Well yeah. You’re measuring the output to a multi-variate function. There should be no expectation that this one output drives other occasionally correlated outputs to other functions.