Computer Adaptive Testing

What We Know Now

Description

A form of com­puter based test­ing that pro­gres­sively adapts to the student’s abil­ity level. The sys­tem starts with a series of ques­tions based upon the esti­mate of the user’s abil­ity. The sys­tem then includes pro­gres­sively more dif­fi­cult ques­tions until the stu­dent starts to answer them incor­rectly at which point the sys­tem accu­rately deter­mines the abil­ity level of the per­son being tested.

Context:

Com­puter adap­tive test­ing is typ­i­cally used mul­ti­ple times over the course of a school year as a for­ma­tive assess­ment tool. The tool is designed to pro­vide feed­back to the teacher and the learner allow­ing for the adjust­ment of instruc­tional meth­ods and activities.

There are a vari­ety of test­ing prod­ucts on the mar­ket today includ­ing com­puter adap­tive tools designed to deter­mine the pro­fi­ciency level of the learner. These tools bill them­selves as a for­ma­tive test­ing sys­tem that responds dynam­i­cally to the learner allow­ing for mul­ti­ple assess­ments through­out the course of the school year.

Considerations or Implications:

It is unclear as to whether the new Smarter Bal­anced Assess­ment sys­tem, the replace­ment for the Wis­con­sin Knowl­edge and Con­cepts Exam­i­na­tion (WKCE), will be con­sid­ered a com­puter adap­tive test­ing sys­tem. Early indi­ca­tions are that this sys­tem will not meet that cri­te­ria and would be more accu­rately char­ac­ter­ized as a computer-​based test that does not level the ques­tions to the abil­ity of the learner. Instead the sys­tem will pro­vide a set of ques­tions based upon the grade level of the learner and assess based upon the stan­dards and bench­marks for that grade level.

Dis­tricts wish­ing to aug­ment their assess­ment tools to pro­vide more for­ma­tive assess­ment data may wish to exam­ine com­puter adap­tive test­ing tools and com­pare them to other more tra­di­tional for­ma­tive assess­ment systems.

Advantages:

Adap­tive tests have the advan­tage of pro­vid­ing uni­formly pre­cise scores for most test-​takers as opposed to stan­dard fixed tests that tend to pro­vide the best pre­ci­sion for users of medium abil­ity. Stan­dard fixed tests have a more dif­fi­cult time pre­cisely mea­sur­ing the abil­ity level for those learn­ers at the high and low extremes.

An adap­tive test can also be up to 50% shorter than a fixed ver­sion which may trans­late into a time sav­ings. This may be espe­cially ben­e­fi­cial given the desire to give for­ma­tive assess­ments mul­ti­ple times over a year.

Disadvantages:

Any adap­tive sys­tem requires a rig­or­ous cal­i­bra­tion to ensure valid­ity and reli­a­bil­ity of results. This cal­i­bra­tion is done by the ven­dor which explains the high costs of this type of system.

Any such sys­tem is heav­ily depen­dent on fre­quent access to the school’s com­puter lab which would limit access for instruc­tional purposes.

Resources:

International Association for Computerized Adaptive Testing
Concerto: Open-source CAT Platform
CAT Central
by David J. Weiss
Frequently Asked Questions about Computer-Adaptive Testing (CAT). Retrieved April 15, 2005
Computerized Adaptive Tests — from the Edu­ca­tion Resources Infor­ma­tion Center


David Gund­lach, WASCD Board of Directors

Close this window

ascd logo

dpi logo

Registered User Login