A composite indicator is a mathematical combination of single indicators that represent different dimensions of a concept whose description is the objective of the analysis. If all the indicators refer to a single latent concept, classical MDA techniques, like Factorial Analysis or Principal Component Analysis, can be used. These techniques allow us to assess the impact of each indicator on the composite indicator. However, often the several indicators used in the construction of a composite indicator express different aspects of a complex phenomenon, and so they can be conceptually split in several blocks of indicators. Each block can be resumed by a composite indicator, which is considered causative with respect to a second-order composite indicator. We refer to this kind of index, which is a synthesis of composite indicators, as complex indicator.
ESPOSITO VINZI, V., TRINCHERA, L. and RUSSOLILLO, G. (2015). Multidimensional and Non-Metric Extensions of PLS Path Modeling for Synthetic Indicators. In: Dealing with Complexity in Society: from Plurality of Data to Synthetic Indicators.