Data　envelopment　analysis　is　a　data-driven　tool　for　building　a　composite　index.　Fare　et　al.　(1995)　proposed　an　attribute-incorporated　data　envelopment　analysis　where　a　Malmquist　productivity　index　including　attributes　was　formulated.　The　attribute-incorporated　Malmquist　productivity　index　was　decomposed　into　quality　change,　efficiency　change,　and　technical　change.　However,　this　attribute-incorporated　Malmquist　productivity　index　may　yield　incorrect　results　due　to　the　fact　that　its　separable　multiplier　only　considers　the　reference　technology　and　ignores　the　corresponding　period　data.　This　current　paper　proposes　a　revised　decomposition　version　of　the　attribute-incorporated　Malmquist　productivity　index.　In　addition,　we　propose　an　aggregate　quality　change　model　which　is　extended　from　Fare　et　al.　(1995),　where　subjective　quality　and　objective　quality　are　considered　at　the　same　time.　The　aggregate　quality　change　model　is　useful　for　measuring　changes　in　the　quality　of　life.　An　application　in　this　paper　shows　that,　within　quality　attributes,　subjective　quality　of　life　has　a　more　significant　impact　on　the　overall　urban　quality　of　life　than　objective　quality　of　life.