lecture 1- intro to methods and experimental designs

How do you run a scientifically sound psychological experiment?

• clear definition of concepts and constructs
• operational definition- procedures used to measure a variable
• large sample size
• well controlled

What is the empirical approach?

Evidence based method that draws on knowledge through the senses.

• wundt- introspection, subjective
• watson- behaviourism, observable, little albert
• neisser- father of cognitive psychology- use reaction times to infer about internal processes, objective

What is the scientific method?

• observation
• hypothesis
• research
• data analysis
• scientific theory

What is a hypothesis?

A scientifically testable prediction.
Criteria:

• operationalised/clearly defined
• non-circular
• deals with observable/measurable phenomena
• falsifiable

What is a theory?

Framework that organises explanations of data. Consists of general principles for outlining and understanding data, guides empirical investigation.
Should include;

• definition/ description
• explanation
• prediction
• be falsifiable.

Why is Asch's conformity study flawed?

• non-conformity was more prevalent than conformity but most textbooks don't report this
• biased sample of psych students- lacks population validity
• artificial task- low ecological validity

Why is Zambardo's prison experiment flawed?

• zimbardo himself took part in study
• small sample size (24)
• sample bias- all stanford students
• most guards didn't exhibit cruel behaviour- contrary to his conclusions
• prisoners undermined the guards
• his conclusions are more reflection of what he wanted to be true
• unethical

What do we need to infer causation?

• co-variation/ correlation- not sufficient alone. Beware of confounds.
• time-order relationship - cause must come before the effect
• eliminate other possible causes of behaviour

What is a natural group design?

• groups are not created by manipulating an IV
• based on a ppt variable
  -ready formed groups e.g. men vs women

What is a between-subjects design?

• independent groups- groups are made up of different people
• each group takes part in one condition
• difference in performance between groups/subjects
• groups can be natural (old vs young) or manipulated (treatment A vs B)

advantages of between subjects design

• Performance not influenced by boredom/fatigue- ppts only do one condition
• no practice effects
• conditions don't affect each other

What is a within-subjects design?

• repeated measure design
• all ppts complete all conditions
• repeatedly measure the same people on the same DV
• controls for individual differences e.g. natural memory ability
• can do it in one session or different times.

advantages of within subjects design

• ppt characteristics not a problem
• more powerful
• requires fewer ppts
• more convenientw

What is power?

The probability of finding a statistically significant effect i.e. correctly reject the null hypothesis.

• variation caused by the experimental manipulation vs error variance
• low error variance= more powerful

What is error variance?

Variation caused by individual differences.
Reducing error variance makes it more likely to find a real result.

What is nominal data?

Data that can be placed in categories (colours, Type A/B personality, types of attachment style).

What is ordinal data?

Data which is placed into order/ranked but the distances between the scores is not known. E.g. likert scales, age categories

What is interval data?

Distance between data points are at equal intervals. No true zero. E.g. temperature- distances are equal, can be minus numbers and ratio of numbers is not equal.

What is ratio data?

Same as interval data, but has a true zero and a true ratio. E.g. height, weight, score on a test.
Can get ratios from it - e.g. when driving at 60mph, going twice as fast as driving at 30mph.