Student perceptions of predictability of examination requirements and relationship with outcomes in high-stakes tests in Ireland

Journal article

Entirely predictable examinations are ones for which the questions are known in advance. Some assessments are designed this way, but in public examinations, predictability is subtler. Students familiarise themselves with the requirements broadly: likely topics that will come up, question formats and how to maximise their marks. If students can predict what they have to do, they can memorise performances, such as essays, and restrict their learning to fit only with examination requirements. The danger is that this focus could undermine curriculum aims. Further, examinations that are overly predictable might produce results that do not generalise to other performances or have predictive validity. This paper presents part of a broader project investigating whether the Higher Level Irish Leaving Certificate Examinations were too predictable. Here, the development of a rating scale for students' views of examination predictability is described. Data were collected from 1,002 Irish Leaving Certificate students taking Higher Level examinations in biology (n=536), English (n=749) and geography (n=387). Students' views on predictability of the examination could be grouped consistently across subject areas into three factors: valuable learning, predictability and narrowing of the curriculum. Belief that narrowing of the curriculum was a good examination preparation tactic had a negative relationshop with examination scores and perceived learning value of examinations was positively associated with students' scores in biology and English. These findings indicate that the scoring system rewards students who believe they must study the discipline broadly.

Authors: Baird, J; Caro, D; Hopfenbeck, T

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: Routledge
• Journal: Irish Educational Studies
• Volume: 35
• Issue: 4
• Pages: 361-379
• Publication date: 2016-09-20
• Acceptance date: 2016-06-08
• DOI: 10.1080/03323315.2016.1227719
• EISSN: 1747-4965
• ISSN: 0332-3315
• Keywords: structural equation modelling; student perception; washback; predictability; examinations