Creating Blocks

The next step is to define the different kinds of blocks that will be used in the experiment. For this experiment, we will define 2 blocks, 1 to run the practice trials and 1 for data collection.

First, let's define the practice block element.

<block covertattentionpractice>
/ trials = [1-4 = noreplace(topleftpractice, bottomleftpractice, topleftpractice, bottomrightpractice)]
/ bgstim = (fixation)
</block>

This block element, named covertattentionpractice, will run just 4 trials randomly selected from the 4 practice trial types. We've specified that the block should randomly select from the 4 trials without replacement. Since we are only running 4 trials and there are 4 to select from, that means that each type of trial will be run exactly one time in the block, and the order in which the 4 types are selected will randomly vary.

Now, lets define the data collection block:

<block covertattention>
/ preinstructions = (begin)
/ trials = [1-20 = noreplace(topleftcongruent, bottomleftcongruent, topleftcongruent, bottomrightcongruent, 
                             topleftcongruent, bottomleftcongruent, topleftcongruent, bottomrightcongruent, 
                             topleftcongruent, bottomleftcongruent, topleftcongruent, bottomrightcongruent, 
                             topleftcongruent, bottomleftcongruent, topleftcongruent, bottomrightcongruent, 
                             topleftincongruent, bottomleftincongruent, topleftincongruent, bottomrightincongruent)]
/ bgstim = (fixation)
/ postinstructions = (performance)
</block>

The block begins by presenting the instruction page named begin as specified by the preinstructions command. It then runs a total of 20 trials randomly selected from the list of data collection trials. Finally, after all the trials have been run it presents an instruction page called performance.

You may have noticed that the congruent trial types appear 4 times each in the trial list, whereas incongruent trials appear only once. Why did we repeat some of the trials? The answer is that we want 80% of the trials in this block to be congruent and remaining 20% incongruent, so we've created a random selection pool where the proportion of congruent to incongrent trials is 4 to 1. Since we are selecting without replacement, we are guaranteed that the proportion of selected trials will match the proportions in the selection pool. Of the 20 trials, a randomly selected 16 will be congruent and 4 incongruent.

That does it for the blocks. Now let's define the experiment.


Creating Trials Creating an Expt