Programming a study in ExperimentBuilder
Starting from the ground up
Last updated
Starting from the ground up
Last updated
Whenever you start programming an experiment, there are two components that you need to worry about: your Experiment Flow, and your Datasource
The first is the flow of your experiment. How long do you want the images to be onscreen before the audio starts? How long should participants see the images after the target word onset? Where onscreen should the images be? All of these are questions that you will answer with how you set up your Experiment Flow.
Experiment flow is determined through a point-and-click interface in ExperimentBuilder. When you open the program, the flow is the first thing you see. It is composed of Actions, which determine what your participant sees or what is being recorded, and Triggers, which determine when a specific action begins or ends.
Actions include "begin trial", "display image", or "play audio".
Triggers include "key press" "timer" or "gaze in interest area".
By combining Actions and Triggers, you can build the flow you want for each trial, and also for your experiment in total, by nesting components inside your larger Experiment flow.
The datasource is where you indicate exactly what you want on the screen during each trial. It also lets you provide multiple orders for your experiment, lets you tag fillers vs. experimental trials, and lets you note anything you want about the details of each individual trial.
Whereas the Experiment Flow looks like a flowchart, the datasource is simply a spreadsheet akin to what you would make on Google Sheets or Excel. The format you choose for your datasource is very important and must be machine-readable. A good datasource will follow the principles of : each row is one trial, each column is a variable, and each cell contains only one value.