Scribd
Upload
48 views

Session 5 Rasch Measurement Model

This document discusses the Rasch measurement model, which was developed by Danish mathematician Georg Rasch in 1960. [1] The model provides a probabilistic relationship between a person's ability and an item's difficulty that allows their locations along a common variable to be measured. [2] Specifically, the Rasch model establishes that the difference between a person's ability and an item's difficulty determines the probability that the person will correctly answer that item. [3] This relationship transforms raw scores into linear measures on an ability or difficulty scale.

Uploaded by

akd fpsi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
48 views

Session 5 Rasch Measurement Model

This document discusses the Rasch measurement model, which was developed by Danish mathematician Georg Rasch in 1960. [1] The model provides a probabilistic relationship between a person's ability and an item's difficulty that allows their locations along a common variable to be measured. [2] Specifically, the Rasch model establishes that the difference between a person's ability and an item's difficulty determines the probability that the person will correctly answer that item. [3] This relationship transforms raw scores into linear measures on an ability or difficulty scale.

Uploaded by

akd fpsi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 17

Session 5

RASCH MEASUREMENT MODEL

Bahrul Hayat
Faculty of Psychology
State Islamic University Jakarta
Pioneers of objective and linear
psychological and social measurement
Rasch Measurement Model
• Complete solutions to Thurstone's requirements did not
emerge until 1953 when Danish mathematician, Georg Rasch
(1960) deduced the mathematical model for objective
measurement.

• A person having a greater ability than another should have the


greater probability of solving any item of the type in question,
and similarly, one item being more difficult than another one
means that for any person the probability of solving the second
item correctly is the greater one. (Rasch, 1960, p.117)
Reasonable model when a person answers an item
• We would like a person v’s ability hat is their location on the
variable, to govern how far along the variable we can expect
them to produce correct responses to items.

• We would also like item i’s difficulty that is its location on the
variable, to determine how far along the variable we can expect
correct responses to that item to occur.

• Indeed that is the only situation in which we can use item


difficulties and a person’s responses to them as the basis for
measuring the person.
The Essential Conditions Causing a Response
Item difficulty

Person Ability 𝑿 𝒗𝒊 Observed


Response

• When the response is "correct", x = 1


• When the response is "incorrect”, x=0.
• Our next step is to decide how we want person ability and
item difficulty to interact in order to produce .

• What is a reasonable and useful way to set up a


mathematical relation between and ?

• Since we require that represent locations along one


common variable which they share, it is their difference (- )
which is the most convenient and natural formulation of their
relation.
Sometimes > fails to answer correctly, and
Sometimes < succeed to answer correctly.

We need probabilistic relationship (model)


not deterministic relationship (model)
Difference between person ability and item difficulty
affect the probability of a correct response

When > so (- ) > 0 and P > 1/2

When < so (- ) < 0 and P < 1/2

When = so (- ) = 0 and P = 1/2


Item
characteristic
curve (ICC)
Rasch Measurement Model
• We want (- ) to govern the probability of what is supposed
to happen when person v uses their ability against the
difficulty of item i.

• But the (- ) can vary from minus infinity to plus infinity while
the probability must remain zero to one.
Rasch Measurement Model
To deal with this we apply the (- ) as an exponent of natural
constant e= 2.71828 and write the result as:

= exp (- )

Form the ratio:

exp (- ) / 1+ exp (- )
Rasch Model for Dichotomous Data

P
Family of Rasch Models
Attribute Data type Rasch
Measured Model
• Ability Dichotomy Dichotomous
• Non Ability
Ability Polytomy Partial Credit

Non Ability Polytomy Rating Scale

• Ability Multi-facet Many Facet


• Non Ability
Rasch model transform ordinal raw scores
into linear measures.

The growth of science depends on the


development of objective methods for
transforming observation into measurement.

Benjamin Wright
Objectivity in Mental
Measurement
ETS Invitational Conference
on Testing Problems
October 28, 1967

Sample-free Test Calibration and Person


Measurement
Benjamin D. Wright
The end

You might also like