Human Cognition

- Wilhelm Wundt established the 1st psychology lab (1979) but used the method of introspection.

              - Analysis of the structure of own mental processes as they occur

              - Subjective self-report - cannot be measured or observed

- Sigmund Freud (1856-1939) proclaimed his work to be the scientific study of human nature.

              - Critics claimed his approach was unscientific because his theory was not falsifiable in that it made no predictions.

- In the middle of the 20th century, behaviourism emerged as a reaction agaisnt introspection and became the dominant approach in psychology.

- Behaviourists researched only that which was directly observable (S-R behaviour) and dismissed introspection as unscientific.

- Aimed to develop psychology as a science by adopting objective/empirical methods.

- In the early 1980s cognitive psychology became the dominant theoretical approach.

- Cognitives psychologists claim that the mind can be studied scientifically.

- Aimed to understand human cognition by the study of behaviour.

              - Human cognition = processes such as attention, perception, learning, memory, language, problem solving, reasoning and thinking.

- Uses a variety of approaches; each serving different purposes and having own strengths and weaknesses.

              - Experimental cognitive psychology

              - Cognitive neuropsychology

              - Computational cognitive science

              - Cognitive neuroscience

- Distinction between approaches not clear-cut; research often combines elements of more than one approach, this enhancing understanding of human cognition and behaviour.

- Aims to understand human cognition via behavioural evidence.

- Experiments on healthy individuals under controlled lab conditions

           - Performance/behavioural responses (accuracy/RT)

           - E.g. attention, memory, language, problem solving.

- Strengths

           - Rigorous design to test specific hypothesis

           - Theories makes specific predictions

- Limitations

           - Ecogological validity (lab vs. normal performance). E.g. search, driver behaviour

           - Behavioural data only provides indirect evidence of underlying process.

- Study of brain-damaged patients as a way of understanding normal human cognition

- Aim to find double dissociations:

           - Patient A: Task 1 normal performance, task 2 impaired

           - Patient B: Task 1 impaired, task 2 normal performance

Patient

Performance

Task 1

Task 2

A

ü

û

B

û

ü

Strenghts:

- Causal links can be shown between brain damage and cognitive performance

- Very influential for modularity of function idea

Limitations:

- Dependence of serendipity of nature

- Single case studies or small n

- Compensatory strategies may mask impact of brain damage

- Neural plasticity; connections might modify

- Damage often affects multiple areas to difficult to disentangle and make sense of findings.

- Internal double dissociations not possible

- Rarely have pre-damage control measures

- Aims to develop computational models to enhance understanding of human cognition and behaviour

Two main types:

- Production systems

         - Consist of a number of "IF... THEN" production rules and a working memory containing information

         - Match working memory content agaisnt IF rules --> execute THEN rules

- Connectionist networks:

          - Also called neural networks or parallel distributed networks

          - Consist of connected nodes and various layers

          - Nodes = neurons and connections = axons

- The networks model cognitive performance by strong patterns of activation in the network.

Strengths:

- Theoretical assumptions are spelt out in precise detail

- Empirical evidence and functional neuroimaging data support underlying assumptions of the models.

Limitations:

- Only rarely used to make new predictions

- Ignore the motivational and emotional factors present in human cognition.

- Uses evidence from brain and behaviour to understand human cognition

- Techniques for studying links between cognitive processes and brain structure/functions

- Uses methods of measuring brain activity:

           - fMRI: Function Magnetic Resonance Imaging

           - MEG: Magnetoencephalography

           - Single cell recording

           - TMS: Transcranial Magnetic Stimulation

- Measures magnetic resistance in brain atomd

- Activity shown in BOLD signals ("blood oxygenation level development")

- Good spatial resolution

- Indirect measure of neural activity

- Statistical maps

- Statistical limitations

             - Under-correction

             - Under-powered

- Association, not necessity.

- Brain activity (ERP: event-related potentials) recorded at surface of scalp

- Excellent temporal resolutions but very limited spatial resolution

- Many trials/subject to artefacts

- Similar to ERG

- Records  brain activity via magnetic rather than electric fields

- Excellent temportal resolution

- Expensive and difficult to maintain/use.

- Allows direct measurement/stimulation of human cortex with great precision

         - Invasive

         - Limited to patients requiring neurosurgery

         - Stress and medications affects performance

         - Time constraints limit the experimental paradigms

         - Retesting is not possible

- Uses electromagnetic pulse to introduce disorder into a normally-ordered system

- Excellent temporal and good spatial resolution

- Shows causal links between brain and behaviour

- Can test functional connections and timing

- Avoids plastiticity problems

- Allowed pre- and re-testing

- Can only stimulate cortical surface - in regions away usually away from muscles

- Precise localisation is uncertain

- Can be a bit unpleasant

- There are risks: Patients/Seizures

Strenghts:

- Great variety of techniques offering excellent temporal or spatial resolution

- Can use some techniques simultaneously to allow assessment or integrated brain processing

Limitations:

- Functional neuroimaging techniques provide essentially correlation data

- Restrictions on the tasks that can be used, particularly in scanners.