PS1-15 ANALYSIS OF EYE GAZE PATTERNS AND DECISION-MAKING STRATEGIES DURING A DISCRETE CHOICE EXPERIMENT

Sunday, June 12, 2016
Exhibition Space (30 Euston Square)
Poster Board # PS1-15

Sian Martin1, Kielan Yarrow1, Kate Mandeville2, Stian Reimers1, Mylene Lagarde2 and Benjamin Cooper1, (1)City University London, London, United Kingdom, (2)London School of Hygiene and Tropical Medicine, London, United Kingdom

Purpose:

  To examine eye gaze patterns and decision-making strategies of discrete choice experiment participants

Method(s):

   21 participants took part in a discrete choice experiment assessing preferences for different types of primary care appointments. An efficient design was used to construct 25 binary generic choice tasks with four attributes. The eye movements of participants were tracked during the completion of choice tasks using a specialised videocamera. We reconstructed the pattern of each participant's gaze for each choice task using Matlab. Qualitative examination of gaze patterns for each participant allowed us to infer common decision-making patterns.

Result(s):

   No participant displayed a consistent eye gaze pattern over all 25 choice tasks, with all participants using a mixture of decision-making strategies. Three participants predominantly used a “within-alternative” assessment of levels before making their choice (Figure 1A). Five participants predominantly assessed levels on one or two attributes between alternatives and then the remaining levels in one alternative (Figure 1B). Two participants switched between these two strategies approximately halfway through the choice tasks. The majority of participants had no clear strategy, with gaze patterns changing frequently between choice tasks. 

Figure 1

Notes: Coloured line indicates eye gaze pattern from a representative choice task for one participant. Dark blue indicates start of participant's eye gaze, evolving to yellow at the end of gaze. A. A typical “within-alternative” stategy. B. A typical strategy assessing levels on one or two attributes between alternatives and then the remaining levels in one alternative.

   Conclusion(s):

Participants employ a mixture of decision-making strategies during discrete choice experiments. Further use of eye-tracking could help identify common patterns to help improve econometric modelling.