Online Parameter Estimation for Student Evaluation of Teaching

Area of Science:

• Educational Measurement
• Psychometrics
• Higher Education Research

Background:

• Student Evaluation of Teaching (SET) is crucial for assessing teaching performance.
• Current SET methods often fail to account for student rating harshness, impacting assessment validity.
• Online SET lacks robust methods for simultaneously estimating teaching proficiency and student harshness.

Purpose of the Study:

• To develop and compare novel statistical methods for SET parameter estimation.
• To address the unaddressed issue of simultaneously estimating teaching proficiency and student harshness in online SET.
• To improve the precision of parameter estimations in computerized adaptive testing for SET.

Main Methods:

• Development of three novel methods: marginal, iterative once, and hybrid approaches.
• Utilizing a simulation study to evaluate the performance of the proposed methods.
• Comparison of the novel methods against traditional scoring techniques.

Main Results:

• The hybrid method demonstrated substantial outperformance compared to traditional methods.
• The proposed methods showed improved precision in parameter estimations.
• The simulation study confirmed the efficacy of the developed techniques.

Conclusions:

• The hybrid method is a promising technique for enhancing the validity of online SET.
• Accurate estimation of teaching proficiency requires accounting for student rating harshness.
• Further research into advanced psychometric models for SET is warranted.