Task Complexity in Individual Stock Control Tasks for Laboratory Experiments on Human Understanding of Dynamic Systems

Abstract

Dynamic stock control tasks have been frequently used in laboratory experiments in behavioural research to investigate understanding of dynamic systems. In these studies, the dynamic system is often represented in the form of a simulation model, and they almost exclusively focus on how the structure of a system (i.e. the simulation model) affects human's inference of system behaviour. In doing so, these studies fail to consider that human's performance on dynamic decision making tasks might also be a function of the complexity embedded in other task components like goals, input, processes, output, time and presentation. Hence, the objective of this paper is to carve out what task complexity entails when applied to dynamic stock control tasks in order to determine its usefulness for future research on human understanding of such tasks. In this paper, task complexity is conceptualized consisting of ten complexity dimensions: (1) size; (2) variety; (3) redundancy; (4) ambiguity; (5) variability; (6) inaccuracy; (7) novelty; (8) incongruity; (9) connectivity; and (10) temporal demand. 2 These factors correspond to Liu and Li's (2012) 'Complexity contributory factors'. Syst. Res

Unrecognized interdependencies in mental models—The case for taking feedback loops into account

“…The attention of many system dynamics researchers has turned to the “stock‐and‐flow error” (SFE), which deals with the apparent difficulties humans have to infer the behavior of stock variables from information about the flows adding to or draining stocks (Baghaei Lakeh & Ghaffarzadegan, ; Bleijenbergh, Vennix, Jacobs, & van Engen, ; Booth Sweeney & Sterman, , ; Capelo & Dias, ; Cronin & Gonzalez, ; Cronin, Gonzalez, & Sterman, ; Qi & Gonzalez, ; Sterman, ; Stouten & Größler, ; Stouten, Heene, Gellynck, & Polet, ; Sweeney & Sterman, ; Sweeney & Sterman, ; Sweeny & Sterman, ). Just like the MoF thread, this line of research investigates decision processes and suggests that pattern‐matching heuristics may be responsible for the difficulty to intuitively understand accumulation.…”

“…Most of the MoF articles also mention the term "dynamic decision making" (DDM) because the experiments expose participants to feedback-driven systems with stock-accumulation, where decisions taken at one point in time change the conditions under which decisions will be taken at later moments. Several authors carrying out DDM experiments have done so in search of inferring information about the participants' mental models (Arango et al, 2012;Gary, Pillinger, & Wood, 2012;Gary & Wood, 2011, 2016Gonzalez, Fakhari, & Busemeyer, 2017;Stouten et al, 2012;Stouten & Größler, 2017). Some researchers report that more accurate mental models have come together with increased performance in decision experiments (Capelo & Dias, 2009).…”

‘Behavioural System Dynamics’: A Very Tentative and Slightly Sceptical Map of the Territory

“…In fact, we know that our tendency to ignore nonlinear relationships and to imagine that relationships are directly functional rather than involving accumulation also play a role (Forrester, ). Indeed, the latter gives rise to arguably the most active ‘pre‐existing’ areas of BehSD, the interest in the problems humans have understanding systems involving stocks (Sterman and Booth Sweeney, ; Ritchie‐Dunham and Puente, ; Cronin et al ., ; Abdel‐Hamid et al ., ; Stouten and Groessler, ; Kapmeier et al ., ).…”

Three Generic Policies for Sustained Market Growth Based on Two Interdependent Organizational Resources—A Simulation Study and Implications

“…The dynamic nature of this task is already challenging because of the interdependencies between them, but its complexity increases when different resources grow at different speeds [11]. In such cases, their trajectories react to one another in counter-intuitive ways: decision-makers build mental models of the situation's structure which contain flaws like the "misperception of feedback" [12][13][14][15] or mentally infer incorrect behavioral implications like in the case of the "stock and flow error" [16][17][18][19][20][21][22]. Such flawed mental models lead to policies that drive the system into an initial phase of strong and exponential growth followed by a crisis and decline often referred to as "boom and bust" [23].…”