Abstract
Examination marking is a cognitively demanding task. In England, examiners of formal, high stakes assessments are tasked with marking hundreds of scripts accurately and consistently within a short, intensive period of time. Despite assessment and marking practices having been the focus of research for many years, there is comparatively little research into the judgement and decision making processes used by examiners when marking high stakes examinations. This thesis adopted a cognitive psychological perspective to investigate the cognitive marking strategies used when marking A-level Psychology responses, with a focus on the marking of extended written responses which have been consistently shown to be the least reliably marked. The thesis consists of five empirical studies. In Study 1, a hybrid thematic analysis of interviews with senior A-level Psychology examiners (n = 5) identified that the cognitive marking strategies used when marking extended written responses were qualitatively different to those previously identified in the marking of GCSE responses. Study 2, a multi-methods study confirmed these findings in a larger sample comprised of novice (n = 30) and experienced (n = 13) markers. The participants completed a marking activity whilst ‘thinking aloud’ followed by an online questionnaire which asked them about their marking practices. Qualitative and quantitative analyses identified that there were few differences in the marking strategies used by novice and experienced markers and that marking accuracy was not associated with marking strategy usage. A model of marking was developed which was investigated further in the subsequent studies. The next two studies investigated marking processes across a three week operational examining period. In Study 3, A-level Psychology examiners (n = 53) completed online surveys which asked them about their marking at four times points across the marking period. Statistical analysis identified that whilst there was an increase in marking speed, this was not the result of a reduction in how thoroughly responses were read, but rather the result of a decreased reliance on the physical mark scheme and less re-reading of material. Interestingly few differences were identified in the marking strategies of accurate and inaccurate examiners, although marking accuracy was found to be associated with the use of an internalised marking schema. Further insight into the model of marking was gained from Study 4, in which a small sample (n = 5) of the Study 3 examiners completed a marking activity whilst having their eye-movements tracked, once at the start of the examining period and again at the end. A semi-structured interview followed the marking activity and included a cued retrospective think aloud (RTA) generated from the examiners watching a replay of their eye movements. Qualitative analysis of the data led to a revised model of marking. In Study 5, aspects of the model were validated using secondary marking accuracy data obtained from the examiners used in Study 3 (n = 53) and the associated population of A-level Psychology examiners (N = 284). The thesis concludes that marking takes place within an individual mental marking paradigm (MeMaP), the values of which are resistant to change. This suggests that ensuring examiners develop and embed a shared understanding of the mark scheme is crucial to marking accuracy.