International School for Advanced Studies, Trieste

Written Exam for Admission to the Cognitive Neuroscience PhD curriculum

October 6, 2005

 

Please answer/address three (3) of the following questions.

Note: Individual papers/experimental paradigms that are central to the way you address one question should not also be central in your response to others.

 

English is the language strongly preferred by the Commission.  Italian may be used if absolutely necessary.  

 

Please write clearly, and concisely. Length is not correlated with quality.

 

*          *            *            *            *

1)      Design a neuropsychological study that you would carry out in order to test a finding of animal neurophysiology, being it in the domain of action, space or memory.

 

2)      Tell us about a neuroscience publication that has had a major influence on you. Use the following format for your answer: (a) title and authors (accuracy not required); (b) the problem addressed in the paper, (c) what was known about the problem before; (d) the methods used to approach the problem; (e) the main results; (f) the authors’ interpretation of the results; (g) how this publication changed your way of thinking. Alternatively, you may do the same exercise for a lecture that you have attended.

 

3)      What experiments would you conceive to advance our knowledge of the uniqueness of language learning to the human species?

 

4)      On departing for a perilous journey through uncharted waters, you want to please as many ancient Greek gods as possible by offering them sacrifices. You are aware, though, of their extreme jealousy: each one of them would be irritated by your offering a sacrifice to one of their arch-rivals, and absolutely incensed if you offer it to two of them, no matter what you offer to the particular god’s friends. Therefore you will have to carefully weigh your choice of patrons. Given that there are 22 gods available and 22 archrival pairs, randomly distributed, produce a quick estimate of how many sacrifices you will be able to offer, without offending anyone, and detail the calculation.

 

5)      Discuss an example taken from any domain of cognition in which the evidence provided by neuroimaging nicely overlaps with that supplied by studying brain-damaged patients, and one in which there is great inconsistency between the two.

 

6)      Bayes’ theorem is commonly cited as a way to calculate the probability that you suffer a given disease, D, given a positive blood test, T: P(D|T). Bayes’ theorem can also be used to evaluate how sensory neurons represent the external world by their firing; here one would want to solve for P(E|F), where E and F are time varying: E is external state of the world and F is the firing of neurons). (a) Why is it interesting to compute P(E|F)? What does it mean? (b) Write out an expression of Bayes’ theorem solving for P(E|F) using the “prior” P(E), P(F|E), and P(F); (c) Explain the application of the above equation (b) to some artificial, laboratory study of processing in a sensory modality you select. In other words, explain what the terms in (b) refer to in your laboratory experiment. (d) Now try to generalize the experiment (c) to more natural stimulus conditions, changing the terms to ones with real meaning. (e)  Then, explain some of the problems neuroscientists face between (b) and (c), that is, generalizing Bayes’ framework from artificial to natural conditions. If you are more familiar with some framework other than Bayes’ (for example, Mutual Information), use your preferred terminology to answer these questions in a clear, consistent manner.

 

7)      Describe the salient characteristics of Darwinian adaptation, using your own simplified Ising spin model (i.e., strings of +/-1 variables) of DNA .

 

8)      Describe a study to compare and contrast full-term and premature infants to evaluate the influence of biology and environment on early cognitive development.

 

9)      In what way does the neuronal representation of the outside world (visual scenes, faces, things heard or touched etc.) differ in "higher-level" (also called "late" or "downstream" or "association") cortical areas as compared to the "lower-level" cortical areas in the same processing stream?  Try to support your assertions from some evidence in the literature.

 

10)  As a prisoner in the highest tower of the castle of Monfort, the only way you can communicate with your beloved one is by dropping from a high, narrow opening the oak leaves you have been offered for your toilette comfort. Such leaves are all green, but maybe you can wrinkle some, or dye them red with your own blood, to produce discernible symbols. Devise a sort of Morse code, and estimate the length of the love poem, in your favourite language, that you will be able to transmit with your provision of 300 leaves.

 

11)  Over the last 15 years, functional imaging has become an increasingly popular technique for investigating higher cortical functions. Take one imaging study that clarified the neural mechanism(s) underlying a cognitive operation that could not be explained before using simple reaction times experiments.

 

12)  What experiments would you conceive to advance our knowledge of how humans process numbers?