5. When does including controls in multivariate regressions not solve problems of omitted variable bias? 0 When you have more than one potential omitted variable 0 When you think that there is some "noise" or error in your measure of your outcome Q When important potential omitted variables cannot be measured