Minimization with equal groups

[Ashish Awasthi]

Hello
I'm trying to divide a group of 30 animals into two groups before conducting an experiment with differential treatments on these groups.
The experiment is expected to have consequences on weight (in grams, continuous) so I'd like these groups to be created so as to have equal means (no statistically different) and variances for this variable, and of course an equal number of animals for each.
Could you please help me on the commands to use (can we include random number seed)? I am using Stata 16.1 MP
Thank you in advance

[Wei Hai Deng]

I hope someone more experienced with the exact statistical issues/assumptions that you need to consider when designing your type of study can answer but you might be looking for something like the splitsample nsplit(2) cluster() balance() command.
Just off the top of my head there are probably several things you need to consider before deciding on a split scheme, eg. how close are the covariates at baseline, assumed distribution of these, etc. Also, you should probably draw a DAG (directed acyclic graph) for your assumptions about the causal relationships, covariates, etc.

More details is almost always necessary to answer these questions. What is your aim, exposure, outcome, examples of things you will be assessing, is this an intervention, etc...

Best,
Wei

[Rich Goldstein]

the question is not completely clear but there are user-written commands for minimization allocation including minmse and rct_minim; use search to find, download and install