A polling organization wants to estimate voter support for a ballot measure. They divide voters by political party (Democrat, Republican, Independent) and interview a fixed number from each party by approaching people at political rallies and campaign events. Compared to stratified random sampling of voters by party, this quota sampling approach is most problematic because of: