Kompleksnost zadatka i odlučivanje u uslovima neizvesnosti

Abstract

Ova studija imala je za cilj da proveri na koji način kompleksnost zadatka, u sadejstvu s drugim potencijalnim faktorima, utiče na odlučivanje u uslovima neizvesnosti. Kompleksnost zadatka je objektivno definisana preko dve varijable: broja potencijalnih događaja u budućnosti (dva, tri i pet događaja) i razlika u neto vrednostima ishoda (male i velike razlike). U oba slučaja, kompleksnost je zasnovana na neizvesnosti. Sprovedena su četiri eksperimenta: u dva eksperimenta zadaci su bili smešteni u ekonomski, a u dva u medicinski domen. Pored dve varijable kojima je operacionalizovana kompleksnost, u svakom od eksperimenata variran je po još jedan faktor, koji se odnosi na karakteristike zadatka: referentna tačka (zona gubitka i zona dobitka), stepen uključenosti u odlučivanje (donošenje odluke za sebe i donošenje odluke za drugog) i okvir (pozitivan i negativan). Zadaci su osmišljeni tako da se na osnovu donesene odluke može jednoznačno odrediti koji od tri modela maximax, maximin, minimax se nalaze u osnovi izbora. Ujedno, ovakva procedura je omogućila objektivno praćenje promene strategije odlučivanja u zavisnosti od eksperimentalnih uslova. Nalazi ukazuju da se prilikom odlučivanja ispitanici oslanjaju na sva tri modela odlučivanja, s tim da je većina odluka donesena tako da smanji rizik od gubitka, tj. na osnovu maximin i minimax modela. Utvrđene su interakcije kompleksnosti zadataka, s jedne strane, i domena, stepena uključenosti i referentne tačke, s druge strane. Kada se radi o kompleksnijim zadacima, tj. u slučaju većeg kognitivnog opterećenja, ispitanici pokušavaju da olakšaju proces odlučivanja, zbog čega dolazi do većeg ispoljavanja kognitivnih pristrasnosti, kao što su uticaj okvira i averzija prema gubitku

The aim of this study is to explore how task complexity and other potential factors impact decision making under uncertainty. Task complexity is objectively defined through two variables: the number of potential events in the future (two, three and five) and the difference in the net value of the outcomes (small and big differences). In both cases the complexity is based on uncertainty. Four experiments were carried out. In two experiments the tasks were placed in an economic domain, whereas the other two tasks were placed in a medical domain. Beside the two variables through which complexity was conducted, in each of the experiment one more factor referring to task characteristics was varied, i.e. the reference point (domain of loss and domain of gain), the level of decision making involvement (either for yourself or for somebody else) and the frame (as either positive or negative). The tasks were designed in a way that, based on the decision made, one can unambiguously define which of the three models – maximax, maximin or minimax – are at the base of the choice. In addition, this procedure enabled objective tracking of the change strategy of uncertain decision making depending on the experimental conditions. The findings show that in the course of decision making process, the participants rely on all three models of decision making, provided that most decisions are made in order to reduce the risk of loss, i.e. based on maximin and minimax model. Interactions of the task complexity, on one hand, and the domain, the level of involvement and the reference point, on the other hand, were determined. In the case of more complex tasks, i.e. in the case of greater cognitive load, the participants try to make decision making easier, which leads to more cognitive biases, such as frame impact and loss aversion