Millisecond Forums

Hello:

I'm using Run Monkey to check my code (as below). It consists of an introduction of the task, followed by 24 surveys, which are supposed to be displayed once, in randomized order.
However, when I checked the iqdat files under 'data', it looks like only a subset of the surveys contained in the 'noreplace' are being sampled.
I believe one iqdat file should be generated for each survey but there were only 18 of them (including one for the introduction).
I tried Run Monkey a couple of times and again only got 18 to 20 files, but never 25, which is the number I was expecting to see.

I'm worrying that this might not be an issue with Run Monkey but rather, that I'm misunderstanding how 'noreplace' works.
I want to make sure that subjects will go through all the 24 surveys (in random order). Any comments will be greatly appreciated!

Thank you very much in advance,
EN

--

<expt MAIN>

/ blocks = [1 = survey.INTRO; 2-25 = noreplace(survey.SCENARIO_01, survey.SCENARIO_03, survey.SCENARIO_05, survey.SCENARIO_08, survey.SCENARIO_10, survey.SCENARIO_11, survey.SCENARIO_13, survey.SCENARIO_14, survey.SCENARIO_16, survey.SCENARIO_17, survey.SCENARIO_18, survey.SCENARIO_19, survey.SCENARIO_21, survey.SCENARIO_22, survey.SCENARIO_26, survey.SCENARIO_27, survey.SCENARIO_28, survey.SCENARIO_29, survey.SCENARIO_30, survey.SCENARIO_32, survey.SCENARIO_33, survey.SCENARIO_35, survey.SCENARIO_36)]

</expt>

You're not misunderstanding noreplace, but your noreplace pool is set up wrongly. You have 23 -- not 24! -- survey elements in the pool, i.e. you're sampling 24 times from a pool of 23 elements, i.e. you're oversampling the pool, and since elements are to be sampled in equal proportions, each element is effectively included twice.

Adjust your pool accordingly. Either add a 24th element or sample only 23 times.