Reasoning about complex probabilistic concepts in childhood

Abstract

The competencies of children, particularly their understanding of the more complex probabilistic concepts, have not been thoroughly investigated. In the present study participants were required to choose the more likely of two events, a single event, and a joint event (conjunctive or disjunctive). It was predicted that the operation of the representativeness heuristic would result in erroneous judgements when children compared an unlikely component event with a likely-unlikely conjunction (the conjunction fallacy) and when a likely component event was compared to a likely-unlikely disjunction. The results supported the first prediction with both older children aged between 9 and 10 years and younger children aged between 4 and 5 committing the conjunction fallacy. However, the second prediction was not confirmed. It is proposed that the basis of representativeness judgements may differ between the conjunctive and disjunctive cases with absolute frequency information possibly playing a differential role.

Judgment and decision making in young children

“…While most studies were limited to elementary school age, the conjunction fallacy (a tendency to consider a more specific event more likely than a more general event), notably, has now been reported from age four (Fisk, Bury and Holden, 2006). Non-normative strategies thus emerge as early as the normative strategies reviewed earlier.…”

“…Other studies found no change, in framing effects (Schlottmann and Tring, 2005;Levin and Hart, 2003), use of the representativeness heuristic (Agnoli, 1991), the conjunction fallacy (Fisk et al, 2006), and anchoring effects (Smith, 1999). Yet others found a decrease, in use of the availability heuristic (Davies and White, 1994), sunk cost decisions (Klaczynski and Cottrell, 2004), and averaging errors (Schlottmann, 2000).…”

Linda is not a bearded lady: Configural weighting and adding as the cause of extension errors.

“…Accordingly, to argue that a patient with disease X is less likely to have a headache than both fever and a headache would be to commit a conjunction error and to argue that a patient with disease X is more likely to have a headache than either a fever or a headache (or both) would be to commit a disjunction error. Conjunction and disjunction errors have received considerable attention (e.g., Bar-Hillel & Neter, 1993; Fisk, 2002; Fisk, Bury, & Holden, 2006; Lagnado & Shanks, 2002; Nilsson, 2008; Rottenstreich, Brenner, & Sood, 1999; Stolarz-Fantino, Fantino, Zizzo, & Wen, 2003; Tentori, Bonini, & Osherson, 2004; Tversky & Kahneman, 1983; Wedell & Moro, 2008) and have been observed in such diverse populations as health educators (Adam & Reyna, 2005), auditors (Lindberg & Maletta, 2003), and bettors (Nilsson & Andersson, in press).…”

Why the Conjunction Effect Is Rarely a Fallacy: How Learning Influences Uncertainty and the Conjunction Rule

“…A well-documented example of this supposed irrationality is the conjunction effect, whereby people consistently rate a conjunction as more probable than its constituents in isolation (e.g., Tverksy and Kahneman, 1983 ; Bar-Hillel and Neter, 1993 ; Fisk, 2002 ; Lagnado and Shanks, 2002 ; Sides et al, 2002 ; Sloman et al, 2003 ; Stolarz-Fantino et al, 2003 ; Tentori et al, 2004 ; Fisk et al, 2006 ; Crupi et al, 2008 ; Nilsson, 2008 ; Wedell and Moro, 2008 ; Moro, 2009 ). Given its salience, the continuing lack of a generally accepted explanation for the conjunction effect is remarkable (Pohl, 2004 ; Nilsson et al, 2009 ; Jarvstad and Hahn, 2011 ).…”