There are 234 observations where there is no variant recorded and
"agroup" and "g" variables are just blank...any idea what these are? The
weird thing is that these people all clicked to either share via
twitter/FB
or take the survey, but didn't receive any treatment. Were these us
testing?
I suspect this is some bad cookies / javascript interaction with some browsers, but it should be investigated a bit more.
I suspect this is some bad cookies / javascript interaction with some browsers, but it should be investigated a bit more.