- Experiments
Independent variable = variable that is changed or controlled in experiments to test the effects on the dependent variable
Dependent variable = variable being tested and measured in experiments
Extraneous variables = variables which are not the IV but could affect the results of the experiment
- Types of extraneous variables
-Situational variables = aspects of the environment that may affect behaviour, e.g. temperature, lighting, etc.
These can be controlled so they are the same for all the participants = Standardised procedures
-Participant variables = the way each participant varies from each other and how this could affect the results, e.g. mood, intelligence, anxiety, nerves, etc.
These can be controlled by using random allocation to the conditions of the IV
- Experimenter/Investigator effects
Experimenter unconsciously conveys to participants with accidental clues as to how they should behave = Experimenter bias
Personal characteristics of the experimenter can also affect the behaviour of the participants
- Demand characteristics
The clues in experiments that hint the purpose to the participants.
P's can be affected by their surroundings, the researcher's characteristics, the researcher's behaviour and their interpretation of what's going on in the situation.
- Laboratory experiments
Experiments that are highly controlled and conducted in lab settings
+ high control so less extraneous variables
+ easily replicable = high reliability
+ shows cause and effects well
- low ecological validity
- demand characteristics
- ethically sound because no deception and stress
- Field experiments
Natural, not artificial like a lab experiment
+ high ecological validity
+ minimal demand characteristics
- low control of variables
- low reliability
- ethically sound because there is consent, deception, privacy, and debriefing
- Quasi experiments
Makes use of existing natural IV's, usually in a lab
+ natural IV's so not expensive
+ high control = less extraneous variables = shows cause and effect well
- low ecological reliability
- low replicability
- demand characteristics
- ethically sound because no deception or stress
- Observations
Observing and recording behaviour, looking for patterns and making sense of them, usually in a natural environment = naturalistic observation
Participant = researcher is part of the group being observed
Non-participant = researcher is not part of the group being observed
Overt = participants are aware of the study
Covert = participants are unaware of the study
Structured = designed coding frameworks/charts for recording
Unstructured = non-specified, a wide range of behaviours relevant to the study
Controlled = conditions are controlled in the experiments
+ different views of behaviour that can't be found in experiments
+ good as preliminary investigations
+ captures spontaneous behaviour better
+ high ecological validity
+ good inter-observer reliability
- can be subject to observer bias
- poor behaviour categories can mean lower reliability
- ethical issues, deception, privacy and no consent
- demand characteristics
- social desirability (saying what seems socially appropriate)
- Self Reports
-Interviews, questionnaires or surveys to gain data from p's
-Questionnaires = generally structured and often contain fillers and lie detectors to raise validity, a researcher isn't always present
-Structured interviews = predominantly closed questions with a fixed order, usually scripted and standardised
-Semi-structured interviews = fixed list of open or closed questions
-Unstructured interviews = starts with standard questions, but then the questions start depending on the answers
+ lots of standardised data from lots of people = generalisable to society
+ highly reliable
+ qualitative data = rich in detail
- structured interviews and questionnaires lack flexibility
- unstructured interviews = low reliability
- demand characteristics + social desirability
- Types of reliability
External reliability = same results will occur in a different situation with the same people
Internal reliability = consistency of questions, consistency within an experiment
Split-half reliability = if we can compare 2 halves of a result and it's similar
Inter-rater reliability = if more than 1 researchers compare results and they correlate positively
- Types of validity
Internal validity = trueness of the test or measure
Face validity = whether the test appears to measure what it's supposed to
Criterion validity = whether a test seems like it can predict an outcome
Concurrent validity = whether a procure produces a similar score for different p's
Population validity = Extent that findings can be generalised to population or cultures
Ecological validity = whether findings can be generalised to real life situations
- Correlations
Collect data, then 2 sets are measured against each other to establish a relationship between variables
Positive correlation = 2 variables increase each other or one increases the other
Negative correlation = one variable increases but the other decreases
No correlation = no relationship found between the variables
+ quantifiable measures so easy to compare/analyse
+ gives information about the relationship between variables
+ a strong negative or positive correlation can form a basis for more research
- cause and effect cannot be proved or established