This directory contains the estimation results of four experiments, raw data of four experiments, matlab code of data generation, review results of three experiments and supp.pdf. exp1.csv, exp2.csv and exp3.csv correspond to raw participant data for the three experiments presented in our paper respectively, plus an additional experiment on mark size. exp1.csv: Id: id of current participant Age: age of current participant Std: residual bandwidth of current trial Slope: slope of current trial OLS_Fit: ordinary least squares fit result of current trial isVali: '1' means this is the validation trail, '0' means this ism't the validation trial Mark: mark type of current trial UserDe: the orientation of participant's estimation in degrees AbsErrorOri: absolute error of orientation for current trial and current participant SignedErrorOri: signed error of orientation for current trial and current participant AbsErrorOri: absolute error of intercept for current trial and current participant SignedErrorOri: signed error of intercept for current trial and current participant Times: response time of current trial and current participant exp2.csv: Id: id of current participant Age: age of current participant isVali: '1' means this is the validation trail, '0' means this is't the validation trial Std: residual bandwidth of current trial Slope: slope of current trial OLS_Fit: ordinary least squares fit result of current trial Location: location of outliers ('null' means no outliers, 'start' means outliers are at the very beginning of data, 'oneThird' means outliers are at first third of data, 'end' means the outliers are at the end of the data) Number: numbers of outliers for current trial Distance: calculated distance between outliers and main cluster for current trial ('0' means no outliers. in the raw data, we user 'n' to denote the outliers are near, 'm' to denote the distance of outliers are meadium, 'f' to denote the outliers are far.) Density: the density difference of outliers('null' means no outliers, 'd' meas the outliers are denser than main cluster, 'm' means the density of outliers are similar to main cluster, 's' means the outliers are sparser than main cluster) Sign: '0' denote no outliers, '1' denote the outliers are on the top of the main cluster, '-1' denote the outliers are on the bottom of main cluster Mark: mark of current trial UserDe: the orientation of participant's estimation in degrees AbsErrorOri: absolute error of orientation for current trial and current participant SignedErrorOri: signed error of orientation for current trial and current participant AbsErrorOri: absolute error of intercept for current trial and current participant SignedErrorOri: signed error of intercept for current trial and current participant Times: response time of current trial and current participant exp3.csv: Id: id of current participant Age: age of current participant B_Slope: slope of cluster B in current trial B_Num: number of points in cluster B in current trial isVali: '1' means this is the validation trail, '0' means this is't the validation trial Mark: mark of current trial UserOri: the orientation of participant's estimation in degrees UserInter: the intercept of participant's estimation Time: response time of current trial and current participant exp1.m,exp2.m,exp3.m are matlab code for data generation for the corresponding experiments. exp1ana.m,exp2ana.m,exp3ana.m are matlab code for figures generation for the corresponding experiments (the paper has been updated, you can download at [Ideas-Laboratory][1]). review_1.csv,review_2.csv,review_3.csv correpond to the post-experiment interview results for each experiment (detailed analysis is in the supp.pdf and paper). The Stimuli folder contains zip files of csv of the actual points used for simuli generation for each experiment. supp.pdf contains screenshots of the experimental interface of experiments and additional experimental results and discussion removed for reasons of space from the main paper. [1]: https://ideaslab.wang/data-driven-mark-orientation-for-trend-estimation-in-scatterplots