Naslov
Can computer based testing provide additional information in the psychometric analysis of test results?

Autori
Ljubotina, Damir ; Marinović, Mladen

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni

Izvornik
19. Dani Ramira i Zorana Bujasa: program i sažeci priopćenja / Ljubotina, Damir ; Kamenov, Željka ; Mikac, Una ; Urch, Dražen - Zagreb : Odsjek za psihologiju Filozofskog fakulteta Sveučilišta u Zagrebu, 2009, 65-65

ISBN
978-953-175-339-5

Skup
19. Dani Ramira i Zorana Bujasa

Mjesto i datum
Zagreb, Hrvatska, 22.04.2009. - 25.04.2009

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Domaća recenzija

Ključne riječi
paper and pencil tests; computer based testing; psychometric parameters

Sažetak
Classical paper and pencil tests use the number of correct responses as the key parameter used in psychometric analyses and test performance assessment. Development of computer based testing can provide additional information related to test performance as well as the processes involved in responding to individual items. Information processing theories offer models that see the speed and accuracy of information processing as two important indicators of intelligence, even though the relationship between these two dimensions has not yet been completely explored. Th e authors of this paper have developed a new computer based testing system, which enables administration of diff erent kinds of psychological tests while registering a number of newly introduced parameters connected with test performance. Th is paper presents the results of a cognitive ability test, consisting of 58 questions with 6 possible answers. Th e test was administered to a sample of 130 participants with time limitations long enough for all of the participants to go through all of the items. Th e system can register a variety of performance indicators such as: number of correct answers, total time spent on each item, time before off ering an answer on each item, total time required to go through the whole test, a total number of changed answers on each item. The aim of this paper was to determine the potential of these measures to provide additional information in psychometric analyses and performance assessments. Th e results obtained suggest a potential use of response time as a measure of item difficulty and validity, as well as in distracter analysis. Item characteristic curves can be used in order to analyze item response time in relation to traits measured by the test. When analyzing individual results, useful information can be obtained through comparing respondents based on the speed and accuracy of their responses.

Izvorni jezik
Engleski

Znanstvena područja
Psihologija

POVEZANOST RADA