Do People Ask Good Questions?

Rothe, Anselm; Lake, Brenden M.; Gureckis, Todd M.

doi:10.1007/s42113-018-0005-5

Cited by 50 publications

(54 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, Brown‐Schmidt, Gunlogson, and Tanenhaus (2008) found that participants naturally asked their partner questions about occluded objects in a task that required mutual knowledge about these objects for success. Indeed, actively highlighting and resolving known sources of uncertainty about a partner's private information may be one of the primary functions of questions in discourse (Hawkins, Stuhlmüller, Degen, & Goodman, 2015; Rothe, Lake, & Gureckis, 2018). Further work is needed to determine the resource‐rational trade‐offs of asking explicit questions about hidden distractors as opposed to implicitly increasing the specificity of one's referring expressions to account for them, as we found here.…”

Section: Discussionmentioning

confidence: 99%

The Division of Labor in Communication: Speakers Help Listeners Account for Asymmetries in Visual Perspective

2021

View full text Add to dashboard Cite

Recent debates over adults' theory of mind use have been fueled by surprising failures of perspective‐taking in communication, suggesting that perspective‐taking may be relatively effortful. Yet adults routinely engage in effortful processes when needed. How, then, should speakers and listeners allocate their resources to achieve successful communication? We begin with the observation that the shared goal of communication induces a natural division of labor: The resources one agent chooses to allocate toward perspective‐taking should depend on their expectations about the other's allocation. We formalize this idea in a resource‐rational model augmenting recent probabilistic weighting accounts with a mechanism for (costly) control over the degree of perspective‐taking. In a series of simulations, we first derive an intermediate degree of perspective weighting as an optimal trade‐off between expected costs and benefits of perspective‐taking. We then present two behavioral experiments testing novel predictions of our model. In Experiment 1, we manipulated the presence or absence of occlusions in a director–matcher task. We found that speakers spontaneously modulated the informativeness of their descriptions to account for “known unknowns” in their partner's private view, reflecting a higher degree of speaker perspective‐taking than previously acknowledged. In Experiment 2, we then compared the scripted utterances used by confederates in prior work with those produced in interactions with unscripted directors. We found that confederates were systematically less informative than listeners would initially expect given the presence of occlusions, but listeners used violations to adaptively make fewer errors over time. Taken together, our work suggests that people are not simply “mindblind”; they use contextually appropriate expectations to navigate the division of labor with their partner. We discuss how a resource‐rational framework may provide a more deeply explanatory foundation for understanding flexible perspective‐taking under processing constraints.

show abstract

Section: Discussionmentioning

confidence: 99%

The Division of Labor in Communication: Speakers Help Listeners Account for Asymmetries in Visual Perspective

2021

View full text Add to dashboard Cite

show abstract

“…One potential answer may be found in the strategies people use to actively explore the world. By asking questions (Coenen, Nelson, & Gureckis, 2018;Nelson, 2005;Rothe, Lake, & Gureckis, 2018;Wu, Meder, Filimon, & Nelson, 2017), directing eye-movements towards informative scenes (Cavanagh, Hunt, Afraz, & Rolfs, 2010;Najemnik & Geisler, 2005;Nelson & Cottrell, 2007), or sampling different options (Hertwig, Barron, Weber, & Erev, 2004;Hills & Hertwig, 2010), people are remarkably efficient at acquiring relevant information for making good decisions (Pirolli & Card, 1999;Todd, Hills, & Robbins, 2012). We investigate this capacity for active learning as a potential source of the gap between human and machine learning.…”

Section: The Gap Between Human and Machine Learningmentioning

confidence: 99%

Asking better questions: How presentation formats influence information search.

Wu¹,

Meder²,

Filimon³

et al. 2017

Journal of Experimental Psychology: Learning, Memory, and Cogni

View full text Add to dashboard Cite

While the influence of presentation formats have been widely studied in Bayesian reasoning tasks, we present the first systematic investigation of how presentation formats influence information search decisions. Four experiments were conducted across different probabilistic environments, where subjects (N = 2,858) chose between 2 possible search queries, each with binary probabilistic outcomes, with the goal of maximizing classification accuracy. We studied 14 different numerical and visual formats for presenting information about the search environment, constructed across 6 design features that have been prominently related to improvements in Bayesian reasoning accuracy (natural frequencies, posteriors, complement, spatial extent, countability, and part-to-whole information). The posterior variants of the icon array and bar graph formats led to the highest proportion of correct responses, and were substantially better than the standard probability format. Results suggest that presenting information in terms of posterior probabilities and visualizing natural frequencies using spatial extent (a perceptual feature) were especially helpful in guiding search decisions, although environments with a mixture of probabilistic and certain outcomes were challenging across all formats. Subjects who made more accurate probability judgments did not perform better on the search task, suggesting that simple decision heuristics may be used to make search decisions without explicitly applying Bayesian inference to compute probabilities. We propose a new take-the-difference (TTD) heuristic that identifies the accuracy-maximizing query without explicit computation of posterior probabilities. (PsycINFO Database Record

show abstract

“…These considerations may also shed new theoretical light on recent results by Rothe et al (2018) finding that in games like Battleship, people often fail to ask maximally informative natural language questions (although they can accurately rank the informativity of pre-generated questions). By embedding questioning in a social, communicative context (see also Shafto, Goodman, & Frank, 2012;Shafto, Goodman, & Griffiths, 2014), our framework suggests that this behavior may in fact be resource-rational under the conditions where most questioning takes place (Griffiths, Lieder, & Goodman, 2015).…”

Section: The Correspondence Between Questions and Goalsmentioning

confidence: 92%

Why do you ask? The informational dynamics of questions and answers

Hawkins¹,

Goodman²

2017

Preprint

View full text Add to dashboard Cite

Asking questions is one of our most efficient and reliable means of learning about the world. Yet we do not often pose these questions to an impartial oracle; we ask cooperative social partners, in dialogue. In this paper, we aim to reconcile formal models of optimal question asking and answering with classic effects of social context. We begin from the observation that question-answer dialogue is motivated by a two-sided asymmetry in beliefs: questioners have a private goal but lack goal-relevant information about the world, and answerers have private information but lack knowledge about the questioner's goal. We formalize this problem in a computational framework and derive pragmatic questioner and answerer behavior from recursive social reasoning. Critically, we predict that pragmatic answerers go beyond the literal meaning of the question to be informative with respect to inferred goals, and that pragmatic questioners may therefore select questions to more unambiguously signal their goals. We evaluate our pragmatic models against asocial models in two ways. First, we present computational simulations accounting for three classic answerer effects in psycholinguistics. We then introduce the Hidden Goal paradigm for experimentally eliciting questioner and answerer behavior in scenarios where there is uncertainty about the questioner's goal. We report data from three experiments in this paradigm and show how our core computational framework can be composed with more sophisticated question semantics, hierarchical goal spaces, and a persistent state over which extended dialogue can unfold. We find that social inference is needed to account for critical aspects of the data.

show abstract

Do People Ask Good Questions?

Cited by 50 publications

References 23 publications

The Division of Labor in Communication: Speakers Help Listeners Account for Asymmetries in Visual Perspective

The Division of Labor in Communication: Speakers Help Listeners Account for Asymmetries in Visual Perspective

Asking better questions: How presentation formats influence information search.

Why do you ask? The informational dynamics of questions and answers

Contact Info

Product

Resources

About