I

I have a question, a colleague said that Minitab leads to wrong interpretations in DoE for high variance and small effects. He would have used 3 replications of the matrix. Since only the mean value is considered with DoEs, a significant effect is shown even though the variance were very high and therefore the effects were significant. Is this true, or is his statement from my colleague wrong? If so, how can this effect be countered? I think the variance is observed by the F-Value/p-Value?

Thank you very much!

Max