The relationship between gaze and decision preference has been widely concerned by researchers in recent years.In order to explore how the formation process of decision-making preference is reflected in fixation trajectory and whether the fixation information from individuals can be used to change decision-making preference.In experiment 1,We asked the subjects to freely observe two pictures containing dots,and to try to determine which one of the two pictures has the higher density of dots,and then press the left and right buttons of the mouse to select.During the experiment,we used Tobii TX120 eye tracker to record the eye movement information of the subjects.Gaze likelihood analyses were conducted following the procedures by Shimojo et al.(2003)and Zommara et al.(2018).Gaze likelihood analyses refer to the calculation of the proportion of subjects’ fixation on the final selected picture in all the trials before the decision.The results showed that the subjects’ gaze was initially torn between two choices,but then gradually shifted to the image they were going to choose.We fitted the raw data points from all tasks to sigmoid curves.The R2 values are all above 0.9,indicating good fit.Therefore,we also found gaze bias effect in simple feature judgment tasks.The gaze bias effect of possible is larger than that of impossible,which supports the Two-Stage Theory of Decision Information Processing proposed by Glaholt and Reingold(2011).In experiment 2,the gaze-contingent response prompt was used to study the effect of gaze on decision preference.Compared with the exogenous manipulations method used in previous studies,our experiment controlled the influence of demand effects on experimental results.We recorded gaze information,and tried to use this information to systematically bias the selection of subjects to randomly determined target options.Specifically,the subjects could freely observe two pictures presented on the screen with internal dots.Before the trial,one of the pictures was randomly selected as the target option.When the target option was cumulative gaze for 750 ms,and the non-target option was cumulative gaze for 250 ms,the subjects were prompted to make a choice.The key dependent measure is the proportion of trials on which the target is chosen.The target was randomly determined on each trial,so any bias toward the target manifests as a deviation from 50% selection.The experimental results show that the subjects are more inclined to choose the target option when the stimulus is undistinguishable.This suggests that an interruption to the naturally evolving pattern of attention does have an impact on decision-making preference.More surprisingly,our study found that the gaze-contingent response prompt did not make significant difference in the cumulative gaze time of subjects on the target option and non-target option,but made the target option be finally looked at in most cases.In experiment 3 we also used the gaze-contingent response prompt to test whether controlled subjects’ fixation could influence their moral decision preference.The experimental procedure of Experiment 3 is the same as Experiment 2,except that moral questions are used as experimental materials.The program presented ethical questions through headphones,then presented two options simultaneously on the screen,and asked participants to choose when the options disappeared.It was found that the proportion of subjects choosing target options remained around 50%.The results show that manipulating gaze did not affect the moral decision preferences of the participants,but the results found that the difficulty of the moral decision seems to also can affect the experimental results.In general,the results show that gaze is closely related to decision making.Gaze can reflect the forming process of decision preference,and it can also influence decision preference by controlling subjects’ gaze.Both the response of gaze to decision preference and the influence of gaze on decision preference are regulated by the difficulty of decision task.The research results support Attention Drift Diffusion Model and the Two-Stage Theory of Decision Information Processing. |