Academics inflated grades during the early stages of the pandemic, apparently reflecting a desire to help students cope with their tougher online environment, a study at a large US public university has found.

, based on data from a single medium-sized public research university, showed significant gains in grades during the second half of the spring 2020 semester after most campuses shut their physical operations.

¡°During the Covid semester, the universities encouraged instructors to be more accommodating,¡± said one study author, Duha Tore Altindag, associate professor of economics at Auburn University.

The data ¨C unique in?their breadth across the entire institution and multiple semesters ¨C also cast wider doubt on the value of online education stretching beyond the effects of the pandemic.

É«ºÐÖ±²¥

That¡¯s because the unidentified institution has a large and long-standing online option for courses, and its data showed that its students had been already getting higher grades while faring worse on longer-term outcomes such as course completion.

That should serve as a caution for any institutions that may be tempted post-pandemic to make any aggressive permanent shifts into online operations, Dr Altindag said. ¡°Online is cheaper, but this comes at a cost,¡± he said.

É«ºÐÖ±²¥

Dr Altindag¡¯s co-authors were Erdal Tekin, a professor of public administration and policy at American University, and Elif Filiz, an assistant professor of economics at the University of Southern Mississippi.

Their partner institution, with about 15,000 students, provided them with grading data for the spring 2019, autumn 2019 and spring 2020 semesters. It included mid-term grades for spring 2020, reflecting student progress just ahead of their movement into online formats.

Their study did not identify causes of grade inflation after the online transition, although it was well known that institutions urged leniency at that time, Dr Altindag said.

The study, posted by the National Bureau of Economic Research ahead of formal peer review, did check for the possibility that cheating played a substantial role in grading increases but found no evidence of that, he said.

É«ºÐÖ±²¥

The authors acknowledged that numerous previous studies have cast doubt on the overall value of online instruction. They argued their database gave a uniquely valuable affirmation of that concern by assessing the phenomenon across an entire institution, rather than individual courses, with a robust set of comparison courses in existing online formats. Students and faculty also were unaffected by any awareness of the study.

The analysis made an attempt to quantify the importance of technical factors, by comparing student performance?with the average speed of internet connections in their neighbourhoods. That showed only a modest advantage, about 0.02 per cent of a grade for those living in the better half of communities as ranked by connection speeds.

The authors attempted to factor out income-related variations, but conceded that lower-income students forced into online situations likely contended with many handicaps beyond internet speed.

paul.basken@timeshighereducation.com