About the Project
Many empirical studies (Rayner, 2009) and several sophisticated computational models (Reichle, Pollatsek, & Rayner, 2003) now provide excellent accounts of the mechanisms involved in word recognition and eye movement control during reading. It has long been established that text comprehension, especially for detail within text, is poorer during skimming compared with reading for comprehension (Just & Carpenter, 1987). However little research has been undertaken to examine how the processes underlying reading are modulated by reading goals, such as skimming text quickly for gist or relevant content (White, Warrington, McGowan, & Paterson, 2015). Research into reading goals has potentially very important theoretical implications for identifying flexibility in the nature and co-ordination of the systems involved in reading. The proposed project will especially focus on how reading goals modulate how text and prior knowledge are integrated during reading.
Methods: The experiments will be undertaken in the established eye movement laboratory within the Department of Neuroscience, Psychology and Behaviour. Recording participants’ eye movements as they read provides a unique insight into what is processed when. Eye movements will be recorded with an EyeLink 1000 eye tracker, which has excellent temporal and spatial resolution. Both of the supervisors have a strong track record of high quality research in this area. The experimental procedure involves participants reading sentences on a computer screen and responding to comprehension questions. Reading goals will be modulated by the frequency and difficulty of comprehension questions as well as participant instruction. Analyses will be undertaken using a suite of eye tracking analysis software and R (for example Linear Mixed Effects models).
Research questions: The aim of the project is to examine the effect of reading goals on eye movement behaviour, and specifically how reading goals modulate integration within sentences and with prior knowledge. A number of research questions could be addressed as part of the PhD, either examining broadly a range of different types of integration (e.g. syntax, context, plausibility, prior knowledge), or instead focusing on a particular aspect, such as the integration of text with existing knowledge. Recent work in our laboratory indicates that prior knowledge is integrated to a lesser extent during skim reading. It would be especially valuable to examine how reading goals modulate the acquisition (and subsequent integration) of knowledge during reading. Crucially the manipulations will be carefully designed such that the detailed eye movement records will provide a unique insight into the level and time course of text processing. For example, by manipulating the veracity of test stimuli, the extent and time course of the influence of text veracity on eye movement behaviour will indicate to what extent text is integrated with prior knowledge. For example, during more careful reading, contradictions of the text with prior knowledge have been shown to modulate initial word reading and re-reading times (Rayner, Warren, Juhasz, & Liversedge, 2004). Reading goals are likely to modulate both the size and the time course of such effects, providing insights into the effect of reading goals on the depth and time frame of integration.
Revealing the nature of text processing during skimming has especially important implications for applied situations for which accurate comprehension is critical. For example, in learning contexts integrating prior knowledge with text comprehension is crucial, with potential implications for educational and other settings, such as patients comprehending medical information leaflets (for which errors are known to occur, Keselman & Arnott-Smith, 2012). Following a recent publication about medical errors (Makary & Daniel, 2016), the Director of the Harvard Global Health Institute (Dr. Ashish Jha) suggested in a Scientific American Guest Blog (2/6/2016) that rapid reading of complex medical documents may be a key contributory factor to medical errors: “I see a patient with literally 50 previous hospitalizations. It would take me days to carefully go through all of those records, so I skim, praying that I don’t miss anything important.” Understanding more about the effect of reading goals on the mechanisms underlying reading will provide a stronger basis both for educating readers about rapid reading and for providing robust critiques of new rapid text display technologies.