you are expected to understand features of investigation design - the planning and construction of research methods.
RESEARCH METHODS can be defined as:
"the means by which explanations are tested"
Quantative and Qualitative
- involved a measurement of numerical quantity
- e.g. the amount of 'yes' or 'no's' answered in a questionnaire
- give participants the option to express & describe their answer more fully
- give a more in-depth answer
- e.g. during an interview, the participant may be asked for their opinion of a controversial subject
Laboratory Experiments take place in a controlled environment with random members of the public as participants.
- the equipment is precise and reliable
- the experiment can be repeated to check accuracy
- it is unrealistic and artificial - cannot be applied to real life situations
- the observers may be biased - 'experimental bias'
Field Experiments are conducted in a natural environment, where the independant variable is manipulated. (for example, something may be place on a street to test reactions etc)
- the participants behaviour is natural - high ecological validity
- no demand characteristics as participants are unaware they are expected of anything
- can't control surrounding activity (e.g. weather)
- difficult to replicate - conditions are never exactly the same again
Natural Experiments are when the experiment is taken into natural surroundings, where the independant variable occures naturaly with no manipulation.
- results can be generalised as natural behaviour
- no demand characteristics as participants are unaware
- ethical issue - don't have the participants consent
- can't replicate the experiment
Correlational Analysis measures the relationship behaviour between two variables to find a trend or pattern (positive, negative or no correlation), commonly on a graph.
- allows predictions
- easy to do (no manipulation of data)
- only works for linear relationships
- only works for extraneous relationships
Naturalistic Observations is when you measure naturally occuring observations - observers a voluntarily involved.
- high external validity - can be generalised
- practical method, no manipulation of people
- no control over other variables
- observers could be biased
Interviews are 'face to face conversations', and they can be formal or informal.
- complex issues are best dealt with face to face, any misunderstandings can be sorted or explained
- easy to replicate
- some answers/responses may not be truthful
- interviewers have to be trained
Questionnaire Surveys are 'self-reports', participants record their own answers to a pre-set list of questions.
- easy to replicate (can just print off x amount of copies)
- takes large samples
- low response rates
- biased samples (only a certain group of people may respond)
Case Studies are in-depth, detailed investigations of one individual or a small group.
- information relates to a single, real person
- provides a great depth of understanding
- no two case studies are a like, so they cannot be generalized
- they are dependent on participants having a clear and full memory
Independent Variable (IV) - the variable that is directly manipulated to see the effect it has on the DV
Dependent Variable (DV) - the variable that is measured, it is dependent on the IV
Extraneous Variables (EV) :D - variables that may have an effect on the DV, that can controlled (e.g. age of participants, time limits etc). we like EV's!
Confounding Variables (CV) :( - variables that may have an effect on the DV that can't be controlled (e.g. weather, mood of participants etc). we don't like CV's, they're 'confusing'
Reliability and Validity
Pilot Study: 'small' scale practice run of experiment
Participant Variables: confounding effects from characteristics of participants
Situational Variables: confounding effects as a result of environmental influences
Experimenter Variables: confounding effects from experimenters themselves (if they are biased etc)
Reliability and Validity
Internal Reliability: within a single experiment, being consistent
External Reliability: getting the same results every time - compare with other studies
Internal Validity: the experiment is measuring what it's supposed to be measuring (no experimenter effects, no demand characteristics, use of standardised instructions)
External Validity: results can be applied to other people, time and place (the wider population)
Ecological Validity: the environment reflects the task (it's realistic)