The authors analyzed the relationship between learner traits and computerized tests scores, and then discussed strategies for putting these findings to practical use in computerized testing. They first conducted a type I quantification analysis using computerized test scores as the criterion variable sand learner traits as the explanatory variables. This analysis revealed that learners who preferred using computers to paper-and-pencil in testing, were faster at keyboard input, had an external locus of control, or took a moderate position on an extrovert-introvert dimension, displayed their abilities to the fullest during computerized testing. Next, the authors again used type I quantification analysis to predict whether each learner had made full use of his or her abilities during computerized testing. The traits of those learners who fell short of their expected scores were analyzed. The results showed that these learners tended to find it difficult to operate computers, especially at the letter-input level, were anxious about the achievement test itself, and exhibited some neurotic traits. The authors emphasized that, when computerized testing is to be conducted, it would be well to coach learners in how to use the keyboard and so on in order to permit them to become familiar with computer operations.
View full abstract