Prof. George Engelhard Delivers a Lecture on Invariant Measurement at FoE UIII
January 13, 2024
Contributor: Maroof Ahmed | Editor: Supriyono
UIII.AC.ID, DEPOK – In a notable academic exchange bridging the United States and Indonesia, the Faculty of Education (FoE) at Universitas Islam Internasional Indonesia (UIII) recently hosted an online guest lecture featuring Prof. George Engelhard from the University of Georgia, who delivered a detailed lecture on Invariant Rater-Mediated Measurement, highlighting the critical importance of unbiased and objective assessment methods in educational settings.
During the lecture, Prof. Engelhard stressed the necessity of maintaining objectivity in educational assessments. He referenced the work of Prof. Ben Wright, asserting, “The calibration of measuring instruments must be independent of those objects that happen to be used for the calibration. The measurement of objects must be independent of the instrument that happens to be used for measuring.”
This principle underpins the need for unbiased measurement approaches in educational assessments, ensuring that the results are not skewed by the instruments or the objects used in the calibration process.
Further exploring the concept of invariant measurement, Prof. Engelhard detailed the three key facets necessary for achieving unbiased assessments: sample-invariant calibration of items, item-invariant measurement of persons, and rater-invariant measurement.
“The measurement of persons must be independent of the particular raters that happen to be used for measuring,” he emphasized, highlighting the need for assessments that remain consistent and fair regardless of the raters involved.
A significant portion of the lecture was dedicated to the methods of achieving high model-data fit, which is crucial for obtaining invariant estimates of person competence across different subsets of raters. Prof. Engelhard explained, “If the degree of model-data fit is high, then invariant estimates of person competence can be obtained across different subsets of raters. Statistical adjustments for rater severity can only be justified when model-data fit is good.” He elaborated that these invariant estimates are instrumental in improving the objectivity and fairness of the overall assessment system.
In concluding his lecture, Prof. Engelhard discussed the importance of achieving good model-data fit for invariant calibration of rater-mediated assessment and rater-invariant measurement. He pointed out that various statistical indices and graphical displays, based on the Facets model, are crucial tools for exploring sources of misfits and identifying potential problems in educational assessments.
The session facilitated by UIII provided a platform for meaningful academic dialogue and exchange in the field of educational measurement, significantly contributing to the global conversation on creating fair and objective assessment methods. Prof. Engelhard’s expertise and in-depth analysis of invariant measurement offered valuable insights to educators and researchers at UIII, underscoring the importance of unbiased and equitable assessment practices in education .