Final Assessment
Instructions
• Prepare short answers to these questions at home
•There are 10 questions. Each is marked out of 10. Final mark out of 100 will equal
your total score/100.
•Number your answers but do not waste word count by copying the question text or writing it out again in your answer.
•Use in-text citations where required [e.g. “Lagnado and Sloman (2002)” or “(Lagnado & Sloman, 2002)”].
• Include a single reference section at the end of your submission if necessary. For
this, you need only include any texts you refer to that are not cited anywhere in
the course materials.
• Your submission should be no longer than 2000 words, excluding the reference section but including in-text citations.
• Use your own words and keep things as simple as possible! I am looking for evidence of understanding these concepts rather than ability to copy paste text from slides.
• Submit your answers on TurnItIn.
1. Briefly explain how the Blicket detector paradigm probes children’s causal thinking, and describe one of the empirical results that have resulted from this paradigm.
2. Why are cases of overdetermination a challenge for dependency theories of causation?
3. Temporal contiguity, reliability, predictability, and order have all been shown to shape causal judgments in human adults. Pick one of these temporal factors and describe how a relevant experiment showed this.
4. What would be an efficient strategy for determining which one switch on a basement fusebox controls an upstairs bathroom light? Describe the steps of the strategy and explain why it is efficient. Can you think of any situation in which it would be suboptimal?
5. Using an example, explain why distinct hypotheses about causal structure cannot always be distinguished with interventions that only fix a single component of the relevant causal system at a time. What intervention strategy would allow you to distinguish hypotheses in this case and how does it do so?
6. What do each of the four experiments in Kemp, Goodman & Tenenbaum (2010) demonstrate about how priors shape our causal judgments?
7. How did Bramley, Mayrhofer, Gerstenberg & Lagnado (2017) test active learning in a continuous time setting? In their experiment, were participants better at using interventions to identify the structure of cyclic or acyclic causal systems in this experiment? What factors might explain this difference and why?
8. What combinations of base rate and causal power are probable under the generic causal priors proposed by Lu et al (2008)? Relative to Lu et al’s proposal, what differences did Yeung and Griffiths (2015) observe in their empirical estimation of people’s priors? What, if anything, does this tell us about causal cognition?
9. What is the difference between a statistical causal claim and an actual causal claim? Explain why one is more relevant for choosing between possible actions, and the other more relevant for attributing responsibility.
10. Describe a behavioural study you learned about outside this course that finds evidence for a causal relationship between two or more variables. What steps did the researchers take in designing the experiment to ensure their conclusions are reliable and generalisable? Do you think they were successful?
