Random Assignment of Participants - by: Joffm

Hi, kporter0,
first: what did you expect, if you have such a small sample.
I had a look at a statistical table to see if there is really a significant difference.
In the binomial test the interval of convidence for 20 in a sample of 120 is: 10.5 - 24.6 (5% convidence.)
So all your group sizes are inside of the interval.
Did you roll a dice 120 times? Which result did you get?

But to solve your problem.

I assume you set relevance to the six groups :
Group1: randNumber==1
Group2: randNumber==2
Group3: randNumber==3
...
well, just change the relevance.
Filled groups: relevance: 0 or even randNumber==10 (anything that's always false)
Group1: (randNumber==1) OR (randNumber=2)
Group3: (randNumber==3) OR (randNumber=4)
Group4: randNumber==5
Group6: randNumber==6
Or whatever you want.

So you can lead more respondents to the less filled groups.

Best regards
Joffm