Blind Mice: Mouse Movements as an Alternative to Eye Tracking

Abstract

The eye tracking while reading paradigm is an excellent tool for answering psycholinguistic research questions, and is generally seen to offer the highest level of precision and accuracy among all self paced incremental processing paradigms. However, eye tracking experiments are costly in terms of time and money, and require specialised equipment. A novel incremental processing paradigm developed by Wilcox et al. (2024) which has proven capable of replicating results from eye tracking replaces eye gaze during reading with mouse movements. This method, which they call ‘mouse tracking for reading’ (MoTR), allows experiments to be implemented fully online, cutting down on costs. This thesis uses MoTR to replicate results from an eye tracking study which found slowdown effects during reading of English sentences caused by semantic and syntactic retrieval interference. Findings of retrieval interference are informative for models of memory access during language use. The eye tracking study, conducted by Mertzen et al. (2023), used longer and more complex sentences than those tested in the original MoTR experiment. Furthermore, Wilcox et al. focused only on syntax related phenomena, whereas our research examined syntactic and semantic effects simultaneously. The aim of the present research was thus twofold: extend what we know about memory access and language, and extend what we know about the effectiveness, strengths, and weaknesses of MoTR. The MoTR study used differed slightly from that of Wilcox et al., allowing this thesis to bring novel insights about mouse tracking. Notably, participants were allowed to use trackpads. Trackpad usage led to higher reading times, but there was otherwise no qualitative difference in reading behaviour between trackpad and regular mouse users. MoTR succeeded in replicating both kinds of retrieval interference effects. The results strengthen accounts of interference based on feature match, such as the cue-based retrieval theory. They show that mouse tracking is capable of capturing multiple complex effects on reading times, and suggest that MoTR can be a useful supplement to or even replacement for eye tracking.