**Repository for White, Palmer & Boynton (2019)**
Alex L. White
University of Washington
2019
This repository contains stimulus materials, raw data, analysis code, and model code for the study reported in White, Palmer & Boynton's 2019 article: "Visual word recognition: Evidence for a serial bottleneck in lexical access."
Here we uploaded a single .zip file that, when opened, contains various folders:
1. Stimuli: this contains two csv files with the stimulus set used in Experiment 1, and in Experiments 2 and 3. Each csv file has 2 columns, one for each category of stimuli, as described by the header row.
2. Data. This contains raw trial-level data for all three experiments. The subfolder "indiv" contains one folder for each experiment, X1, X2 and X3. Within each of those folders is a csv file with demographic information about each subject (eg, Expt1Subjects.csv). Demographic information includes their scores on the TOWRE reading efficiency tests.
Individual subject data files are tab-delimited .txt files, with one row for each response and many columns that describe what happened on those trials. Note that each single-task *trial* has 1 row (because the subject made just 1 response), but each dual-task trial has two rows.
After running the analysis code, results files will be saved in the Data folder (e.g., Expt1_MainRes.mat).
3. AnalysisCode: Two subfolders of MATLAB code:
AnalyzeData: the Matlab files here process the raw data files and store results in .mat files in the Data folder. There is one script per experiment: Expt1Analysis.m, Expt2Analysis.m, and Expt3Analysis.m. Running each of those will do all the analyses. Each calls various functions to analyze each subject (e.g., Expt1_AnalyzeSubject.m), and to analyze subsets of trials for each subject (e.g., AnalyzeTrials.m).
FiguresAndStats: the Matlab files here create all the figures in the manuscript and print out statistics. This is accomplished by running a single script: MakeFiguresAndStats.m.
Those functions load in results files from the Data folder, print figures as .eps files to the Figures folder, and statistics as .txt files to the Stats folder.
4. Model: this folder contains Matlab files that support the computational modeling of fixed-capacity parallel processing and all-or-none serial processing in our dual tasks. Several functions are used to generate the Attention Operating Characteristics, plot the model predictions on them, and compare the data to those predictions (e.g., plotAOCWithPredictions.m; compareToFixedCapacityModel.m).
Another set of functions are used to generate the predictions for the stimulus processing tradeoffs: SerialModelTradeoffSimulation.m, which calls SerialDualTaskModel.m. The resulting prediction is saved as SimulatedAccuracyTradeoff_AllOrNoneSerial.mat.
5. Figures: after running MakeFiguresAndStats, eps files will be saved here.
6. Stats: after running MakeFiguresAndStats, txt files will be saved here.
7. Utils: this folder contains various Matlab functions that are needed for the rest of the code. All were written by Alex White except:
- rm_anova2.m, by Aaron Schurger, from the Matlab file exchange;
- exportfig, by Ben Hinkle, from the Matlab file exchange;
- boyntonBootstrap, by Geoff Boynton and edited by Alex White.
All the code was developed and tested with Matlab 2018a, using their statistics toolbox.
Note: Version 2 of the zip file contains two small edits to the figure generation code, which were applied when revising the manuscript for publication in *Attention, Perception & Psychophysics*. V2 was uploaded on December 6, 2019.
