100% satisfaction guarantee Immediately available after payment Both online and in PDF No strings attached 4.2 TrustPilot
logo-home
Class notes

Lecture 6 statistiek 4.4

Rating
-
Sold
-
Pages
11
Uploaded on
25-01-2022
Written in
2021/2022

Uitwerking lecture 6

Institution
Course









Whoops! We can’t load your doc right now. Try again or contact support.

Connected book

Written for

Institution
Study
Course

Document information

Uploaded on
January 25, 2022
Number of pages
11
Written in
2021/2022
Type
Class notes
Professor(s)
Oliver lindemann
Contains
6-8

Subjects

Content preview

Lecture 6 statistiek blok 4.4


6A: ANCOVA DEEL 2
Stappen ANCOVA:
1. Measurement level of the variables
2. Check for outliers
3. Check assumptions
• Normality (for each group)
• Homoscedasticity
• Linearity
• Homogeneity of regression slopes
• Independence of treatment and covariate
• (maybe) multicollinearity
4. Check quality of prediction

Checking for outliers: univariate à save à leverage,
cooks, standardized residuals.




Als je niet zeker bent welke je moet kiezen
dan voor het grotere criterium (3(k+1)/n.


Normally distributed errors à als groep varianties gelijk zijn, kunnen we de assumptie overall
kijken (niet per groep). SPSS: save standardized residuals, maak histogram (check voor outliers)




Homoscedasticity (errorvarianties in elke groep moeten gelijk zijn) à Levene’s test.

Als Levene’s test sign.
is dan significant
verschil in
errorvarianties. Dit wil
je niet!!




1

, Lecture 6 statistiek blok 4.4


Linearity (tussen covariate en dependent variabele) à bij elke groep! Graphs à scatterplot

X-as: covariate
Y-as: dependent variable

Er is een lineaire relatie voor de drie groepen.




Homogeneity of regression slopes (bij second continue variabele;covariate)
à Heel belangrijk, VIOLATED? à GEEN ANCOVA
1. See parallel regression lines in previous scatterplot
2. Test violation of assumption by: repeating ANCOVA en including interaction effect:
covariate x group
o Analyze à General Lineair Model à Univariate à model à build terms à
(group en T0QL tegelijk selecteren en dan met het pijltje naar links verschuiven
dan krijg je het interactie effect).
o Sign? dan interactie en kan je stoppen want dan zijn de lijnen niet paralell
o Xsign? dan kan je verder gaan dan is er geen interactie tussen group en covariaat,
dus geen indicatie voor verschillende regressieslopes.




Haal interactie-effect er weer uit, want beïnvloed F-value. Dit doe je door bij model op full-
factorial te klikken.

Independence of treatment and covariate (bij second continue variabele; covariate)
Average size of covariate moet hetzelfde zijn voor de verschillende groepen.


p > .05 dit wil je want dan
geen verschil.



(maybe) multicollinearity (bij second continue variabele;covariate)
= different predictor variabelen niet correleren. Multiple covariates Xcorrelated.




2

Get to know the seller

Seller avatar
Reputation scores are based on the amount of documents a seller has sold for a fee and the reviews they have received for those documents. There are three levels: Bronze, Silver and Gold. The better the reputation, the more your can rely on the quality of the sellers work.
JOPsychologie Erasmus Universiteit Rotterdam
Follow You need to be logged in order to follow users or courses
Sold
13
Member since
9 year
Number of followers
13
Documents
19
Last sold
1 year ago

5.0

1 reviews

5
1
4
0
3
0
2
0
1
0

Recently viewed by you

Why students choose Stuvia

Created by fellow students, verified by reviews

Quality you can trust: written by students who passed their tests and reviewed by others who've used these notes.

Didn't get what you expected? Choose another document

No worries! You can instantly pick a different document that better fits what you're looking for.

Pay as you like, start learning right away

No subscription, no commitments. Pay the way you're used to via credit card and download your PDF document instantly.

Student with book image

“Bought, downloaded, and aced it. It really can be that simple.”

Alisha Student

Frequently asked questions