Study finds gender gap in academic performance is disappearing in Science Education

Swansea University researchers have discovered an apparent lack of gender bias in performance when using ‘elimination testing’ assessment methods on life science students. The findings of the new study are published online in the international journal PLOS ONE.

Dr Geertje van KeulenThe work, led by Dr Geertje van Keulen (pictured) of the University’s College of Medicine, focuses on how knowledge is assessed, specifically through the use of multiple choice questionnaires – or MCQs – which are increasingly used as the assessment method in many life science degrees.

Funded by the Higher Education Academy and with a high level of voluntary student participation, the study shows that the previously-existing performance gap between genders in science, technology, engineering, and mathematics (STEM) subjects has decreased to no significant difference.

In addition, it shows that the student experience and performance improves when changing the MCQ format to reward partial knowledge, a technique known as ‘elimination testing’.

MCQ tests can be used to objectively measure factual knowledge, but may also introduce gender bias dependent on topic, instruction, scoring, and difficulty. The ‘Single Answer’ (SA) test is the one that most of us will remember where you’re asked to choose one correct answer and are unable to demonstrate partial knowledge.

It is in these types of tests, without negative marking conditions, that female students often perform less well by displaying risk-adverse behaviour.

Elimination testing (ET) is an alternative form of MCQ, which discriminates between levels of knowledge while rewarding demonstration of partial knowledge.

The study’s result show that life science students were significantly advantaged by answering the MCQ test in elimination format compared to single answer format.

Importantly, the study found no significant difference in performance between genders in either cohort for either MCQ test under negative marking conditions.  Surveys also showed that students generally preferred ET-style MCQ testing over SA-style testing. And students reported feeling more relaxed taking ET-style MCQ tests and agreed it improved their critical thinking skills.

Dutch-born Dr Van Keulen, who joined Swansea University in 2007 from the University of Groningen in The Netherlands, said: “This is the first time a proper study has been done on the different methods of assessment and the results are very promising.

“Here in Swansea, we have a reputation for good practise in innovative teaching methods and publishers are showing a real interest in the study’s findings, so we could see the elimination method becoming much more widely-adopted.”

Dr Van Keulen will speak about the results of the study at the Higher Education Academy’s Annual Learning and Teaching STEM Conference, which will be held in Birmingham from April 17-18.  

Dr Nathan Pike, discipline lead for Biological Sciences in the Higher Education Academy, which funded the research, said:  “Practitioners within the Biological Sciences hold a variety of views on the pedagogical value of MCQs in assessment and learning.  However, I know that they would speak with one voice in acknowledging the great value of studies (like this one) which help us to improve MCQs as pedagogical tools for appropriate use in HE.”

Dr Janet De Wilde, Assistant Director and Head of STEM at the Higher Education Academy, said: “The outcome of this study is very important to the Biosciences community and I look forward to the results being presented at the HEA STEM Annual Learning and Teaching conference 2013, which will be held at the University of Birmingham in April.”

Sarah Broadley, Marketing Manager for HE Science, Oxford University Press, said: “Oxford University Press is pleased to have supported this research. We hope these findings will help to improve undergraduate teaching and learning by providing evidence for assessment methods that better reward students’ knowledge.”

The full study, entitled “Negatively-marked MCQ assessments that reward partial knowledge do not introduce gender bias yet increase student performance and satisfaction and reduce anxiety”, is available at: http://dx.plos.org/10.1371/journal.pone.0055956.