A longitudinal study of database-assisted problem solving

The current study examines the effects of database assistance on clinical p roblem solving across three cohorts of students and two database interfaces . Medical students attempted to solve problems presented as clinical scenar ios, first from their personal knowledge and then, on a second pass, assist ed by one of two versions of a bacteriology database. One version had a for m-filling interface in which each term was typed into the form; the other h ad a direct manipulation interface that allowed students to select terms wi th a mouse. Data were collected on three assessment occasions: just before the bacteriology course, just after the course, and six months after the co urse, Personal Knowledge Scores varied by occasion, being highest just afte r and lowest just before the course, The Database-Assisted Scores were simi lar just before and after the course, but were higher six months after the course. The differences between the interfaces did not have any statistical ly significant effect on database-assisted performance, There was a margina lly significant interaction between interface style and assessment occasion , suggesting that performance with the direct manipulation interface was le ss affected by variation in the students' personal knowledge than performan ce with the form-filling interface. (C) 2000 Elsevier Science Ltd. All righ ts reserved.