This research tries to propose a general construct for computational models handling affect dedicated to complex and collective decision-making. The importance of integrating emotional, personal, and social intelligences, in complex individual and collective decision-making, is highlighted. Complex decision-making is approached from human to computational perspectives with the main perspective of complex problem solving. The objective of this paper is hence to: 1) examine how emotional, personal, and social intelligences capabilities contribute to effective collective decision-making in complex environments, 2) investigate how these capabilities can be computationally modeled to enable agents to build internal representations of the systems they manage, learn to process and respond to highly complex and dynamic information, and execute deliberate, prioritized cognitive and behavioral strategies to achieve desired outcomes in real-world problem solving, 3) identify current methodologies and approaches that integrate these forms of intelligence in agent-based systems, and 4) highlight promising future research directions and alternatives emerging from initial findings in this field. The main results are that this study identifies seven core mechanisms through which individual and group affect influence complex collective decision-making, integrating bottom-up and top-down emotional workflows into a single agent-based model. The implications of this study are that by combining affective, cognitive, and environmental parameters — weighted using statistical, knowledge-based, and machine learning methods — the model enables more adaptive, human-like behavior in artificial general intelligence systems.