We must consider these findings in light of several limitations to the present study. Participants were all 21 to 26 year old Chinese- and Korean-American college students so the conclusions reached from this sample may not generalize to other ethnic subgroups (e.g., Macgregor et al., 2009; Whitfield, 2002), age groups, educational levels, or samples selected for alcohol dependence or cancers (e.g., Yokoyama et al. 1999; 2003). In addition, our endorsement rates for flushing was higher than reported in several previous studies, which may be due to differences in study methodology, including level of alcohol consumption referenced (1-2 standard drinks vs. 0.25 to 1.0 standard drinks), symptom definitions (general flushing/blushing of the body vs. facial flushing), time frames (ever vs. first year or two of drinking, current, or always), and assessment method (interview self-report vs. questionnaire self-report, patch test, laboratory alcohol consumption, or observer rated); such differences likely result in different endorsement rates of symptoms that make it difficult to directly compare across studies. It is also important to note that retrospective self-reports may be biased in ways that differ across
