An application of a weighting method to adjust for nonresponse in standardized incidence ratio analysis of cohort studies