Models

Question 1

What is the difference between statistical and theoretical models?

Answer 1

Statistical- mathematical relationship between variables, that hold under specific assumptions

Theoretical models- a description of the relationship between different mental processes, that makes assumptions about the nature of these processes

Question 2

What is the difference between behaviourisma and cognitive science in terms of models?

Answer 2

• behaviourists not interested in the mental processes, only with input and output

Question 3

What are box and arrow models?

Answer 3

Cognitive science- used with what is going on between input and output, multi staged mind

Question 4

What are the 2 subcategories within theoretical models?

Answer 4

• informal and formal models
• formal are the mathematical description between mental processes, usually expressed through computer code

Question 5

What is meant by simplification and abstraction in models?

Answer 5

simplification- making model only contain simpler parts
abstracion- generating general rules and concepts for specific information (right level depends on the question we are asking and what we are trying to convey)

Question 6

What is meant by prediction and what can they be?

Answer 6

• predictions can be directional or numerical
• models that provide numerical predictions can be more or less accurate
• non-scientific theories explain after the fact but cannot provide falsifiable predictions

Question 7

How do we use models to predict and explain? 5 steps

Answer 7

• Framework- conceptual system (cog psych)
• theory- a scientific proposition that provides relation between phenomena (early-selection theory)
• model- (Broadbent’s model)
• hypothesis (irrelevant stimuli that contain target defining feature will automatically be detected)
• data (new ‘gorilla’ experiment, detection rates, t-test
• data

Question 8

What are some examples of explanation without exact prediction?

Answer 8

• models of scz can indicate causes but cannot yet predict individual cases
• the model be able to predict group differences, but not individual cases

Question 9

What is meant by prediction without explanation?

Answer 9

• some models can predict whether an individual will develop Alzheimers even though we aren’t close to understanding the factors that explain Alzheimers

Question 10

What are the strengths of formal models?

Answer 10

1. More accurate predictions- by having numberical stimulation, we can see if the model provides unreasonable predictions (easier to reject bad models), helps us select which experiments to perform, can provide a more subtle form of hypothesis testing by having numberical predictions
2. Counter-intuitive predictions- a model can more clearly describe which predictions follow from a model, with informal models its hard to notice when they make counter-intuitive predictions. Formal models clearly produce such predictions
3. benefits of explicit assumptions- by making assumptions explicit, we can reveal unanswered questions, flaws in our reasoning, contradictory or unreasonable assumptions

Question 11

What is evidence for counter-intuitive predictions?

Answer 11

• model assumes when people make a decision between leftward and rightward choice, they accumulate noisy evidence over time
• you would expect decision making to slow if more noise is added
• however, response time gets shorter (counter intuitive)
• formal models, because it gives specific predictions, can reveal these predictions whereas informal models can’t

Question 12

What are the cons of formal models?

Answer 12

• require expertise
• idea of transparency means only for experts
• prediction- sometimes numerical predictions are premature
• progress- changing the model is costly time wise, can limit progress
• theory (1)- a computational model may give the semblance of scientific validity (neural network models)
• theory (2)- making a model simulate a cognitive task doesn’t necessarily teach much about cognition

Question 13

What is meant by the hype timeline?

Answer 13

• stages when a new concept comes in eg. AI
• innovated trigger
• peak of inflated expectations
• trough of disillusionment
• slope of enlightenment
• plateau of productivity

Question 14

Who is Marr?

Answer 14

• British Mathematician 1945-1980, worked on visual processing
• asked how can we understand information processing systems like the brain
• said an algorithm is more likely to be understood more readily by understanding the nature of the problem being solved than examining the mechanism in which its embodied

Question 15

What does Marr propose on how to understand the brain?

Answer 15

• we can only ever hope to sample from a tiny sample of brains activity in a tiny fraction of a bit of brain. Therefore, the way to make sense of brain data was to break any brain problem into 3 levels:
  Computation- the problem being solved
  algorithms- the steps/rules to solve it
  implementation- the actual machinery

Question 16

What is a bottom up approach on neuroscience and AI?

Answer 16

• Implementation- AI technology, machine learning, LMM, neural network
• Then Rules- what can we study with this technology?
• Then problem- what do these algorithms tell us about cognition?

Question 17

What is a top down approach on problems in neuroscience?

Answer 17

• Problem- identify it
• Then rules- what representations and algorithms can solve this problem?
• Then implementation- how can these representations and algorithms be implemented in neural circuits

Question 18

Which approach did Marr prefer?

Answer 18

top-down