Are uneven groups a sign of high risk of bias?

In general, you should be concerned if there’s a large imbalance between the two groups. Dramatically different treatment groups are less likely to occur by chance and could affect a study’s results.

Also, if the study seems poorly designed in other ways, this might be another red flag that something’s not quite right.

If you come across an issue like this when you’re assessing a study, check to see whether the study authors used statistical tools in their analyses to account for these differences.

Now, let’s move on to the next domain for bias: “allocation concealment.”