This is a visual English summary of the Measuring and Diagnosing 2 course at the VU Amsterdam (second year). It includes everything from the lectures and some repetition from MD1 and Stat 1 for extra understanding. Good luck studying!
LECTURE 1 : Introduction
,
CHI
A
psychological
MEASUREMENT IN PSYCHOLOGY variable or construct
a
Psychological the assignment of to characteristics
measurement =
numerals
psychological
of individuals ( according to rules )
psychometrics = the Science concerned with the attributes (characteristics ) of
psychological tests
(e.g Validity
.
,
construct bias etc .
)
hypothetical construct
A
Psychological variables are latent variables / latent traits =
they are not
directly observable (but are real )
↳ so their measurement depends on the measurement of observable behavior
↳ To link observable behavior to (unobservable ) traits need :
psychological ,
we
1 .
A psychological
theory
2 .
Causality → t operational definitions
3
.
Statistics
Diagram
'
latest variable
' '
the observed variable
'
representation of & relation
7 .
Latent variable
latent →
Can be
variable e.g .
Psychopathology ,
personality , cognition .
development
^ r n
b b b
observed variables
.
2 .
Response items
items
item 2
items →
These items are all indicators of the latent variable
f
eS -
. items on a questionnaire about behavior or reaction time etc .
3 .
Error
err or
,
→ All measurement is subject to measurement errors
+ The observed variable ( response item) is related to the latent variable ,
but never perfectly
(reliability is
hardly ever 1)
THE ROLE OF STATISTICS → needed for :
1 . Psychometric analysis =
analysis of individual differences in item responses
}
-
For latent variables : Variance standard deviation
,
descriptive statistics
-
for observable variables : Variance ,
standard deviation ,
mean &
probability
linking latent & observable traits linear → Yi a t b Xi tei
=
with model
2 .
a
regression
stope
t ↳ error
→ linear when the dependent variable distributed intercept
regression is
continuously
→ But logistic regression when the dependent
we use a
variable is
binary or dichotomous (e.g .
o or 1)
For between variables ( items )
}
establishing association
.
↳ The Pearson product moment correlation coefficient expresses the linear
relationship between variables in a correlation matrix
→
Items should be related bc they should all measure the same latent variable
,THE ROLE OF PSYCHOLOGICAL THEORY
The identifies the
1 .
psychological theory observable items of interest to measure the latent variable
↳ So it's needed for an operational definition of the latent variable in terms of observable variables
→
e.g ,
relevant observable behavior to measure neuroticism
2 .
Psychological theory informs statistics
↳ We based the theory
' '
make distributional assumptions on
↳ The the distribution
theory informs our expectations of of the latent variable
→
e.g .
Continuous , latent classes or discrete t ordinal
,
nominal ,
interval ,
ratio
THE ROLE OF CAUSALITY
→ Implicit y ,
we adhere to a causal model
↳ We assume there is a causal relation between the
latent variable & the observable variable
↳ Thus : You're position on the latent variable of
Reflective model =
depression should inform your item responses
-
latent variable causes reflective items ↳They are reflective items (also called reflective indicators )
-
Reflective items reflect latent variable (of the latent variable )
→
Reflectivity is a theoretical assumption
( thus based on the
theory of the latent variable)
But not all test items are reflective
↳ e. g. APGAR test on babies : A low score doesn't cause bad breathing
↳ A low score is a manifestation of bad
breathing
→ aka the observable variable is formative to the score on the latest
Formative model =
variable test
- The item scores determine the test score
ya
.
-
The test score is not causal to the observed items The difference between reflective I
formative modal will be on the exam !
Causality can also be applied to
group differences
↳ This idea adds an extra box to our previous model
bsex
latent c Group difference =p e. g .
The influence of sex on the latent variable
variable
spatial ability ( in red )
^ r n
↳ If bsex > 0 then sex differences will be observed
bsexz
in the latent variable I thus the item responses
observed variables L l
to
.
item y item 3
item 2
= Causal mediation model
f ↳ The latent variable mediates the relationship
between sex & the item scores
error
7
Bc . Sex predictive of spatial
is
ability
.
And spatial the item
ability predicts responses
→ However : It is
possible there's a violation of the causal
mediation model → e. g .
Sex influences item 3 only or
way more ( in blue ) ( bsexz.IO)
means there is difference that's not
directly caused by spatial ability
→ in
a sex item 3
D H 's even more embarrassing if bsex = o but bsexz to GO ) → Then the difference is no
longer
at the latent variable level → Basis for construct bias
, Causality can also be used to look for the influence ( relations between two latent variables
e. g. spatial
ability
e. g.
memory
latent latent
Variable 2
Variable 7
n n n
^ r n
b
observed variables
observed variables
. L .
item 7 item 3
item 7 item 3 item 2
item z
t q q f q q
error error
error error error
error y §
^ 72
z B
-
b. should be 0
Because it that item 3 for spatial
→ if not ,
means
ability is
picking up
signal from another latent variable leg memory )
.
Violation of
= Uni
dimensionality
↳ The items should be reflective of latent variable
only one
INFORMATION FROM THE BOOK
Psychological tests must
-
be
capable of
comparing
→ Inter individual differences ( between people)
→
Intra individual differences ( within a
person )
- Tests can in : Content
vary
-
Response option ( e.g closed
questions)
-
or
.
open
Methods used to administer the test
-
-
Reference used
↳ Criterion off Ccr !
referenced = use of a cut -
! er
:
on
)
↳ Norm referenced Use of to reference sample
score
(norm group )
= a comparison a
↳ Needs to be representative
-
Speeded us . Power tests
↳
Speeded ±
Counts the nr . of questions answered in on allot ed time period Aine - limit )
↳ Counts
Power = nr ,
of correct answers ( no time - limit
,
wants all questions answered )
Psychometrics
'
'
is the science concerned
→
3 attributes of interest
with
evaluating the attributes of psych tests
.
1 .
Type of info (e.g .
Scores) generated by these tests
The the data
2 .
reliability of
3 .
The validity of the data
^
Challenges to measurement in
psychology
1 .
The complexity of psychological phenomena
2 .
Participant reactivity = the act of measurement can
change how people act
, affecting what's
being measured
↳ Demand characteristics figure
=
trying to out the research purpose & acting in accordance to it
↳ Social behavior to the
desirability =
changing impress researcher
↳
Malingering =
changing behavior to leave a
poorer impression on the researcher
3 .
Biases (e.g ,
observer bias )
( scale)
4
.
Score sensitivity = The measure chosen needs to be sufficiently sensitive to pick up small differences
Psychological the assignment of to characteristics
measurement =
numerals
psychological
€How
of individuals ( according to rules )
is assignment operationalized ?
the assignment
↳ By of numerals to objects according to rules
r .
Objects are reflective indicators
2 . You assign numerals G. 2 or a
,
b etc) to these indicators
3 .
With this you arrive at the assignment
,
of numerals
at the latent level ( inference )
Numerals are informative
,
they San convey
:
↳ So not all do convey all things
1
Identity
e. g .
You assign people the numeral of N for normal or MD for manic depression for o
,
1)
Through doing this you can now tell when people are identical ( NIN or MD IMD )
& when people are different ( Nd MD )
LD The numerals are mutually exclusive
of meaning each individual can
only be assigned one
-
exhaustive value at time
single unique given
-
a
↳ But note that the assignment
-
May change over time (no stability )
-
May be incorrect due to Meas .
error
2 Order
e- 9 .
Having 3 categories of normie ( N ) ,
dy sthenic ( D ) I major depression ( MD ) , implies Nfo) L DH L MD 14
→ Rank order
along a
single dimension implies transitivity : 277 and I >o implies 220
3
Quantity
↳ The adds information the to the
property of
quantity about amount numeral
e.
g length -
Identity = Tall us .
Short
→
Order = Tall ,
normal short
,
-
Quantity =
length expressed in centimeters f- an arbitrary scale
,
works
by agreement)
→
Quantity requires a definition of units of measurement (e.g ,
centimeters ) (often divisible)
Rank with the of quantity also implies transitivity
order
property
→
Absolute zero us .
relative zero
↳ A scale may have an absolute zero ( then o means the absence of the attribute e.
g. distance travelled )
↳ Or a scale can have a relative zero (here o does not mean
B
e -
g .
temperature in Celsius )
↳ Most (except time )
psychological measurement uses a relative o reaction
Voordelen van het kopen van samenvattingen bij Stuvia op een rij:
Verzekerd van kwaliteit door reviews
Stuvia-klanten hebben meer dan 700.000 samenvattingen beoordeeld. Zo weet je zeker dat je de beste documenten koopt!
Snel en makkelijk kopen
Je betaalt supersnel en eenmalig met iDeal, creditcard of Stuvia-tegoed voor de samenvatting. Zonder lidmaatschap.
Focus op de essentie
Samenvattingen worden geschreven voor en door anderen. Daarom zijn de samenvattingen altijd betrouwbaar en actueel. Zo kom je snel tot de kern!
Veelgestelde vragen
Wat krijg ik als ik dit document koop?
Je krijgt een PDF, die direct beschikbaar is na je aankoop. Het gekochte document is altijd, overal en oneindig toegankelijk via je profiel.
Tevredenheidsgarantie: hoe werkt dat?
Onze tevredenheidsgarantie zorgt ervoor dat je altijd een studiedocument vindt dat goed bij je past. Je vult een formulier in en onze klantenservice regelt de rest.
Van wie koop ik deze samenvatting?
Stuvia is een marktplaats, je koop dit document dus niet van ons, maar van verkoper FabienneDenberg. Stuvia faciliteert de betaling aan de verkoper.
Zit ik meteen vast aan een abonnement?
Nee, je koopt alleen deze samenvatting voor €8,99. Je zit daarna nergens aan vast.
