Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Save to My Library
Look up keyword or section
Like this
3Activity
P. 1
An Introduction to Modern Measurement Theory

An Introduction to Modern Measurement Theory

Ratings: (0)|Views: 84 |Likes:
Published by AZOGTHOTH

More info:

Published by: AZOGTHOTH on Aug 15, 2011
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

01/29/2014

pdf

text

original

 
 
An Introduction to Modern Measurement Theory
This tutorial was written as an introduction to the basics of item response theory (IRT) modelingand its applications to health outcomes measurement for the National Cancer Institute’s Cancer Outcomes Measurement Working Group (COMWG). In no way this tutorial is meant to replaceany text on measurement theory, but only serve as a stepping-stone for health care researchers tolearn this methodology in a framework that may be more appealing to their research field. Anillustration of IRT modeling is provided in the appendix and can provide better insight into theutility of these methods. This tutorial is only a draft version that may undergo many revisions.Please feel free to comment or ask questions of the author of this piece:Bryce B. Reeve, Ph.D.Outcomes Research BranchApplied Research ProgramDivision of Cancer Control and Population Sciences National Cancer InstituteTel: (301) 594-6574Fax: (301) 435-3710Email: reeveb@mail.nih.gov
 
 2
Outline
Topic Page Reference Terms……………………………………………………………………………3 An Introduction to Modern Measurement Theory……………………………………………. 4 A Brief History of Item Response Theory……………………………………………….. 5 Item Response Theory Basics……………………………………………………………. 7 Assumptions of Item Response Theory Models……………………………………. 11 Item Response Theory ModelsTwo Theories towards Measurement…………………………………………………….. 13 The IRT Models………………………………………………………………………….. 14 The Rasch Simple Logistic Model…………………………………………………... 15The One-Parameter Logistic Model…………………………………………………. 17The Two-Parameter Logistic Model………………………………………………… 18The Three-Parameter Logistic Model……………………………………………….. 19The Graded Model…………………………………………………………………... 20The Nominal Model…………………………………………………………………. 22The Partial Credit Model……………………………………………………………..23The Rating Scale Model……………………………………………………………... 24 Trait Scoring……………………………………………………………………………... 25 Classical Test Theory and Item Response Theory……………………………………….. 31 Applications of Item Response Theory in ResearchItem and Scale Analysis………………………………………………………………….. 34 Differential Item Functioning……………….…………………………………………… 38 Instrument Equating and Computerized Adaptive Tests………………………………… 45 Conclusion……………………………………………………………………………………. 47 References…………………………………………………………………………………….. 51 Illustration of IRT Modeling…………………………………………………………... appendix
 
 3
Reference: Some Terms You Will See in this Tutorial as well as in the Literature
Assumption of Local Independence – A response to a question is independent of responses toother questions in a scale after controlling for the latent trait (construct) measured by the scale.Assumption of Unidimensionality - the set of questions are measuring a single continuous latentvariable (construct).Construct (a.k.a. trait, domain, ability, latent variable, or theta) - see definition of theta below.Information Function (for a scale or item) - an index, typically displayed in a graph, indicatingthe range of trait level
θ
over which an item or test is most useful for distinguishing amongindividuals. The information function characterizes the precision of measurement for persons atdifferent levels of the underlying latent construct, with higher information denoting more precision.Item – Question in a scaleItem Characteristic Curve (ICC, a.k.a. item response function, IRF, or item trace line) – The ICCmodels the relationship between a person’s probability for endorsing an item category and thelevel on the construct measured by the scale.Item Difficulty (Threshold) Parameter 
b
– point on the latent scale
θ
where a person has a 50%chance of responding positively to the scale item.Item Discrimination (Slope) Parameter 
a
- describes the strength of an item's discrimination between people with trait levels (
θ)
below and above the threshold
b.
The
a
parameter may also be interpreted as describing how an item may be related to the trait measured by the scale.Scale (measure, questionnaire, or test) – A scale in this tutorial is assumed to measure a singleconstruct or domain.Slope Parameter – See Item DiscriminationTest Characteristic Curve (TCC, a.k.a. Test Response Function) - The TCC describes theexpected number of scale items endorsed as a function of the underlying latent variable.Theta (
θ
)– The unobservable (or latent) construct being measured by the questionnaire. Theseconstructs or traits are measured along a continuous scale. Examples of constructs in healthoutcomes measurement are depression level, physical functioning, anxiety, or social support.Threshold Parameter – See Item Difficulty

Activity (3)

You've already reviewed this. Edit your review.
1 thousand reads
1 hundred reads
verai1131 liked this

You're Reading a Free Preview

Download
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->