Testen en Meten colleges
De rode draad ............................................................................................................................................... 3
Hoorcollege 1 ................................................................................................................................................ 4
Why psychological tes/ng ma2ers to you..........................................................................................................4
Observable behavior and unobservable a2ributes ............................................................................................4
Wat is a psychological test? ...............................................................................................................................4
Type of tests .......................................................................................................................................................5
Challenges to measurement in psychology ........................................................................................................5
The importance of individual differences ...........................................................................................................5
Fundamental issues with numbers .....................................................................................................................5
Unit of measurement (hoofdstuk 2) ...................................................................................................................6
Addi/vity and coun/ng ......................................................................................................................................7
JASP ....................................................................................................................................................................8
Hoorcollege 2 .............................................................................................................................................. 10
The nature of variability (de aarde van de spreiding) ......................................................................................10
The importance of individual differences .........................................................................................................10
Variability and distribu/ons of scores ..............................................................................................................10
Quan/fying the associa/on between distribu/ons ..........................................................................................11
Variance and covariance for composite variables ............................................................................................12
Binary items .....................................................................................................................................................13
Interpre/ng test scores ....................................................................................................................................13
Hoorcollege 3 .............................................................................................................................................. 16
Test dimensionality...........................................................................................................................................16
Factor analysis .................................................................................................................................................18
Matrices ...........................................................................................................................................................21
JASP ..................................................................................................................................................................22
Factor analyse uitvoeren .............................................................................................................................22
Roteren ........................................................................................................................................................23
Hoorcollege 4 .............................................................................................................................................. 24
Overview of reliability and classical test theory ...............................................................................................24
Four ways to think about reliability..................................................................................................................25
Reliability and the 𝑠𝑒𝑚 (standard error of measurement)...............................................................................26
Toepassing van 𝑠𝑒𝑚 (H7) .................................................................................................................................27
Hoorcollege 5 .............................................................................................................................................. 28
Alternate forms of reliability (paralleltestbetrouwbaarheid) ...........................................................................28
, Test-retest reliability (test-hertestbetrouwbaarheid) .......................................................................................28
Internal-consistency reliability (interne consisten/e betrouwbaarheid) ..........................................................28
Methode 1 – split-half es<mates of reliability (splitsingsmethode) ............................................................29
Methode 2 (bekendste) – Cronbachs alfa ...................................................................................................29
Methode 3 – coëfficiënt KR20 .....................................................................................................................29
Sample homogeneity and reliability generaliza<on ....................................................................................31
Reliability of difference scores (betrouwbaarheid van verschilscores) .............................................................31
JASP ..................................................................................................................................................................32
Hoorcollege 6 .............................................................................................................................................. 33
Applied behavioural prac/ce (evalua/on of an individual’s test score) ...........................................................33
Effect size (effectgroo2e) .................................................................................................................................34
A2enua/e ........................................................................................................................................................34
Test construc/on and refinement .....................................................................................................................35
,De rode draad
Meten met tests en vragenlijsten
- Introduc5e
o Het belang van meten (H1)
o Verschillende meetniveaus (H2, WG1); gemiddelde, varian5e, samenhang
(H3)
- Hoe maak je meetwaardes?
o Testscore 𝑋! is de som van een aantal itemscores “composite score” (H3,
JASP-opdracht)
o Gemiddelde, varian5e en samenhang van composite scores (H3, JASP-
opdracht)
- Hoe interpreteer je meetwaardes?
o Normen (H3, jASP-opdracht)
o Bewerkte scores: z-score, percen5elen, IQ-score etc. (H3, JASP-opdracht)
- Hoe goed zijn de meetwaardes?
o Validiteit “meet de test wat die moet meten”: dimensionaliteit (H4, JASP-
opdracht, ook H8 t/m H12)
o Betrouwbaarheid “hoe nauwkeurig is de meetwaarde” (H5, H6, H7)
Grijs op de slides betekent niet voor op tentamen
, Hoorcollege 1
Why psychological testing matters to you
- Sollid understanding of tes5ng
o Test afnemenen (hbo-niveau)
o Test begrijpen (wo-niveau) nodig voor
§ Interpreteren van testscores (ook in JASP-opdracht)
§ Correct gebruik van tes5nforma5e
o Is het zinvol om een bepaalde test bij kleurters af tenemen?
o Hoe betrouwbaar is de test/hoe valide?
Observable behavior and unobservable attributes
- Werkgeheugen (P5) (o.a. onderdeel van de WISC) (oefening/test met cijfer reeks
onthouden)
o Defini5e: working memory is a cogni/ve system with a limited capacity that
can hold informa/on temporarily (wikipedia)
o Werkgeheugen is een niet observeerbaar constuct (hypothe5sche construct,
lantente trek, latente variabele) (kan je niet zien, alleen indirect meten)
o Onze constructen zijn bijna al5jd latent (observeerbaar) à weergegeven door
een elips (als je het grafisch weergee`)
o Wel observeerbaar zijn:
§ Aantal cijfers onthouden (digit-span)
§ Aantal woorden onthouden (reading-span)
§ Aantal rekensommen goed bij rekentest à is niet per se
werkgeheugen of rekenvaardighoud
§ Test zegt iets over construct, maar hoe` niet al5jd perfect te zijn
o Theorie nodig om hypothe5sche construct en observeerbaar gedrag te linken
§ Theorie is erg belangrijk om te begrijpen wat je meet
o Empirische onderzoek nodig om de “validiteit” en “betrouwbaarheid” te
bepalen
Wat is a psychological test?
(P6) Cronbach (1960): a systema5c procedure for comparing the behavior of two or more
people
- Gedragingen
- Gedragingen worden systema5sch verzameld
- Gegdragingen worden gebruikt om mensen te vergelijken
- Test hoe` geen groepsafname te zijn, zolang er normen zijn
- Niet alle tests hebben dezelfde kwaliteit
- Testen en meten: inleiding in kwaliteitskenmerken van tests, en leren hoe je deze
onderzoekt?
o Testscores
o Dimensionaliteit
o Betrouwbaarheid
o Validiteit
De rode draad ............................................................................................................................................... 3
Hoorcollege 1 ................................................................................................................................................ 4
Why psychological tes/ng ma2ers to you..........................................................................................................4
Observable behavior and unobservable a2ributes ............................................................................................4
Wat is a psychological test? ...............................................................................................................................4
Type of tests .......................................................................................................................................................5
Challenges to measurement in psychology ........................................................................................................5
The importance of individual differences ...........................................................................................................5
Fundamental issues with numbers .....................................................................................................................5
Unit of measurement (hoofdstuk 2) ...................................................................................................................6
Addi/vity and coun/ng ......................................................................................................................................7
JASP ....................................................................................................................................................................8
Hoorcollege 2 .............................................................................................................................................. 10
The nature of variability (de aarde van de spreiding) ......................................................................................10
The importance of individual differences .........................................................................................................10
Variability and distribu/ons of scores ..............................................................................................................10
Quan/fying the associa/on between distribu/ons ..........................................................................................11
Variance and covariance for composite variables ............................................................................................12
Binary items .....................................................................................................................................................13
Interpre/ng test scores ....................................................................................................................................13
Hoorcollege 3 .............................................................................................................................................. 16
Test dimensionality...........................................................................................................................................16
Factor analysis .................................................................................................................................................18
Matrices ...........................................................................................................................................................21
JASP ..................................................................................................................................................................22
Factor analyse uitvoeren .............................................................................................................................22
Roteren ........................................................................................................................................................23
Hoorcollege 4 .............................................................................................................................................. 24
Overview of reliability and classical test theory ...............................................................................................24
Four ways to think about reliability..................................................................................................................25
Reliability and the 𝑠𝑒𝑚 (standard error of measurement)...............................................................................26
Toepassing van 𝑠𝑒𝑚 (H7) .................................................................................................................................27
Hoorcollege 5 .............................................................................................................................................. 28
Alternate forms of reliability (paralleltestbetrouwbaarheid) ...........................................................................28
, Test-retest reliability (test-hertestbetrouwbaarheid) .......................................................................................28
Internal-consistency reliability (interne consisten/e betrouwbaarheid) ..........................................................28
Methode 1 – split-half es<mates of reliability (splitsingsmethode) ............................................................29
Methode 2 (bekendste) – Cronbachs alfa ...................................................................................................29
Methode 3 – coëfficiënt KR20 .....................................................................................................................29
Sample homogeneity and reliability generaliza<on ....................................................................................31
Reliability of difference scores (betrouwbaarheid van verschilscores) .............................................................31
JASP ..................................................................................................................................................................32
Hoorcollege 6 .............................................................................................................................................. 33
Applied behavioural prac/ce (evalua/on of an individual’s test score) ...........................................................33
Effect size (effectgroo2e) .................................................................................................................................34
A2enua/e ........................................................................................................................................................34
Test construc/on and refinement .....................................................................................................................35
,De rode draad
Meten met tests en vragenlijsten
- Introduc5e
o Het belang van meten (H1)
o Verschillende meetniveaus (H2, WG1); gemiddelde, varian5e, samenhang
(H3)
- Hoe maak je meetwaardes?
o Testscore 𝑋! is de som van een aantal itemscores “composite score” (H3,
JASP-opdracht)
o Gemiddelde, varian5e en samenhang van composite scores (H3, JASP-
opdracht)
- Hoe interpreteer je meetwaardes?
o Normen (H3, jASP-opdracht)
o Bewerkte scores: z-score, percen5elen, IQ-score etc. (H3, JASP-opdracht)
- Hoe goed zijn de meetwaardes?
o Validiteit “meet de test wat die moet meten”: dimensionaliteit (H4, JASP-
opdracht, ook H8 t/m H12)
o Betrouwbaarheid “hoe nauwkeurig is de meetwaarde” (H5, H6, H7)
Grijs op de slides betekent niet voor op tentamen
, Hoorcollege 1
Why psychological testing matters to you
- Sollid understanding of tes5ng
o Test afnemenen (hbo-niveau)
o Test begrijpen (wo-niveau) nodig voor
§ Interpreteren van testscores (ook in JASP-opdracht)
§ Correct gebruik van tes5nforma5e
o Is het zinvol om een bepaalde test bij kleurters af tenemen?
o Hoe betrouwbaar is de test/hoe valide?
Observable behavior and unobservable attributes
- Werkgeheugen (P5) (o.a. onderdeel van de WISC) (oefening/test met cijfer reeks
onthouden)
o Defini5e: working memory is a cogni/ve system with a limited capacity that
can hold informa/on temporarily (wikipedia)
o Werkgeheugen is een niet observeerbaar constuct (hypothe5sche construct,
lantente trek, latente variabele) (kan je niet zien, alleen indirect meten)
o Onze constructen zijn bijna al5jd latent (observeerbaar) à weergegeven door
een elips (als je het grafisch weergee`)
o Wel observeerbaar zijn:
§ Aantal cijfers onthouden (digit-span)
§ Aantal woorden onthouden (reading-span)
§ Aantal rekensommen goed bij rekentest à is niet per se
werkgeheugen of rekenvaardighoud
§ Test zegt iets over construct, maar hoe` niet al5jd perfect te zijn
o Theorie nodig om hypothe5sche construct en observeerbaar gedrag te linken
§ Theorie is erg belangrijk om te begrijpen wat je meet
o Empirische onderzoek nodig om de “validiteit” en “betrouwbaarheid” te
bepalen
Wat is a psychological test?
(P6) Cronbach (1960): a systema5c procedure for comparing the behavior of two or more
people
- Gedragingen
- Gedragingen worden systema5sch verzameld
- Gegdragingen worden gebruikt om mensen te vergelijken
- Test hoe` geen groepsafname te zijn, zolang er normen zijn
- Niet alle tests hebben dezelfde kwaliteit
- Testen en meten: inleiding in kwaliteitskenmerken van tests, en leren hoe je deze
onderzoekt?
o Testscores
o Dimensionaliteit
o Betrouwbaarheid
o Validiteit
