Answers are grounded in this paper and cite their sources.
Qualitative case study explores students' understanding of unbiasedness in maximum likelihood estimation, suggesting instructional improvements.
The maximum likelihood estimation (MLE) provides a foundational framework for statistical inference and is often prized for its theoretical elegance and asymptotic efficiency. However, in small-sample contexts, MLE frequently lacks unbiasedness and this creates a significant gap between theory and practice. Despite its widespread application, mathematics and statistics education research has not yet sufficiently explored how students mentally construct their understanding of the concept of unbiasedness of the MLE. This study therefore used the qualitative case study design and the APOS Theory to explore some undergraduate statistics students’ mental construction of the concept of unbiasedness of the MLE. We conveniently sample eight students to participate in the study. Data collection procedures involved test and semi-structure interviews. We analysed both sets of data concurrently for triangulation using the genetic decomposition constructed for the test items. The results showed that, while most of the participants demonstrated an Action conception of the unbiasedness of the MLE, they had not interiorised these Actions into a Process, mainly as a result of failing to conceptualise repeated sampling and confusing a single estimate with the expected value of the estimator. It is recommended from the study that in order to foster a robust conceptual understanding of the concept of unbiasedness, instructional approaches should prioritise linking mathematical construction to practical application. Steps must also be taken to ensure that students first achieve an object conception of the fundamental summation principles to prevent procedural errors during derivation.
No takes yet. Share an insight, caveat, or question.
Soku et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: