Randomization Issue

dng36 — Tue, 15 Apr 2025 19:13:36 GMT

Hi,
We have 8 experimental conditions, and participants are randomly assigned to each condition. This does seem to work, but while we do want it to be random, we also want fairly equal numbers in each condition for analysis, and right now it seems a little off. Is there a way to randomize but in equal amounts to each condition? The code we're using is below:

UPfp trial - UP, famous person
<expt 1>
/ blocks = [*blocks*]
/ groups = (1 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete = 1]
</expt>

PUfp trial - PU, famous person
<expt 2>
/ blocks = [*blocks*]
/ groups = (2 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete =1]
</expt>

UPobj trial - UP, object
<expt 3>
/ blocks = [*blocks*]
/ groups = (3 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete =1]
</expt>

PUobj trial - PU, object
<expt 4>
/ blocks = [*blocks*]
/ groups = (4 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete =1]
</expt>

UPfp trial - UP, famous person
<expt 5>
/ blocks = [*blocks*]
/ groups = (5 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete = 1]
</expt>

PUfp trial - PU, famous person
<expt 6>
/ blocks = [*blocks*]
/ groups = (6 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete = 1]
</expt>

UPobj trial - UP, object
<expt 7>
/ blocks = [*blocks*]
/ groups = (7 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete = 1]
</expt>

PUobj trial - PU, object
<expt 8>
/ blocks = [*blocks*]
/ groups = (8 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete = 1]
</expt>

This is the frequency analysis for each experimental condition:

2 and 6 being low and 4 being high really offsets the amount in the obj vs. fp condition.

RE: Randomization Issue

Dave — Tue, 15 Apr 2025 19:13:36 GMT

+x[quote]

[b]dng36 - 4/15/2025[/b]
Hi,
We have 8 experimental conditions, and participants are randomly assigned to each condition. This does seem to work, but while we do want it to be random, we also want fairly equal numbers in each condition for analysis, and right now it seems a little off. Is there a way to randomize but in equal amounts to each condition? The code we're using is below:

UPfp trial - UP, famous person
<expt 1>
/ blocks = [*blocks*]
/ groups = (1 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete = 1]
</expt>

PUfp trial - PU, famous person
<expt 2>
/ blocks = [*blocks*]
/ groups = (2 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete =1]
</expt>

UPobj trial - UP, object
<expt 3>
/ blocks = [*blocks*]
/ groups = (3 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete =1]
</expt>

PUobj trial - PU, object
<expt 4>
/ blocks = [*blocks*]
/ groups = (4 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete =1]
</expt>

UPfp trial - UP, famous person
<expt 5>
/ blocks = [*blocks*]
/ groups = (5 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete = 1]
</expt>

PUfp trial - PU, famous person
<expt 6>
/ blocks = [*blocks*]
/ groups = (6 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete = 1]
</expt>

UPobj trial - UP, object
<expt 7>
/ blocks = [*blocks*]
/ groups = (7 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete = 1]
</expt>

PUobj trial - PU, object
<expt 8>
/ blocks = [*blocks*]
/ groups = (8 of 8)
/ postinstructions = (end)
/ quit = [values.quit ==1]
/ onexptbegin = [
values.pre = list.baseline_primes.nextindex;
values.post = list.baseline_primes.nextindex; ]
/ onexptend = [values.complete = 1]
</expt>

This is the frequency analysis for each experimental condition:

2 and 6 being low and 4 being high really offsets the amount in the obj vs. fp condition.
[/quote]

You're assigning to conditions by group number. You have provided no information what the group number generation settings for your web experiment are. The latter is the relevant information, not your <expt> elements.

Millisecond Forums » Millisecond Forums » Inquisit 6 » Randomization Issue

Randomization Issue

RE: Randomization Issue