General linear model and ANOVA Flashcards
what is the general linear model?
- Framework for comparing how several variables affect different continuous variables
what does general linear model include?
- correlations
- t- tests
- regression
- ANOVA/ ANCOVA
why does general linear model work?
- because it assumes that the residuals will be normally distributed and that different statistical tests combine linearly
why is it known as linear?
- because different stats tests add together in a simple way
what are the two main purposes of the general linear model?
- strengths and directions of relationships and differences
- strengths of interactions and manipulations
what do all general linear model tests analyse?
- all analyse variance
what differences do ANOVAs find?
- categorical differences between different conditions or different groups in a study
what are the two special kinds of linear regression?
- ANOVA
- T test
what is regression?
- statistical test to analyse relationship between a dependent variable (target) and one or more independent variables ( predictor)
what does ANOVA test?
- tests differences between two or more means
how is ANOVA different from a t-test?
- t- test generally used for comparing two means whereas ANOVA used for three
describe the variables of an ANOVA test
one categorical predictor (group= IV)
+
one continuous outcome (measure= DV)
what is a one way ANOVA?
- compares the means of three or more independent groups to determine if there is a statically significant difference between the means
how many groups in one- way ANOVA?
- three or more groups
how many variables, predictors and outcomes in one way ANOVA?
- one outcome variable
- one categorical predictor
- one continuous outcome
what is factorial (or multi- way) ANOVA?
- used when you’d like to understand how two or more factors affect a response variable and whether or not there is an interaction effect between the factors and the response variable
how many groups in factorial ANOVA?
- two or more groups
how many variables, predictors and outcomes in factorial ANOVA?
- two or more categorical predictor variables (sex, group)
- one continuous outcome variable(measure)
what is the grid created in factorial ANOVA? can individuals be in more than one group?
- 2 x 2
- four individual groups
- each person can only be in one
what is repeated measures ANOVA also known as?
- within- subjects
what is repeated measures ANOVA?
- compares means of three or more groups in which the same subjects show up in each group
how many groups in repeated measures ANOVA?
- one group
how many predictors and outcomes in repeated measures ANOVA?
- two or more categorical predictors
- one outcome variable (measure)
what is mixed ANOVA?
- compares mean differences between groups based on two factors : within and between subjects
how many groups in mixed ANOVA?
- two or more groups
how many predictors and outcomes in mixed ANOVA?
- one or more categorical predictors
- one outcome variable
what are the three things you should work out before deciding the appropriate test?
- between group variable
- within group variable
- levels of variable
what does the graph need to have? and why?
- needs to have error bars
- to show significance
- how big is the difference relative to the size of error
what does a small standard deviation mean?
- low spread
- data around mean
what are the four other different types of test?
- MANOVA
- ANCOVA
- logistic regression
- multivariate regression
what is MANOVA?
- used for multiple outcome variables
e.g., height and weight
what is ANCOVA?
- used for continuous predictor variable
e.g., cholesterol
what is logistic regression?
- discrete outcome variances
e.g., diabetes/ not
what is multivariate regression?
- multiple predictor and outcome variables
e.g., cholesterol, height, diabetes/ not, weight
how do you use one way ANOVA on jamovi?
- select one way ANOVA
- chose mean difference and report significance
what type of ANOVA isn’t on jamovi, what do you pick instead?
- Factorial ANOVA isn’t on jamovi
- just ‘ANOVA’
what is mixed ANOVA under on jamovi?
- under repeated measures
how do you report one way ANOVA results?
F (df1, df2), p value
how do you find df1?
number of groups - 1
how do you find df2?
number of values- number of groups
what do you run after finding a significant difference among three or more groups/ conditions? what are they?
- run post- hoc tests to see which groups were significantly different from each other
- they are t- tests but with some corrections applied to the p- values
what corrections are applied to post- hoc tests?
- corrections consider multiple comparisons
what is interaction ?
- between two variables
- differences in one variable are affected by differences in another variable
what does the interaction speciality depend on?
- your theory
- how you entered the data into ANOVA
how are interactions indicated?
- by a star *
what is sphericity?
- variances of the differences between all combinations of related groups (levels) are equal
does all data need to be corrected?
- no correction if variability is similar for the conditions
what is non- sphericity?
- variances of the differences between all combinations are not equal
what changes when you report results after correction?
- degrees of freedom changes
what is the assumptions about observations in ANOVA?
- observations are independent e.g., from different people; different times
what is the assumption about variance in ANOVA?
- variance of different groups or conditions should be equal
what is the assumption about residuals in ANOVA?
- residuals will be normally distributed
what are residuals?
- unexplained variance/ error
what is the assumption in repeated measures ANOVA?
- differences between levels of a variable should have equal vairance
if you have three or more groups/ conditions in a variable in a RM ANOVA what should you check?
- check sphericity