One area of research that has seen particularly strong boost is the study of what author's broadly refer to here as lexicosyntactic alternations. Given the ease with which frequencies of choices involving lexical material can be extracted from corpora, it comes as no surprise that there are many reference works and studies that report and utilize frequencies of occurrence of the alternants that make up alternations. The chapter draws attention to small set of additional factors whose inclusion would benefit corpus-based research on alternative linguistic choices. The first kind of random effect that distorts overall frequencies but is usually readily available from the fully annotated, case-by-variable format is speaker-/file-specific variation. In fact, an even better kind of analysis would also take into consideration the fact that speakers/ files are nested into varieties, which are in turn nested into variety types, and variability in future choice can be manifested at each of these levels of resolution.