10. Discriminant validities. We need a subsequent design that will permit us to assess the discriminant validity of the test with respect to g-based measures of abilities.
11. Range of skills tested. We need to develop a wider range of behaviors to be sampled, including behaviors from more varied jobs and behaviors that occur outside of jobs.
12. Translation. The procedures for translation were rather informal in this study and resulted in a few items that, in retrospect, were not ideally translated. The result was a lowering of the cross-sample correlation (which was nevertheless .91). A more careful procedure for ensuring the accuracy of translation is needed in a subsequent study, as well as for ensuring that the situations sampled adequately represent apposite situations in the two cultures assessed as well as others.
13. Focus groups. In subsequent work, we also plan prior to validation to have focus groups in each culture that will evaluate the relevance of each potential inventory situation for cultural appropriateness. Scenarios that are not deemed to be culturally valid by two-thirds of the focus group will be replaced.
14. Occupational groups. We will score by occupational groups in order to determine whether different occupational groups tend to prefer different responses to options.
15. Cultures. We would hope to introduce a third cultural context in subsequent work. Sternberg, Grigorenko, and Gil are all willing to participate in subsequent phases of this work.

In sum, the ESJI appears to be a promising inventory for the assessment of situational judgments, but further work is needed in order to refine the existing measure.