Eye-movement Dataset for Mandarin Chinese Sentence Production

This data set was collected in the eye movement room from September 21, 2020 to October 27, 2020. The eye movement device is the desktop mode of Eyelink 1000 (SR Research Ltd., Kanata, ON, Canada). The refresh frequency of the computer is 60Hz, and the distance between the subjects and the screen is 60cm. The data set contains 88 subjects' original eye movement data and experimental procedures (Original Data/ExperienceProcedure/DataAnalysis/Code).The data for analysis is stored in the data analysis folder (DataAnalysis), which contains the folder for recording reaction time (speech onset time) and the selection of sentence structure (RT_S) and the active and passive sentence folder (EYE) for eye movement time course analysis.The experiment is a picture description task. The picture contains action events between two characters, in which the sender of the action is the agent and the receiver of the action is the patient. At RT_ S folder, In the table (RT_S.csv), Trial is the target image, with a total of 32 target images (T1-T32); SubID is the serial number of the subject (1-88), PA is the animacy of the patient (N: non-human, H: human), SF is the sentence structure (A: active sentence, P: passive sentence), FF is the position of the first fixation (a: the agent interest area, p: the patient interest area), EC is the difficulty of event coding (the greater the value, the more difficult the event is), and ECOD is the difficulty of event coding. According to the median of the value, it is divided into High (high) and Low (low) groups, RT is the time from the beginning of the picture to the beginning of the first sound (sound latency), in milliseconds. The Active.csv/Passive.csv file is stored in the EYE folder. The data stored is the collected sample rate data of the active and passive sentences. In the table, Trial is the target image, with a total of 32 target images (T1-T32); ParticipantName is the subject number (1-88), Agent represents the agent interest area (TRUE indicates that the sampling point falls in this area; False indicates that the sampling point does not fall in this area), Patient represents the patient interest area (TRUE indicates that the sampling point falls in this area; False indicates that the sampling point does not fall in this area), TrackLoss indicates the missing value (FALSE indicates that the data sampling point is normal, and TRUE indicates that the sampling point is missing), TimeFromTrialOnset refers to the time of the sampling point (taking the image presentation as the starting point, intercepting the 0-2000ms time period for data analysis) SF refers to the sentence structure, which is used to mark the active sentence in this table, PA refers to the life of the patient (N: non-human, H: human), EC refers to the difficulty of event coding (the greater the value, the more difficult the event is), and ECOD refers to the difficulty of event coding, according to the median value, They are divided into High and Low groups. The data stored in Passive.csv is the eye movement sampling point data of passive sentences, and the variable is the same as Active.csv. In the Code folder, there are codes for response time and structure (RT_S. R) and codes for time process analysis of active and passive sentences (GCA_Active. R/GCA_P. R), respectively. The corresponding statistical methods are linear mixed model and growth curve analysis.