions about the distribution of the scores. The assumptions include:
- level of measurement must be interval or ratio.
- Scores should be normally distributed
- homogeneity of variance - meaning data in each group should have the same variance.
- observations should be independent - so no collusion
What is the characteristic of anova?
Anova is fairly robust so an anova can be still used on data if it doesn’t meet all the assumptions.
advantages of anova
- anova means you don’t have to do multiple t-tests which increases the chance of type 1 error occurring which is a false positive.
- a factorial anova allows for multiple independent variables (factors) to be investigated at the same time.
What is a type 1 error?
Type 1 error is a false positive which means the null hypothesis is rejected when it is actually true.
Definition of factorial anova
A factorial anova is an anova that investigates multiple factors (independent variables) at the same time.
Definition of Factor
Factor in anova is independent variable
Definition of levels
Levels of a factor is the groups/conditions of an independent variable.
What does the p value have to be for the difference to be considered significant
p value has to be less than 0.05 to be considered significant.
What does an anova do?
anova finds out if there is a statically significant difference in the mean between two or more groups by calculating the f ratio which compares variation by comparing explained variation due to manipulation of Independent variable (ms treatment) and unexplained variation due to individual differences (ms error)
MS error
MS error is unexplained variation due to individual differences in f ratio.
MS treatment
Ms treatment in f-ratio is explained variation due to manipulation.
Why does an anova allow you to avoid doing multiple t-test?
Factorial anova allows you to investigate multiple independent variables (factors) at the same time.
How do you report degrees of freedom in anova?
You report 2 degrees of freedom in an anova. You report smaller one first.
what does ANOVA stand for?
ANOVA means analysis of variance
How does you report a 1-way anova?
a 1-way between subjects analysis of variance (ANOVA) revealed a statically significant effect of task, F (12, 36) = 0.29 p = 0.02.
How many levels can a factor have in an anova?
A factor in an anova can have 2 or more levels.
what is equation for f ratio?
f = ms treatment/ms error
ms treatment = variation due to manipulation of independent variable.
ms error = variation due to individual differences.
Disadvantage of ANOVA
- F ratio tells you whether there is a difference but it does not tell you whether the difference is. so you have to do post hoc test to find out where the difference lies.
- anova cannot be done on all data as anova is parametric so data has to meet parametric assumptions but anova is fairy robust so you can have minor departures from the assumptions.
How to perform 1-way between subjects anova in SPPS?
To perform a 1 way anova in spss you press analyse, compare means and one way anova.
When do you report levene’s test (homogeneity of variance)?
You report levene’s test if it is significant so less than 0.05
How to perform one way within subjects anova on SPSS?
To perform a 1-way within subjects anova you press analyse, general linear model and repeated measures
