section 2: methods

• Laboratory experiments take place in a controlled setting using standardised procedure. They are easy to replicate and control extraneous variables, however have low ecological validity and lack mundane realism.
• Field experiments are done in a natural environment where participants are unaware of being in a study. It has higher levels of mundane realism and reduced participant effects however they are impossible to exactly replicate and lack informed consent which is an ethical issue.
• Quasi experiments have some but not all features of a real experiment. Randomisation of participants is not possible because participants are already assigned to a condition due to a personal characteristic e.g. gender.
• Natural experiments have a naturally occuring independent variable. They are typically the only way to test certain phenomena because it would be impossible or unethical to cause them otherwise. Strengths include high levels of mundane realism and fewer demand characteristics and weaknesses include a lack of definite cause/effect relationships and difficulty in replication and generalisation.
• A case study is a study of an individual or small group of people due to a unique or unusual circumstance. Often they use both qualitative and quantitative data, are longitudinal and rely on retrospective information.
• A strength of case studies is that they allow research to be done on individuals whose circumstances could not otherwise be manufactured for research, e.g. it would be unethical to purposefully cause the brain damage that Clive Wearing suffered from. A weakness is that equally, they are incredibly difficult to replicate, and because only very small samples are used it is hard to avoid the influence of participant variables.
• Covert vs. Overt observation:

  • in a covert observation participants are observed without their knowledge which means behaviour is natural and unmanipulated. This rules out demand characteristics but does not have informed consent so is unethical
  • in an overt observation participants know that their behaviour is being observed and recorded. This is more ethical, but social desirability bias and demand characteristics are more likely.
• Participant vs. Non-participant observation:

  • in participant the observer is a part of the group and interacts with them. This means participants may behave more openly and naturally however deception is used and the observer's perception of participants behaviour is more likely to be biased.
  • in non-participant the researcher remains separate from the participants. This means they are more likely to remain objective but cannot build up a rapport with the participants.
• Naturalistic vs. Controlled observations:

  • in naturalistic a researcher observes natural behaviour with no manipulation or intervantion. These are more reflective of people's true behaviour but are less reliable and cause and effect relationships cannot be established.
  • in controlled behaviour is studied in a controlled and structured environment, e.g. Mary Ainsworth's Strange Situation. This is easily replicable and data collected is easier to analyse as it is quantitative however they can lack validity because of demand characteristics.
• In order to collect quantitative data for observed behaviour, behavioural categories must be operationalised into target behaviours. This can be done via time sampling (what is the participant doing at every interval of a certain time e.g. 5 seconds) or event sampling (a tally is taken for every instance of a behaviour).
• Questionnaires are a form of self-report technique where participants answer written questions. These could be open, closed (yes/no) or on a scale. They are easily repeatable so a large pool of data can be collected however answers may be affected by leading questions or social desirability bias.
• Interviews are a self-report technique that involve a participant a set of verbal questions face to face. They can be structured with set answers which produces quantitative data or unstructured which produces qualitative data. They can provide more detailed information but can be affected by leading questions from the interviewer, however this can be mitigated by using pilot studies.