This Paper aims to clarify fundamental aspects of assigning fuzzy scores to conditions with multiple attributes in fuzzy set Qualitative Comparative Analysis (fsQCA). Fuzzy Multiple Attribute Conditions (FMAC) are conditions that are built using different types of concepts. In relation to FMAC we can find at least two issues: 1) the calibration strategy that requires one to consider the complexity of the descriptive and causal context; and 2) the aggregation strategy that should be able to capture conceptual properties of membership and similarity. We will employ an empirical example in order to deal with causal and descriptive heterogeneity. After discussing the disadvantages of the aggregation techniques used in QCA, we individuate an axiomatic framework for defining logical conjunction operators that allows one to aggregate parts of concepts in accordance with membership and similarity. Then we propose a technique to assign fuzzy scores to FMAC using the Arithmetic Mean Based Compensatory Fuzzy Logic (AMBCFL). This techniques has several qualitative advantages: a) it allows one to compute concepts by using the mean of fuzzy numbers because it better mimic natural language and transforms linguistic variables into fuzzy sets; b) it aggregates data without losing important information such as the internal ranking of attributes; c) it makes the researcher think about how a condition is conceptualized and how to choose the right attributes; d) it better deals with causal and descriptive heterogeneity during the process of transformation of raw data into fuzzy numbers; and e) it allows one to better represent reality in fuzzy numbers and locate the cases in the XY plot. This technique allows one to individuate a more reliable solution formula(s) following the QCA analysis and better locate cases in the XY plot during the post-QCA analysis.
