Measuring Disaggregation Quality

We build a modified measure of disaggregation quality (MDQ), based on the disaggregation quality (DQ) measure from Chen, Miao, and Shevlin (2015). Like DQ, MDQ is a parsimonious measure of data disaggregation quality in the annual report that can be constructed through the percentage of nonmissing financial statement items in Compustat. However, we take additional screening steps to minimize counting errors by resolving missing values through accommodating GAAP changes over time and using a more comprehensive set of accounting equations from Casey et al. (2016). Consequently, MDQ relates more consistently to information quality proxies than DQ does, in pooled regression, across subsamples, and for various sample periods. Further analyses in other research settings demonstrate that MDQ and DQ can lead to different inferences