Doing Psychology Flashcards
Quantitative Methods
focus on development and testing of explicit (formal) theories which can be used to make mathematical predications which can be tested by collecting and analysing data.
Qualitative methods
focus on development of verbal theories (more open ended and explanatory)
What is the goal of the quantitative method?
The development of formal theories that can be tested
What is a theory?
A principle (or set of principles) that explain a body of facts. A good theory is one that specifies the (causal) relation between states.
- goal = understanding
- often expressed as an explanation of a system
- must predict or it is not useful.
It is NOT a description, set of data or a diagram
Science of the mind (David Marr, 1980) - 3 answers to ‘how does it (the mind) work?’
In order to understand a system (or model), you must be able to describe it on all three of these levels:
- Computational Theory
- Representation and Algorithm
- Physical implementation
- Computational Theory (of the mind)
What is the problem being solved?
- what are the constraints on the solution?
- what is the nature of the problem/function being computed
(Usually a mathematical expression)
- Representation and Algorithm (of the mind)
How are we solving the problem?
- what info does the system represent?
- how is it being represented?
- what is the input/output?
- what are the steps in between?
Generally, we study cognition at this level as level 1 is too abstract and level 3 is too complex
- Physical implementation (of the mind)
How is this (the problem/solution) realised in the physical brain?
- how are the representations and algorithms realised in the hardware device (the brain) itself?
- Computational theory (of a model)
Specifies a function mapping input state to output state - what are the mechanisms?
does not say ‘how’ the input to output occurs
- Representation and Algorithm (of a model)
specifies representations for input and output (e.g. content and format) - precise series of operations
- Physical Implementation (of a model)
how does the model occur in the real world?
how could it be implemented or understood?
Process models
these models are most common in cognitive and behavioural neuroscience. There are two broad classes:
- symbolic
- connectionist
Symbolic (process) models
Symbolic Representations
Symbolic data structures have basic (or atomic) elements and rules for composing elements to make more complex structures e.g. words and grammatical rules. Any variable is a legal proposition which can be combined by an operator.
Symbolic Processes
We can apply symbolic rules or operations to a data set. In general, cognition is treated like a traditional computer programme (variables and operators run to produce an outcome).
Production systems (symbolic models)
Production systems are prototypical symbolic models with three components:
1. data base = the knowledge the system has
2. inference rules = rules the system knows e.g. if x is larger it is heavier (the rules can produce incorrect answers)
3. executive control structure = how the data base and the rules interact/ decides which rules fire when / requires a specific algorithm
Operation of a production system (symbolic models)
Current state = current contents of the database/facts known about the system e.g. current state of a chess board
State space = the set of all possible states e.g. all legal pieces on chess board
Goal state = the state you want your database to be in e.g. check mate
State transition = moving from one state to another
Search = algorithm for travelling the state space and finding the best path for moving from the current state to the goal state (decide which rules fire in which order)
