In 2012 Michigan implemented Great Start to Quality, a voluntary quality rating and improvement system (QRIS) that uses a multidimensional assessment to rate the quality of early childhood education programs. Changes to the rating calculation approach announced in 2013 provided an opportunity for Regional Educational Laboratory (REL) Midwest to examine how the changes would affect program quality ratings. Under the revised calculation approach (version 2.0), approximately one-third of programs had a higher self-assessment rating, though the underlying data and measures of program quality were unchanged. A simple alternative total score approach developed by REL Midwest that eliminated criteria for domain scores on the self-assessment yielded rating distributions that were nearly identical to those from the version 2.0 approach. These findings suggest that incremental changes to how QRIS ratings are calculated can alter inferences about program quality. (author abstract)