An Australian university has been given the green light to publish student course evaluation data, despite concerns that the information can be used to ¡°reverse engineer¡± staff league tables.

UNSW Sydney has successfully appealed a decision from Australia¡¯s industrial relations umpire, the Fair Work Commission (FWC), banning it from giving students broad access to data compiled from quantitative course ratings.

The information includes students¡¯ appraisals of course quality, resources, assessment and feedback on a six-point scale. UNSW says it does not intend to publish students¡¯ qualitative commentary about the courses or their evaluations of teachers.

The university only plans to publish quantitative data on courses that involve more than one teacher ¨C so that individual staff cannot be directly identified ¨C and where at least 10 responses have been received, so that aberrant assessments cannot unduly influence the data.

³ÉÈËVRÊÓƵ

But the National Tertiary Education Union (NTEU) said that the information allowed individual staff to be identified by academics in the same discipline or school, and ¨C in cases where courses were taught by pairs of tenured and casual staff ¨C by academics across the university.

¡°This is a public university, not a reality TV show,¡± the union¡¯s New South Wales branch tweeted on the eve of a FWC hearing in August.

³ÉÈËVRÊÓƵ

FWC commissioner Leigh Johns sided with the union. He dismissed the university¡¯s argument that the data¡¯s publication did not directly identify staff because ¡°detective work and additional effort¡± was needed to match courses to teachers.

But an FWC appeals panel has??that decision, finding that the university¡¯s plans comply with a strict reading of the enterprise agreement.

¡°We accept that in some cases a person can be identified without being named,¡± the 7 March judgment says. ¡°A reference to [an] attribute could identify the person, even though the person¡¯s name might not be used. But the proposed form of the data to be published simply does not do this. No staff are identified.¡±

UNSW believes that publishing the data will convince students that their feedback is being ¡°considered and utilised¡±, encouraging more of them to fill out course evaluation surveys and improving the questionnaires¡¯ response rates and accuracy. ¡°Our students¡­take such care in providing all this feedback,¡± said deputy vice-chancellor Merlin Crossley.

³ÉÈËVRÊÓƵ

¡°This gives us an opportunity to celebrate and showcase many courses that provide an outstanding student experience, and work with staff to develop courses that still need work.¡±

But NTEU state secretary Damien Cahill said that the ¡°very disappointing¡± ruling potentially sent a signal to managers at other universities who ¡°want to use these surveys as performance metrics¡±.

Dr Cahill said student evaluations were useful for teacher and course development when they were ¡°framed and conducted¡± properly. ¡°But there¡¯s a lot of scholarly literature on¡­the inherent gender and racial biases [and] limitations based on the context. Whether a course was online, the time of day, the room in which it was held, the dynamics of the cohort of students ¨C all of those situational factors play a role in determining the outcome of those evaluations.¡±

In the , published in the?Journal of University Teaching & Learning Practice, a Victoria University analysis of more than 22,000 students¡¯ teaching evaluations found no differences in the score ratings for teacher gender. But male students¡¯ commentary about the teaching style of female academics became increasingly negative during Covid-related lockdowns.

³ÉÈËVRÊÓƵ

john.ross@timeshighereducation.com