Provide statistical evidence that the construct, the method and the individual items behave equivalently in every intended population. Construct equivalence can be examined through factor analysis, weighted multidimensional scaling or comparison of nomological networks, and method bias should be traced to sources such as motivation, test familiarity, speediness, response format or response style. Item equivalence should be tested by differential item functioning analysis of each item, matching respondents on the measured trait, ideally with more than one matching criterion when the test is multidimensional. Translation faults and cultural context are the main suspects behind flagged items, and some items may need rewriting or removal.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.