"Logit" is a contraction of "Log-Odds Unit" (pronounced "low-jit"). It is no more obscure a measurement unit of an underlying and invisible variable than an "Ampère" is of invisible electric current. The essential ingredient of Amps and logits is that they be additive.
Real apples are not additive. One Apple + One Apple = Two Apples. But Two Apples are twice as much as One Apple only when the Two Apples are perfectly identical. Real apples are not perfectly identical. When we say One Amp + One Amp = Two Amps, we say "all Amps are identical," wherever they appear on the Ammeter. Logits form an equal interval linear scale, just like Amps. When any pair of logit measurements have been made with respect to the same origin on the same scale, the difference between them is obtained merely by subtraction and is also in Logits. This is how Amps work.
Like an Ammeter, the logit scale is unaffected by variations in the distribution of measures that have been previously made, or by which items (resistances) may have been used to construct and calibrate the scale. The logit scale can be made entirely independent of the particular group of items that happen to be included in a test this time, or the particular samplings of persons that happen to have been used to calibrate these items.
We construct a logit scale in the same way that we construct an Amp scale. We deduce a theory that produces equal interval, linear measures and derive a method for applying that theory. In the case of qualitative ordered observations (right/wrong, present/absent, none/some/all), the necessary and sufficient theory is the Rasch model, and the method of application is numerous administrations of similar agents (test items) to relevant objects (persons).
The theory is
This is a "linear" model because all elements can be represented as fixed positions along one straight line. In games of chance, the (Probability of Success)/(Probability of Failure) is called the "odds of success". "Log_{e}[(Probability of Success)/(Probability of Failure)]" is called log-odds. The units of measurement constructed by this theory are called "log-odds units" or "logits".
Not all numbers represent equal interval scales, no matter how equally spaced their values appear. Rank orders are counted with equally spacing, but rank order numbers do not specify whether the distance between 1 and 2 is equal, greater or less than the distance between 2 and 3.
How do we know that logits are equal interval? By observing that when data fit the theory, the specification that a one logit positive difference between any person and any item anywhere on the scale always has the same stochastic consequence.
When data fit, the interval specification of the theory is realized in the data. For these data, the interval scale is established. The implication is that for similar data the scale will continue. This implication, however, is always tested when fit is analyzed for each new application, just as wise use of an Ammeter requires that it too be continually checked. When in the new application, the fit criteria are met, then the linear scale continues - the logit unit is maintained.
Benjamin D. Wright
Logits? Wright BD. … 1993, 7:2 p.288
Logits? Wright BD. … Rasch Measurement Transactions, 1993, 1993, 7:2 p.288
Rasch Publications | ||||
---|---|---|---|---|
Rasch Measurement Transactions (free, online) | Rasch Measurement research papers (free, online) | Probabilistic Models for Some Intelligence and Attainment Tests, Georg Rasch | Applying the Rasch Model 3rd. Ed., Bond & Fox | Best Test Design, Wright & Stone |
Rating Scale Analysis, Wright & Masters | Introduction to Rasch Measurement, E. Smith & R. Smith | Introduction to Many-Facet Rasch Measurement, Thomas Eckes | Invariant Measurement: Using Rasch Models in the Social, Behavioral, and Health Sciences, George Engelhard, Jr. | Statistical Analyses for Language Testers, Rita Green |
Rasch Models: Foundations, Recent Developments, and Applications, Fischer & Molenaar | Journal of Applied Measurement | Rasch models for measurement, David Andrich | Constructing Measures, Mark Wilson | Rasch Analysis in the Human Sciences, Boone, Stave, Yale |
in Spanish: | Análisis de Rasch para todos, Agustín Tristán | Mediciones, Posicionamientos y Diagnósticos Competitivos, Juan Ramón Oreja Rodríguez |
Forum | Rasch Measurement Forum to discuss any Rasch-related topic |
Go to Top of Page
Go to index of all Rasch Measurement Transactions
AERA members: Join the Rasch Measurement SIG and receive the printed version of RMT
Some back issues of RMT are available as bound volumes
Subscribe to Journal of Applied Measurement
Go to Institute for Objective Measurement Home Page. The Rasch Measurement SIG (AERA) thanks the Institute for Objective Measurement for inviting the publication of Rasch Measurement Transactions on the Institute's website, www.rasch.org.
Coming Rasch-related Events | |
---|---|
June 23 - July 21, 2023, Fri.-Fri. | On-line workshop: Practical Rasch Measurement - Further Topics (E. Smith, Winsteps), www.statistics.com |
Aug. 11 - Sept. 8, 2023, Fri.-Fri. | On-line workshop: Many-Facet Rasch Measurement (E. Smith, Facets), www.statistics.com |
The URL of this page is www.rasch.org/rmt/rmt72e.htm
Website: www.rasch.org/rmt/contents.htm