• Aucun résultat trouvé

Consistency in inconsistent responses to multiple-choice tests

N/A
N/A
Protected

Academic year: 2022

Partager "Consistency in inconsistent responses to multiple-choice tests"

Copied!
1
0
0

Texte intégral

(1)

CIPE 2016 | VIII CONGRESO INTERNACIONAL DE PSICOLOGÍA Y EDUCACIÓN | VIII INTERNATIONAL CONGRESS OF PSYCHOLOGY AND EDUCATION

1743

www.cipe2016.com ORALES

Consistency in inconsistent responses to multiple-choice tests

Doval, E.1, Riba, M.D.1, Fuentes, M.2 y Renom, J.3

1Departament de Psicobiologia i de Metodologia de les Ciències de la Salut, Universitat Autònoma de Barcelona, Bellaterra, España

2Departament de Psicologia Básica, Evolutiva i de l’Educació, Universitat Autònoma de Barce- lona, Bellaterra, España

3Departament de Metodologia de les Ciències del Comportament, Universitat de Barcelona, Bellaterra, España

[email protected]

The assessments conducted by multiple-choice tests often used as an indicator of the level of knowledge or skills the score obtained from the number of questions answered correctly. This score, however, may be obtained from diferent responses, some of them setting an aberrant response pattern (ARP), for example, such as when difficult questions are answered correctly, yet very easy questions are an- swered wrong.

In this work we aim to provide evidence on whether the presence of ARP in mul- tiple-choice tests corresponds to isolated features, or else it can be considered a consistent style of the person.

The answers to two tests for evaluating basic competencies of Spanish and Catalan applied during the academic year 2012-13 to students of 4 ESO in Catalonia are an- alyzed. 56158 students completed the Spanish test and 56450 the Catalan test. The average proportion of correct answers in the tests was 0.83 and 0.81 respectively, with standard deviations of 0.09 and 0.14. The percentage of students who had a score associated with PAR was 29.3% in the Spanish test and 24.5 % in the Catalan test. 8.5% of students answered during the two tests following an atypical pattern.

Atypical answers to a test cast doubt on the validity of the resulting scores. The fact that this form of answer is kept in different tests can indicate the presence of any structural problems in the teaching-learning process. Some of these problems may be related to deficits in classroom work, with poor study guidelines, with test anxiety or simply with test cheating.

Key words: multiple-choice tests; aberrant response patterns; learning styles

Références

Documents relatifs

This note evaluates the scrambled questions penalty using multiple choice tests taken by first-year undergraduate students who follow a microeconomics introduc- tory course.. We

This study investigated government and community responses to water crises in three rural communities in Newfoundland and Labrador- Flat Bay, Black Tickle, Labrador and Steady

This paper studies the optimal scoring of multiple choice tests in which the marks for wrong selection and omission jointly minimize the mean square dierence between score

Patient-Reported Outcomes (PRO) are now widely used in health-related research, some of them to assess Health-Related Quality of Life (HRQL), usually via self-administered

no spécial difficulty and adduces the question, "How can a version make something far away from it?", saying this question seems to give us no concern, and concluding

Our results showed that although landscape composition explained significant variation within studies, pest and enemy abundances, predation rates, crop damage, and yields each

However, they report that average item difficulty for the generated questions was above 0.5 (i.e., considered difficult) which can be explained by our similarity theory [4] since

Our results showed that although landscape composition explained significant variation within studies, pest and enemy abundances, predation rates, crop damage, and