In general, the measurement of direct/objective outcomes (e.g. production errors, staff turnover rate) is more likely to be valid and reliable than that of self-report/subjective outcomes (e.g. trust among team members, stress).
In addition, when the evidence comes from scientific research that uses a questionnaire, check if it was also used in other studies. If it was ‘self-made’ it is less likely to be valid and reliable.