The statistical properties that surround the detection of the interaction should be specified and discussed. Is the interaction significant on an additive or multiplicative scale? Investigators should provide evidence that detected interaction effects are robust to transformation of scale and/or provide a strong defense of the scaling employed. If it is not robust to transformation, the implications of this should be discussed. Common artifacts associated with nonlinearity should also be evaluated and ruled out. Similarly, if the interaction is detected under some conditions or using some measures of the environment but not others, this should be reported and incorporated into the theory behind what the findings contribute to the literature. Perhaps most importantly, the raw data along with 95% confidence intervals should always be plotted and reported in the publication, not just the model-derived regression lines, which can be misleading, especially when some combinations of the independent variable have small sample sizes and when control covariates are included in the model. Because apparent cGxE effects can be due to stratification, genotyping artifacts, and even gender (Keller, 2014), careful attention must
