Preference for textual information and acting on support devices in multiple representations in a computer based learning environment for statistics

Two studies researched the relevance for knowledge construction in statistics of learners preference for representational system and the mediating role of learners acting with information delivered in textual, graphical and mathematical-symbolic mode.The data were collected in a broader experimental research setting investigating the effectiveness of a computer based independent learning environment for statistics. Preference for representational system, achievement in statistics, self-efficacy in mathematics and attitude towards computers were captured for 211 students enrolled in the first-year statistics course at the department of Psychology and Educational Sciences at the University of Ghent in Belgium. In the second study learners acting on devices in multiple representations was investigated in an aselect sample of 60 students, drawn from the 211 learners. These learners were assigned to the computer based independent learning condition where acting on devices was registered in log-files.The study reveals that preference is related to performance in statistics, self-efficacy in mathematics and to the behavioral and cognitive component of attitude towards computers.The findings suggest the existence of a compensation mechanism, induced by acting on graphs demanding high cognitive processing.The overwhelming majority, 3 against 1, of learners with textual preference raises challenges for instructional designers of service statistics courses. Preference is an important aspect of individual differences to be included in the research concerning cognitive style and self-regulation.