09-02-2020, 09:26 AM
This error occurs when using the Prompt for Data Set feature if the external text file you load contains trials that do not exist in the project's main Data Source.
When you use an external data set file, it acts as a filter for the "master" Data Source that is saved within your Experiment Builder project.
Every trial (i.e., every row) in your external data set file must also be present in the project's main Data Source. You cannot introduce new or different trials through an external file; you can only select a subset of the trials that already exist in the project.
Alternatively, depending on the structure of your task, you may also want to use the Split by property within the TRIAL sequence itself to limit the number of trials performed. This method does not use external data sets.
Instead, it divides the trials from the "master" Data Source into smaller blocks. For example, if your main Data Source contains 12 trials, you could set the Split by property to [4] to run only the first 4 of those 12 trials. This is useful for controlling the number of trials run from the main data list without creating separate files.
Use the Randomizer Tool
The easiest way to create valid data set files is to use the Randomizer application that comes with Experiment Builder. This tool is designed to take your "master" Data Source and generate randomized subset files that are guaranteed to be compatible.
You can find the Randomizer tool here:
When you use an external data set file, it acts as a filter for the "master" Data Source that is saved within your Experiment Builder project.
Every trial (i.e., every row) in your external data set file must also be present in the project's main Data Source. You cannot introduce new or different trials through an external file; you can only select a subset of the trials that already exist in the project.
Alternatively, depending on the structure of your task, you may also want to use the Split by property within the TRIAL sequence itself to limit the number of trials performed. This method does not use external data sets.
Instead, it divides the trials from the "master" Data Source into smaller blocks. For example, if your main Data Source contains 12 trials, you could set the Split by property to [4] to run only the first 4 of those 12 trials. This is useful for controlling the number of trials run from the main data list without creating separate files.
Use the Randomizer Tool
The easiest way to create valid data set files is to use the Randomizer application that comes with Experiment Builder. This tool is designed to take your "master" Data Source and generate randomized subset files that are guaranteed to be compatible.
- Windows: Start Menu -> All Programs -> SR Research -> Experiment Builder -> Randomizer
- macOS: Applications -> ExperimentBuilder -> Randomizer