Reports of survey findings can be of questionable value because of weaknesses in study design or because of deficiencies in implementation: inadequate sample size, biased sampling methods, low subject participation rates, uncertain measurement reliability, and incomplete or erroneous data analyses. Even with surveys of unquestionable integrity, comparison of findings between studies can be problematical if examination methods are non-uniform, or if there are inconsistencies in the definitions of conditions/diseases. Also, between-study comparisons may not be particularly meaningful for surveys based on clinic or other special populations of limited representativeness. This presentation will review these potential shortcomings, and illustrate how they were addressed in a recent series of six surveys of refractive error and visual impairment in school-age children in Chile, China, India, Nepal, and South Africa.