Saturday, January 29, 2011

Modern Measurement: Modern Scaling with Item Response Theory : Chapter 10


 

Introductory Description of IRT

-          What is IRT

o   Psychologically based theory of mental measurement that specifies information about latent traits and the characteristics of stimuli used to represent them.

o   IRT statistics are not non-parametric

-          Relation of IRT to CTT

-          Cautionary note on studying IRT

o   IRT : theory about latencies and the way they can be estimated

 

IRT and Invariant Measurement for Items and Persons

-          Problem of lack of an independent scale in CTT

o   Difficulty in comparing "low esteem" in one test vs another test.

o   No common zero point

o   Converting scores to z scores does not solve problem – then you only have scores expressed in same metric

-          Group Dependent Items and Item Dependent Groups

o   Left with relative comparisons

o   Eg, test on history – what measurement best represents difficulty of test à depends on group who takes test à eg, primary school vs college

o   Difficulty value is group dependent

o   Reference group

o   Measuring examinee ability is item dependent

-          IRT as item and person invariant measurement

-          Notion of invariant measurement

o   Invariance is an estimable concept  [GW – not sure I understood or agreed with this section]

 

Introduction to IRT Models

-          Some commonly used IRT Models

-          Models are usually identified by number of characteristics they estimate about a test's stimuli

o   One – parameter

o   Two / more

-          Most popular models

o   One-parameter

§  Only item difficulty is estimated

o   Two parameter

§  Estimates separate difficulty and discrimination parameters for each item

o   Three parameter

§  Includes examinee's probability of guessing or pseudochance

 

Assumptions

-          Centrality of assumptions to IRT

-          Unidimensionalty of items and tests

o   Given test item or exercise is directly targeted at single cognitive process, and in theory it fills that latent space completely

-          Local independence

o   Examinees response to a given specific measure reflects an independent and autonomous reference to a latent trait in cognition.

o   Examinee responds to stimulus of test item or stimulus, also approaches the stimulus without also thinking about other items or exercises.

o   Degree of learning as more items are encountered

-          Item characteristic curve

o   Defined characteristics of test stimuli are reliably estimable functions

-          Certainty of response

o   Optimal performance

 

ICC and IRC

-          Specifying ICCs generally

-          Inflection point

-          Scales allow a trace line to describe functional relationship between characteristics of an item and the trait level of examinee

 

 

IRT Models

-          Likelihood function

o   Examinee of particular ability level has certain probability of getting an item correct

o   .likelihood function is joint probability of getting several items correct or incorrect

o   More on working with the log scale

o   The two parameter model

 

The one-parameter IRT model and Rasch

-          The Rasch Model

 

Other IRT Models

 

-          Nominal and graded response models for polytomous items

-          Richly cognitive models

 

Estimating Item and Ability Parameters

-          Iterative estimation procedures

-          Developing priors

-          Test information function

-          Some estimation procedures

 

Computer Programs Available for parameter estimation

 

 

Brief History and major Contributors to IRT

 

 

 

 

 

 

No comments: