Deciding Whether There is Statistical Independence or Not?

Abstract

The study investigated how students assess statistical independence (SI). A sample of 98 students was available to the researcher. One group was probed on their understanding of statistical independence, a second group was given a proficiency test and a third group was presented a proof for statistical Independence. Interview data from the first two groups exposed student misconception on their understanding and assessment of SI. Those who judge the relation from a causal framework, tend to explore the "representativeness of the relation," if the relation is familiar, students judge it as being dependent; and if the relation is incidental, it is judged as being independent. The results indicate that student proficiency in the calculation and conceptual formalization of the joint probability and conditional probabilities are a prerequisite to the formalization of SI. The study proposes a continuous feedback to student responses through a conceptual mediated approach and use of formal algebra as a viable method to alter strongly held misconception of SI assessment.