Degrees of freedom Flashcards

1
Q

DF for F test for linear regression

A

F= MSR/MSE
df numerator (MSR) : k –» nb of predictors
df of denominator(MSE): n-(k+1)–» df of errors

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

DF for T test

A

only df you really need is errors DF, so
n-(k+1)
As usual, total df is n-1 and if you don’t have df of regression and need it to compute MSR, its k (nb of predictors)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Df to use for Nested F-test

A

Df Numerator: k-g: si ca te melange en vrai c juste le nombre de predictors que tu rajoutes en passant du R: reduced model au C: complete model
DF denominateur: df des erreurs du modèle complet

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

DF when testing for equal variance (when testing to see if there is multicollinearity)

A

Group 1(numerateur): df= n1-k-1
Group 2(denominateur): df= n2-k-1

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Df pour compute cook’s D

A

Df numerateur: k+1
df dénominateur: n-(k+1)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Anova

A

Df des treatments: p-1, so si ton treatment c’est pays et y’en a 4 bah ton df c 3
les erreurs c n-p

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Df for bonferroni correction

A

df= n-p, p est le nombre de populations que tu compares

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Df avec replication

A

sont sur la feuille de formule mais FYI, la variable r veut dire nombre de replications, pis tes df des interaction c’est la multipliaction des df des terme qui la constitue

How well did you know this?
1
Not at all
2
3
4
5
Perfectly