Fichas
week 4
Fichas del conjunto (11)
- 1usually we code the IV with 0 and 1. how would we interpret the b's if we'd code them with -1 and 1?b0 = the average [DV] of all subjects, regardless of [IV] --> average depression for all subjects, regardless of the intervention.
b1 = the difference in average [DV] between [IV=0] and the whole group; and also the difference in average [DV] between [IV=1] and the whole group --> the difference in average depression between the sitting group and the whole group; and also the difference in average depression between the running group and the whole group. - 2meaning of centringsubtracting the mean from every score of a variable so that the new average will be 0. Now, x = 0 represents the average of variable x instead of an actual value of 0.
you do this when a value of 0 for the IV is impossible or not meaningful. - 3when you have a GLM for multiple regression, how would you interpret b0 and b1?b0 = expected DV when both IV 1 and IV 2 are 0.
b1 for continuous IV = difference in expected DV when IV 1 increases by 1 while IV 2 remains constant (!!!)
b1 for nominal IV = difference in expected DV between IV2 group 1 and IV group 2 while IV1 remains constant. - 4explanation of R2 and its interpretationthe amount of variance in the DV that explained by the predictor(s).
simple regression: .01 is small, .09 is medium, .25 is large.
multiple regression: .02 is small, .13 is medium, .26 is large. - 5when do you use logistic regression?when you have a nominal dependent variable (e.g. recovery or no recovery).
you then calculate the probability that the outcome (y=1) occurs at a certain value of the predictor for subject i. - 6to test the fit of the model in logistic regression, the log-likelihood is used. what is that?the LL is the amount of UNexplained variance after the model. therefore, the larger the LL, the worse the model fits.
it thus tests the INACCURACY of the model.
the LL of a model is compared with the baseline model. if the model is better, there should be a significant reduction in inaccuracy. - 7what is used as a baseline model?the outcome that is observed most. it's the simplest model, without predictors (e.g. more recovered than not recovered, then baseline model predicts recovered).
- 8the odds ratio for logistic regression is Exp(b). how can we interpret that?= the change in odds of the outcome if the predictor increases by 1.
larger than 1 means that the DV will increase, smaller than 1 means that the DV will decrease. - 9what does this table tell you?whether adding IV2 to the model significantly improves the prediction of the DV COMPARED TO THE PREVIOUS MODEL.
it tells you the improvement of fit and whether this improvement is significant. - 10what does this table tell you?you can find the b-values and the exp(b) odds ratio here. both also tell you in which direction the IV changes the DV.
- 11check het einde van WA4 voor belangrijke notesmaak kaartjes van die notes
Información
- Descripción
- university evidence-based CP Fichas sobre week 4, creado por Sabine - el 29/09/2022.
- Creado por
- Sabine -
- Visitas
- 1
- Valoración