Different Research Methods
Laboratory experiments - takes place in a controlled environment and enables the experimentor to test the effect of the IV (independent variable) on the DV (dependent variable). In order to establish a difference and so detect cause and effect relationships, the IV is systematically varied between to conditions. This is good because cause & effect can be established and it can be replicated. However, it is not a real life situation and demand characteristics may be involved.
Field experiments - take place in a natural setting, e.g. a work environment. The experimenter has control of the IV and so casual relationships can be established. This is good because it is ecologically valid and can be generalised. However, field experiments are time consuming and expensive.
Quasi-experiments - exist when the experimenter cannot control the IV; it is said to be naturally occuring. E.g. experiments involving age, gender, class or cultural differences would all be clased as quasi-experiments because the experimenter cannot manipulate any of these as the IV. However, the experimenter has control of the setting. These are good because they are easy to set up and they can be generalised. However, it is hard to get rid of confounding variables and they are time consuming
Natural experiments - a kind of quasi experiment, but the researcher has no control over the IV or the research setting. These are good because they are high in ecological validity and demand characteristics do not exist. However, the IV cannot be directly manipulated and they require ethical sensitivity.
Correlational analysis - a technique that measures the strength of the relationshop between two variables. The paired scores of two variables are analysed to establish the strength and direction of the association. This is good because it is a useful pointer for further research and it calculates the strength of a relationship. However, you cannot assume cause and effect and it could be non-linear.
Observational techniques - involves examining behaviour in a natural setting with minimal intrusion from the researcher as it aims to observe natural behaviour. Participants may be aware they are being observed (overt) or not (covert). They are good because they are practical in studying small children and animals and when covert there are no demand characteristics. However, they lack ecological validity and when covert there are ethical issues.
Interviews - can take many different forms: non-directive, informal, guided, clinical or structured. They usually take place face-to-face and can yield rich, in-depth data. They are high in validity and are a primary source of data. However, they lack validity and can contain interviewer bias.
Questionnaires - a type of interview. Can be conducted face-to-face, via telephone or by post. They consist of a standard set of questions that are either closed or open. Questionnaires are used to survey attitudes, beliefs and behaviour. They are reliable and representative if a good sample size is used. However, they lack validity and usually have a low response rate.
Case studies - in-depth study of an individual or small group. They are good because they gather extensive information and the data can be both qualitative and quantative. However, only one case is involved and so it cannot be generalised and they can be very time consuming.
Independent groups design - Two seperate groups try one condition each. e.g. "are blondes less intelligent than brunettes" - there is one group of blondes and one of brunettes and each do an intelligence test once. ADVANTAGES - no demand characteristics and not time consuming. DISADVANTAGES - individual differences.
Repeated measures - One group tries both conditions. e.g. "the effect of alcohol on reaction times" - one group would test their reaction times before alcohol, and again after alcohol. ADVANTAGES - no individual differences and less participants are needed. DISADVANTAGES - participants could get better at the task or could get fed up, demand characteristics and time consuming.
Matched pairs - Participants are carefully matched up with a similar other person - closest to them in terms of individual differences. e.g. "the effect thatp raise and positive feedback has on improving performance in a maths test". ADVANTAGES - eliminates individual differences. DISADVANTAGES - time consuming to match pairs.
Validity And Reliability
Validity - the extent to which a test measures what it is supposed to measure.
Case studies, Unstructured interviews, Semi-structured interviews, Naturalistic observation and Field experiments are all high in validity.
Reliability - the extent to which a test is repeatable and yields consistent scores.
Lab experiments, Structured interviews, Questionnaires, Content analysis and Correlations are all high in reliability.
Making A Hypothesis
Hypothesis - A prediction made by a researcher prior to an experiment stating what they expect they may find from their results.
There are two types of hypothesis:
Directional hypothesis - A prediction which states exactly the link expected to be found, rather than just stating that a link will be present. e.g. "blondes are less intelligent than brunettes"
Non directional hypothesis - A prediction that only states that a link will occur, but does not say which way the link will go. e.g. "intelligence levels differ between blondes and brunettes"
Positive correlation - when one variable increases so does the other
Negative correlation - when one variable increases the other decreases
ADVANTAGES - can calculate a relationship & it is a good pointer for further research
DISADVANTAGES - cannot establish cause and effect