Al-kazzaz et al (2010) described hybrid adaption technique to generate innovative designs from heterogeneous precedents using shape grammars. An evaluation of the degree of innovation in the hybrid designs gave feedback to grammar users before and after applying a rule. Innovation was assessed using variables derived from the internal structure of the grammar such as: the number of antecedents in the corpus having the same rule, the number of rules in a subclass rule set having the same geometry, etc. However, the validity of the innovation assessment was unclear and the use of the feedback measures was not demonstrated. Accordingly, this study aims to verify the credibility of the innovation measures and to identify the independent variables that a user can control to achieve a significant impact on each innovation measure as a dependent variable.