An interaction effect occurs when the effect of one process factor on a response depends on the setting of another factor. In design of experiments (DOE), interactions help explain why changing a temperature, pressure, or other operating setting produces different outcomes under different process conditions.
Consider a hypothetical bonding experiment with two temperatures and two pressures. Raising the temperature from 150°C to 170°C increases average bond strength by 2 MPa at low pressure and by 10 MPa at high pressure. The difference between those temperature effects indicates an interaction. A recommendation about temperature therefore needs to specify the pressure setting under which the expected improvement applies.
In practice, a factorial DOE allows the team to estimate interactions alongside main effects. An interaction plot displays the response across factor combinations; nonparallel lines suggest an interaction. Statistical analysis then assesses the evidence against experimental error. The operating decision also depends on whether the estimated effect matters against requirements, such as a minimum bond strength.
Interaction analysis belongs within DOE and supports Improve-phase decisions in DMAIC. One-factor-at-a-time trials can miss interactions because they examine changes while other settings remain fixed. After selecting a factor combination, confirmation runs test whether the predicted response is reproduced. Pilot testing can then assess performance under routine operating conditions.
