lab experiments Flashcards
practical
✓ Control to establish cause & effect – The logic of the experimental method is that the researcher can manipulate (alter) the variables in which they are interested, in order to discover what effect they have. By doing this, a researcher has a huge amount of control compared to other methods & this can allow them to establish a cause-&-effect relationship between two factors
.
× Difficulties in identifying & controlling variables: A lab experiment is only effective if all the variables that can influence the outcome can be identified & controlled. The complexity of social interactions with a large number of influences at work means that it is impossible to identify, let alone control, all variables. Human beings have free will, consciousness & choice which many argue means our behaviour cannot be explained in terms of cause & effect as it is much more complex.
× Limited application – As only small-scale interactions can be studied in a laboratory experiment, this excludes many of the most important sociological issues, such as large-scale social change (e.g. religions or voting patterns). It is also impossible to study past events, or events of long duration, using experiments.
× Time & cost: Time consuming method due to lengthy planning, application & analysis which also makes this a potentially costly method. Also, the artificial setting may have to be paid for, researchers trained, equipment hired, participants paid, etc.
× Personal skills: a researcher may need specific interpersonal skills to use this method effectively (e.g. to be a ‘people person’ if they are directly involved with participants).
ethics
× Raise a number of ethical issues: Many experiments involved some kind of ‘blind’ to conceal the real aims of the research from the subjects, so as to avoid this knowledge influencing their actions. This means that the researcher must often deceive participants with regards to the true nature of the study & therefore cannot obtain their informed consent. Another issue is the possible emotional & psychological effects some experiments may have on those involved.
reliability
✓ High in reliability: Once an experiment has been conducted, other researchers can then replicate it (repeat it in exactly in every detail). Laboratory experiments are therefore highly reliable for two reasons: (i) The original experimenter can specify precisely what steps were followed in the original experiment so other researchers can repeat these in the future. (ii) It is a very detached method: the researcher manipulates the variables & records the results, so their personal feelings & opinions should have no effect on the conduct or outcome of the experiment.
× Errors/Inaccuracies/bias: Despite the highly controlled environment, it is still possible for recording errors to be made &/or personal bias in connection to the outcome of the experiment to effect results.
validity
× Low in validity: A laboratory is a highly artificial environment & it is doubtful whether the results of experiments can be transferred to the real social world. How people react in an artificially constructed lab tells us very little about the way they act in real life situations.
× The Hawthorne effect (or experimental effect) – Even if the subjects are misled as to the real purpose of an experiment, their knowledge that they are in an experiment is likely to affect their behaviour. This reduces the validity of the experiment.
examples
- Stanley Milgram’s (1963) experiment on obedience to authority. The purpose of the experiment was to test people’s willingness to obey orders to inflict pain. Against all estimates, 65% of the participants were prepared to administer (what were actually fake) shocks of 450 volts to a stranger.
- Albert Bandura’s (1961) Bobo doll experiment. Set out to demonstrate Bandura’s social learning theory by testing how children’s behaviour would alter after watching an adult be aggressive towards a bobo doll. Bandura found that the children exposed to the physically aggressive model were more likely to behave aggressively themselves.
Representativeness
× Often not representative: The laboratory is a small place, so only small-scale social interaction can be studied. The participants are unlikely to a typical cross-section of the larger group as samples are often made up of a particular ‘type’ of person – e.g. university students or those willing & able to answer an advert to take part. This would then mean that generalisations cannot be made from the research.
theory
✓ Favoured by positivists: Argue that there is a measurable objective social reality ‘out there’. They take a scientific approach using standardised methods of research to obtain quantitative data that allows them to produce generalisations & cause-&-effect statements. They regard laboratory experiments as very reliable, because they can be repeated exactly, allowing previous findings to be checked. Experiments also meet the positivists’ requirement that data should be quantitative & scientifically collected.
× Rejected by interpretivists: Claim that experiments do not translate easily to the study of social behaviour & they produce data that is low in validity.
× However, even positivists rarely make use of them in sociological research due to a number of problems such as their artificiality & limited application to the wider social world.
