*************************************************************** ** readme file associated with manuscript: INSERT TITLE ** *************************************************************** *This dataset contains the information on the movement and mouse clicks of human participants inside a two-dimensional virtual environment. Please refer to the associated publication for details on the study and data analysis. *This data is fully anonymous and cannot be related to particular individuals. *All data collected and prepared by Nikolai W.F. Bode who should also be contacted with questions (nikolai.bode@cantab.net). *Two types of data are included in this dataset: (1) the raw data output from the interactive computer-based task for 464 human participants (2) processed data for the same participants that was used for Bayesian model fitting (see associated manuscript) The presentation of this data is described in more detail below. Throughout, missing data (e.g. age, gender) is denoted by 'NA'. **************** (1) raw data *this data can be found in the archive 'raw_data.zip' *the data consists of 464 plain text files, one for each human participant. *file names give the unique ID assigned to participants that was also used to determine the experimental treatment participants experienced. *file names are not consecutive from 1 to 464, because of data that could not be used (participants closed program before data was saved) and participants using different computers were distinguished by incrementing the start ID on different computers by multiples of 1001. *the first 6 rows of the files for participants give gender and age of the participant, as well the treatment the participant experienced, the on-screen direction of the path indicated by arrows in the training task, which exit was wider, if any, and which route was shorter (recall that participants could not necessarily determine which route was shorter at the start of the simulated evacuation, see treatment S). *the eight different experimental treatments are encoded by numbers, as follows (using notation of the associated manuscript): 1: S 2: control 3: M+S 4: M 5: W+S 6: W 7: W+M+S 8: W+M *the remainder of the files provide the movement and mouse clicks of human participants and the movement of simulated agents inside the two-dimensional virtual environment. *positions are given in metres in the virtual environment. *each row presents data from one simulated time-step in the virtual environment. *the first column lists the time step number, starting from zero at the start of the first task. This counter is set to zero again at the start of the simulated evacuation. In the manuscript, only data from the simulated evacuation is used. *columns 2 to 181, give the x and y coordinates for the positions of 1 human participant and 89 simulated agents inside the virtual environment. Columns 2 and 3 give the x and y coordinates for the human participant, respectively. columns 4 and 5 give the x and y coordinates for agent 1 and so forth. *columns 182 and 183 give the x and y coordinates of the last mouse click location of the human particpant inside the virtual environment. Changes in this position indicate that the participant has clicked onto a new location. *the data output stops when the human participant has reached the final exit from the virtual environment. ** The virtual environment ** *see also figure 1 in the associated manuscript. *the virtual environment is square, has a side-length of 15m and is situated between the corners (0m,0m) and (15m,15m). *the central room has a side-length of 10m. *the centres of the exits from the central room are located at (13m,2.5m) and (13m,12.5m). These coordinates refer to the top left and the bottom right hand corner on-screen, respectively. *simulated pedestrians have diameter 0.5m. **************** (2) processed data *this data can be found in the archive 'processed_data.zip' *this data consists of one plain text file called 'model_fitting_data.txt'. *this file contains processed data for all 464 human particpants. *the file has 29 columns and the first row provides short descriptive column names. *data for human participants is given in the remaining rows. All data for one participant is given in consecutive rows. There is no gap in rows between consecutive participants, but information in the columns can be used to distinguish between participants. *the first three columns give the gender of, age of and experimental treatment experienced by human participants. *the fourth column indicates whether the participant is moving towards the top exit (value 1) or not (value 0) at that moment in time. *the fifth column indicates the difference in exit width between exits (0: no difference, -1: bottom exit is wider, 1: top exit is wider) *the sixth column indicates which route to the final exit is shorter (1: bottom route is shorter, -1: top route is shorter). *column 7 and 8 give the distance between the position of the agent controlled by the human participant and the top and bottom exits out of the central room, respectively. *columns 9 to 18 give the number of simulated agents queuing at the top exit, one data output step ago, two data output steps ago, up to ten data output steps ago. Time points prior to the start of the simulated evacuation were given the same value as subsequent time points. *columns 19 to 28 provide the same information as columns 9 to 18, but for the bottom exit. *column 29 gives the number of simulated agents that block the direct path of the human controlled agent to the opposite exit. *please see associated manuscript for details on how this processed data was computed from the raw data. ****************