The Causality and Mind Lab

Newsletter

Brown University

2011-2012

Cognitive, Linguistic and Psychological Sciences
Brown University
Box 1821 * Providence, RI 02912
Phone: (401) 863-3527
Email:
/

Sometimeduring the past two years, your child participated in a study at our lab. In this newsletter, you can find descriptions of some of the experiments that have been published or that are still underway. These projects do not represent all our projects, and your child might have participated in something completely different. If you would like a copy of one of these or any of our published reports, you can download them at our website

Just click on the “publications” link

Remember, our lab is able to contribute to scientific research because of good word of mouth. If you enjoyed your experience, please pass along our contact information to a friend. We would also love to hear from you if you would like to participate again! You can call us at (401) 863-3527 or email us at . You can also go to our website, and sign your child up!

Thank you again for volunteering your and your child’s time!

David Sobel

Associate Professor

Cognitive and Linguistic Sciences

Brown University

Diagnostic Reasoning and Scientific Thinking

Most of our work shows that young children are really good at reasoning about cause and effect. Our studies show that even infants have rudimentary causal reasoning capacities.

But there is also a lot of research that suggests that older children struggle at scientific reasoning. There’s a worry among both researchers and educators about how difficult it is for children to engage in scientific thinking and to get children interested in the scientific process.

So, if children are so good at causal reasoning, why is it that they’re so bad at science?

There’s no easy answer to this question. We do have a few ideas. The first is that most studies on causal reasoning focus on a kind of inference called predicting: figuring out what’s going to happen next. Scientific thinking is all about diagnosis: figuring out what caused an event to happen. These might seem similar, but it turns out that they’re very different ideas. And children reason about them quite differently.

With my students Phil Fernbach and Deanna Macris, we examined the difference between young children’s predictive and diagnostic reasoning capacities. We showed 3- and 4-year-olds a machine that activated when certain blocks were put on it (informally, in the lab, we call these blocks “blickets” and the machine a “blicket detector” – but we didn’t use those labels in these experiments).

In one experiment, we showed children the detector and three blocks. Two blocks made the machine go and one did not. In one condition, we simply asked the children what they thought would happen when we put each block on the machine again. Children were really good at answering these questions.

In the other condition, we put up a big screen between the child and researcher, so that the children could not see what was happening, and we activated the machine. We then removed the screen, and asked the children which block we had used to make the machine go. Regardless of what children answered, we told them (politely and in a supportive way) that their choice wasn’t incorrect, and we asked them to try again.

The idea behind this condition is to see whether children can revise what they believe in a plausible manner. There are two blocks that could have been used. We wanted to see if children would pick these two blocks. We found that 4-year-olds were pretty good at this – better than the 3-year-olds. This was the first sign of children’s diagnostic reasoning developing.

In another experiment, we also showed 3- and 4-year-olds three blocks. In this experiment, we showed children that one block made the machine go, the second did not, and we never placed the third on the machine (so children don’t know what it can do). We then did the same thing – introduce the screen, activate the machine, asked children what we used to make it go, and then had them revise their belief after their first choice.

The point of this experiment is that while one block obviously makes the machine go and another does not, what the third block can do is unknown. It could make the machine go. Specifically, when asked (particularly the second time) which one made the machine go, it’s a plausible answer. We examined how often children picked these two blocks – the one that obviously made the machine go and the one with unknown efficacy. The answer is that 3-year-olds rarely made this response. Four-year-olds were slightly better, but only did so 39% of the time. Clearly, they have a lot of room to improve!

My student Chris Erb and I have expanded on these findings – we ran similar experiments with children between the ages of 3-7. Children’s diagnostic reasoning abilities improved between these ages, but even the 7-year-olds weren’t as good as adults.

Our goal now is to start to relate these diagnostic reasoning abilities to children’s scientific reasoning. The experiments we’re currently running attempt to bridge this gap.

To Read More:

Fernbach, P., Macris, D. M. & Sobel, D. M. (2012). Which one made it go?: The emergence of diagnostic reasoning in preschoolers. Cognitive Development, 27, 39-53.

Sobel, D. M., & Erb, C. (in progress).Diagnostic reasoning and belief revision in children.Brown University.

Phil Fernbach completed his Ph.D. in 2010. He stayed at Brown as a postdoctoral researcher in 2011 and is now an Assistant Professor at the University of Colorado.

Deanna Macris is a 3rd-year graduate student.

Chris Erb is a 2nd-year graduate student.

How Children Learn from Other People

Imagine if a child never encountered another person. She would acquire pieces of knowledge that aredirectly observable(e.g., concepts like unsupported objects fall or objects don’t change shape, even if she would have no language to describe these concepts). But what she would lack is any knowledge of convention or an understanding of structures that involve hidden or unobservable events. In general, children do not spontaneously form beliefs about Santa Claus, angels, germs, Scientology, vitamins, the meaning of words, and the rules of Chutes and Ladders without other people.

How do children do this? When are they capable of doing this? Is this kind of learning different from learning from observing information in the world? We have been addressing these questions in a set of studies.

First, consider language: if you meet an individual who mislabels common objects (i.e., calls a pen a “shoe”), that person probably isn’t a good source of future information (you probably wouldn’t believe his novel label for a novel object). My graduate student Deanna Macris and I examined this, and we found that 4-year-olds certainly treat this person’s information differently than a person who correctly labels common objects. With Julie Sedivy (a colleague at the University of Calgary) and former graduate student Dave Buchanan and undergraduate student Rachel Hennesey, we found that not only did children make different inferences about these individuals, but their ability to track the meaning of their statements differed as those individuals were talking.

Second, consider expertise: in some ways, we’re all language experts – we all use language competently (more or less). But, some of us have specialized knowledge that others do not. My wife is an avid bird watcher, and can label different kinds of egrets easily. I just look at the egrets and say “bird.” Who would you trust for novel information about chickadees? My former undergraduate student Kathleen Corriveau and I asked 4-year-olds this question (about a novel kind of expertise), and found that they responded similarly to adults.

But it turns out that the 4-year-olds were really smart. They knew to ask a particular expert for novel information in that person’s area of expertise, but not for any kind of novel information. Consider my wife and I – without knowing anything more about us, who would you ask about the meaning of a novel word? If you answered “either one of you” – then you responded just like our 4-year-olds. They didn’t just rely on the individual who was expert in one thing – they assessed whether what was being asked about was related to each person’s expertise.

More generally, we hypothesized that what you know influences how you appreciate someone as a source of knowledge. Our recent experiments (mostly done by Katie Green, an undergraduate in my lab, and I) assess what children know about a particular type of information and then presented children with individuals who knew the same amount as they did or a different amount. What we found is that children’s own knowledge influenced whom they trusted for novel pieces of information – even if that knowledge was objectively wrong from an adult perspective. As children’s knowledge develops, their ability to learn from others follow suit.

To Read More:

Sobel, D. M., & Corriveau, K. H. (2010). Children monitor individuals’ expertise for word learning. Child Development, 81, 669-679.

Sobel, D. M., & Macris, D. M. (in press).Children’s understanding of speaker reliability across linguistic domains. Developmental Psychology.

Sobel, D. M., Sedivy, J., Buchanan, D. W., & Hennessey, R. (2012).The role of speaker reliability in children’s inferences about the meaning of novel words.Journal of Child Language, 39, 90-104.

Sobel, D. M., & Green, K. F. (2012).Preschoolers’ existing knowledge influences whom they trust: The case of false belief. Manuscript submitted for publication, Brown University.

Kathleen Corriveau graduated in 2003 and got her Ed.D. from Harvard in 2010. She is now an Assistant Professor at Boston University.

Dave Buchanan completed his Ph.D. in 2011. He is now a Research Scientist for IBM. He works on the Watson Project (the computer program that won Jeopardy). He’s teaching it how to reason about causality.

Fantasy and Reality

Are young children especially prone to magical or fantastical thought? Many parents believe they are, and several studies find that children easilyaccept fantastical events and creatures as real. But other studies find that 4-year-olds clearly discriminate between reality and fiction. My postdoc Deena Weisberg and I examined this tension by exploring whether children demonstrate fantastical thinking when they create and understand fictional stories. We asked 3- and 4-year-olds to to construct either realistic or fantastical stories by choosing whether an event that violated a real-world law or a normal, non-violation event should occur next. Three-year-olds were fairly random in their responses, but 4-year-olds showed a robust preference for choosing the ordinary events, regardless of whether they were creating stories based on a minimal realistic or fantastical context, continuing context-rich stories in a book, or continuing visually salient movie clips. These results suggest that preschoolers are strongly committed to affirming real-world laws, even in fictional contexts where these laws could possibly be broken. Preschoolers are not fantasy-prone when it comes to constructing fiction; they are reality prone!

We also examined how children treat differences between impossible and possible events. We constructed stories in which children had to compare impossible events with weird but possible events (events they have never seen, like playing soccer with a square soccer ball). Most research in the past has found that preschoolers treat improbable or strange events as impossible. We found, however, that 4-year-olds were able to tell the difference between genuinely impossible and these possible, novel, strange events. Taken together, we believe that children’s understanding of fiction is relatively sophisticated, particular in terms of what rules govern afictional world.

Our most recent work is about how children learn from fantasy characters – and specifically whether they learn differently from characters they are engaged with as opposed to not. Rachel Zolno, an undergraduate completing her senior thesis, is studying this idea using gender boundaries. Do girls learn better from princesses and boys from superheroes than the other way around?

To Read More

Sobel, D. M., & Weisberg, D. S. (2012).Tell me a story: How children’s developing domain knowledge affects their story construction. Manuscript submitted for publication, Brown University.

Weisberg, D. S., & Sobel, D. M. (2012). Young children discriminate improbable from impossible events in fiction.Cognitive Development, 27, 90-98.

Weisberg, D. S., Sobel, D. M., Goodstein, J., & Bloom, P. (in press). Young children are reality-prone when thinking about stories.Journal of Cognition and Culture.

Zolno, R., & Sobel, D. M. (in progress). Preschoolers’ ability to learn from gender-specific fantasy characters. Brown University

Deena Weisberg is now a Research Assistant Professor at the University of Pennsylvania.

Rachel Zolno is a senior, completing her honors thesis.

Other Announcements

The Lab is thankful to the National Science Foundation for its continued support. We have been supported by NSF grants DLS-0518161 and BCS-0744898 and are currently supported by grant DLS-1223777. Deanna Macris won a Galton Fellowship from the Institute of Brain Sciences at Brown, which supported her research last year.

The Lab congratulates Brianna Doherty, Katie Green, Naomi Heilweil, and Katherine Williams who all defended their senior thesis projects and graduated with honors. Brianna won a Rhodes scholarship and is now studying developmental neuroscience at Oxford University in England. Katie is a research assistant at the Psychiatric Neuroimaging Lab at Brigham and Woman’s Hospital in Boston. She’ll be applying to grad school next year. Naomi is now a graduate student in the developmental psychology program at Yale University. Katherine is now in medical school at the University of Maryland.