What does regression analysis do?
It creates a parametric model of data to understand how predictor variables (Xs) explain variation in an outcome (Y)
In jamovi, where do you put continuous vs. dichotomous predictors?
Continuous predictors go in Covariates, dichotomous predictors go in Factors.
Formula for simple linear regression?
Y=b0+b1X+ϵ.
Formula for multiple linear regression?
Y=b0+b1X1+b2X2+…+ϵ
.
What does the intercept (𝑏0) represent?
The expected value of Y when all predictors are zero.
How do coefficients in multiple regression differ from simple regression?
They represent the unique effect of each predictor on Y, adjusted for the influence of other predictors
What are unstandardised coefficients (b)?
Raw-score coefficients, expressed in the units of Y (e.g., “grumpiness points per hour of sleep”)
What are standardised coefficients (β)?
Coefficients expressed in standard deviation units, allowing comparison of predictor importance
Why are standardised coefficients useful?
They put all predictors on the same scale, so we can judge which predictor has a stronger effect.
How does regression handle dichotomous predictors?
By coding groups (e.g., 0 = control, 1 = treatment), regression produces results equivalent to a t-test
What is the relationship between regression coefficients and t-tests?
b/SE=t, so regression subsumes the logic of t-tests
What does 𝑅2 represent?
The proportion of variance in Y explained by the predictors.
Why is overall fit important in psychology research?
It shows how well our model captures systematic variance in behaviour, but high R2 is rare because human behaviour is influenced by many factors.
What are the 4 main assumptions of regression?
Linearity, Independence, Normality, Equality of variance (homoscedasticity)
What happens if assumptions are violated?
Usually reduced precision (wider CIs), but regression is fairly robust. Non-linearity is most problematic.
Where is the outcome (Y) placed in jamovi regression?
In Dependent Variable.
