Determinants of Explanation-Seeking Curiosity Across Development

• In Event: 3-081 - The Emergence of Causal Reasoning and Causal Explanations in Childhood: Diverse Definitions and Methodologies

Integrative Statement

Much of human learning is achieved by seeking explanations, yet little is known about what triggers explanation search. Here we study explanation-seeking curiosity (ESC), the desire to know why, as a window onto human learning across the lifespan.
In Study 1, 187 adults received eight why-questions drawn from a set of 80 sampled from the Explain Like I’m Five Reddit (e.g., “Why do our taste buds change over time?”) or created from a semantic network (e.g., “Why are dogs a type of canine?”). Each question was rated for ESC (“How curious are you about why…”) as well as 13 potential determinants of ESC, derived from accounts of attribution (e.g., Bruckmüller et al., 2017), curiosity (e.g., Dubey & Griffiths, 2017; Loewenstein, 1994), need for explanation (e.g., Grimm, 2008; Wong & Yudell, 2015), self-explanation (e.g., Lombrozo, 2016), or the division of cognitive labor (e.g., Lutz & Keil, 2002). A factor analysis on these determinants revealed three factors (see Figure 1), corresponding to “learning” (how complex and informative the question’s answer was expected to be), “export” (how useful the answer was expected to be in future contexts), and “surprise” (how inconsistent the explanandum was with prior beliefs). A regression analysis predicting ESC from the resulting factor scores revealed that all three factors were significant predictors, with learning (B = 0.38, 95% CI [0.35, 0.42]) and export (B = 0.37, 95% CI [0.33, 0.41]) notably stronger predictors than surprise (B = 0.20, 95% CI [0.16, 0.23]).
In Studies 2 and 3, we repeated Study 1, again with adult participants (NStudy 2 =87, NStudy 3 = 82), but using why-questions asked by children. 40 questions asked by Kindergarteners through 3rd graders (Study 2; e.g., “Why does jelly wobble?”) and 40 questions asked by 4th through 6th graders (Study 3; e.g., “Why do spiders have so many eyes?”) were sampled from www.madsci.org. Factor analyses revealed remarkably similar three-factor structures in comparison to Study 1 (see Figure 1), However, in both Studies 2 and 3, regression analyses predicting ESC from these factor scores revealed that learning remained a strong predictor (B-Study 2 = 0.52, 95% CI [0.47, 0.57]; B-Study 3 = 0.40, 95% CI [0.33, 0.47]), while export did not (B-Study 2 = 0.22, 95% CI [0.16, 0.28]; B-Study 3 = 0.20, 95% CI [0.14, 0.27]) (see Figure 2).
These studies suggest that ESC is driven by surprise and by the potential to learn about the explanandum itself, but also by the anticipated value of what can be learned (“export”). The importance of export in driving why-questions seems to increase across the lifespan, with a corresponding decrease in the importance of learning. However, these conclusions are derived from adults’ ratings of ESC; does children’s own curiosity support the same conclusions? In ongoing work, we are studying whether children’s ESC is driven by these same three factors. Preliminary analyses from CHILDES suggest that it is. Together, these studies provide deeper insight into the factors that drive explanation search across the lifespan.

Authors

• Emily G. Liquin, Princeton University
  Presenting Author

• Tania Lombrozo, Princeton University
  Non-Presenting Author