09-02-2020, 11:34 AM
You can allow projects to have both practice and experiment trials within the same Data Source by using the Blocking Levels within the Randomization Settings. The image below shows the Data Source columns from the Picture Response example from the Experiment Builder Video Tutorial Series:
By having the 'practiceStatus' as the first blocking level, Experiment Builder separates practice and experimental trials into separate blocks. Note that to ensure that practice trials always come before experimental trials, the Randomize option is unchecked at the 'practiceStatus' blocking level and that the practice trials are above the experimental trials in Data Source. The example has a counterbalance between subjects and also implements a splitting column. In the Data Source either rows 1-10 (list 1) or 11-20 (list 2) were presented to the subjects, from these the practice trials are rows 1-2 or 11-12.
Finally, like many experiments, this experiment has fewer practice trials per block than experimental trials. It also shows an instruction screen after the practice trials and at the end of each block. This is achieved using the Split by property in the Trial sequence properties. This property determines the number of trials in each iteration of the Trial sequence before returning to the sequence above (e.g. the Block sequence). In this case, this the number of trials per block (2 trials in the practice, 4 trials in block 2 and 4 trials in block 3).
Please see either the Randomization Options in the Video Tutorials or the Webinar on Data Source Randomization to get more detail on trial and block randomization.
By having the 'practiceStatus' as the first blocking level, Experiment Builder separates practice and experimental trials into separate blocks. Note that to ensure that practice trials always come before experimental trials, the Randomize option is unchecked at the 'practiceStatus' blocking level and that the practice trials are above the experimental trials in Data Source. The example has a counterbalance between subjects and also implements a splitting column. In the Data Source either rows 1-10 (list 1) or 11-20 (list 2) were presented to the subjects, from these the practice trials are rows 1-2 or 11-12.
Finally, like many experiments, this experiment has fewer practice trials per block than experimental trials. It also shows an instruction screen after the practice trials and at the end of each block. This is achieved using the Split by property in the Trial sequence properties. This property determines the number of trials in each iteration of the Trial sequence before returning to the sequence above (e.g. the Block sequence). In this case, this the number of trials per block (2 trials in the practice, 4 trials in block 2 and 4 trials in block 3).
Please see either the Randomization Options in the Video Tutorials or the Webinar on Data Source Randomization to get more detail on trial and block randomization.