Week 2 doing social psychology Flashcards
What is the scientific method and why is it used in social psychology?
method of research in which a problem is identified, relevant data is gathered, a hypothesis is formulated from this data, and the hypothesis is empirically tested.
Social psychologists believe that a true understanding of the causes of social behaviour can only be obtained through a systematic scientific approach, and that is why they conduct scientific research. Social psychologists believe that the study of social behaviour should be empirical—that is, based on the collection and systematic analysis of observable data.
What are a hypothesis and a theory? What does each attempt to do?
Theory
Set of interrelated concepts and
principles that explain a phenomenon.
Empirically testable predictions about what goes with what, or what causes what.
The scientific method dictates that no theory is ‘true’ simply because it is logical and seems to make sense, or because one simply believes it to be true. On the contrary, the validity of a theory is based on its correspondence with publicly verifiable fact. So, social psychologists develop hypotheses or predictions based on theories or past observations. They then collect data to test if the hypothesis is correct.
What are the two types of research methods used in social psychology, and what are their characteristics?
Laboratory experiments
high internal validity (control the independent and dependent variables
Field experiment High external validity but confounding variables.
Be sure that you can define the terms ‘independent variable’ and ‘dependent variable’. This is important to know about psychology in general.
Independent variables
Features of a situation that change of their own accord, or can be manipulated by an experimenter to have effects on a dependent variable.
Dependent variables Variables that change as a consequence of changes in the independent variable.
Your text discusses three sources of bias in experiments. Identify these and provide an explanation for each.
Subject effects Effects that are not spontaneous, owing to demand characteristics and/or participants wishing to please the experimenter.
can cause participants’ behaviour to be an artefact of the experiment rather than a spontaneous and natural response to a manipulation. Artefacts can be minimised by carefully avoiding demand characteristics (Orne, 1962), evaluation apprehension and social desirability
Demand characteristics Features of an
experiment that seem to ‘demand’ a certain response.
Experimenter effects Effect that is produced or influenced by clues to the hypotheses under examination, inadvertently
communicated by the experimenter.
The experimenter is often aware of the hypothesis and may inadvertently communicate cues that cause participants to behave in a way that confirms the hypothesis. This can be minimised by a double-blind procedure, in which the experimenter is unaware of which experimental condition they are running.
Make sure that you understand that laboratory experiments are intentionally low on external validity (mundane realism) but high on internal validity (experimental realism). Be able to define these terms.
this controls confounding variables that would impacts the effect of the IV on the DV in the normal social environment.
in a laboratory in order to be able to control as many potentially confounding variables as possible. The aim is to isolate and manipulate a single aspect of a variable, an aspect that may not normally occur in isolation outside the laboratory. Laboratory experiments are intended to create artificial conditions.
. Laboratory experiments are intentionally low on external validity or mundane realism (i.e. how similar the conditions are to those usually encountered by participants in the real world) but should always be high on internal validity or experimental realism (i.e. the manipulations must be full of psychological impact and meaning for the participants)
Mundane realism Similarity between circumstances surrounding an experiment and circumstances encountered in everyday life.
Internal validity or Experimental realism Psychological impact of the manipulations in an experiment.
Imagine you wish to conduct two experimental methods: (a) the laboratory experiment; and (b) the field experiment. For each method, ensure that you can: (i) describe the method; (ii) provide a research example that might use it; (iii) list the advantages of using the method; (iv) list the disadvantages of using the method.
Labatory experiment would be in a controlled environment out of the natural setting with a clear IV something that is manipulated to effect the DV. Their could be any number of conditions depending on the number of IVs. advantages are that is controls confounding variables (strong internal validity) , disadvantages are that it cannot be generalized to real setting (weak external validity).
in a field experiment the external validity (participants usually unaware of the experiement)
However, there is less control over extraneous variables, random assignment is sometimes difficult, and it can be difficult to obtain accurate measurements or measurements of subjective feelings (generally, overt behaviour is all that can be measured).
Why do researchers use non-experimental methods?
Doing an experiment is usually the preferred research method in social psychology. However, there are circumstances where it is very difficult or impossible to properly
test a hypothesis experimentally. For example, theories about the relationship between biological sex and decision making are not amenable to experimentation – we cannot manipulate biological sex in an experiment and see what effects emerge. Social psychology also confronts ethical issues that can prevent experimentation. For instance, hypotheses about the effects on selfesteem of being a victim of violent crime are not at all easily tested experimentally – we would not be able to assign participants randomly to two conditions and then subject one group to a violent crime and see what happened!
What sorts of conclusions can be drawn from these methods? Be sure that you understand the distinction between correlation and causation.
Take an example. Suppose we try to compare the selfesteem of people who have been victims of violent crime with those who have not. We might like to think that any differences will be due to violent crime, but unfortunately they could be due to other differences between the two groups. We can only conclude that self-esteem and being the victim of violent crime are correlated. There is no evidence that one causes the other – being a victim may lower self-esteem or having lower self-esteem may increase the likelihood of becoming a victim. It is also possible that there is actually no causal relationship at all – a third variable such as chronic unemployment might both reduce self-esteem and expose one to violent crime. In general, nonexperimental methods involve the examination of correlation between naturally occurring variables and as such do not permit us to draw causal conclusions.
Causal conclusion would have to prove that one variable directly impacts another variable. not that either could impact each other.
Your text identifies five non-experimental methods: (a) archival research; (b) case studies; (c) discourse analysis; (d) survey research; (e) field studies. For each method, ensure that you can: (i) describe the method; (ii) provide a research example that might use it; (iii) list the advantages of using the method; (iv) list the disadvantages of using the method.
Archival research is a nonexperimental method that is useful for investigating
largescale, widely occurring phenomena that may be remote in time – for example a stock market crash, or a disastrous decision made by a government, such as groupthink, an idea explored by Irving Janis (1972; see Chapter 6). The researcher has to make do with whatever is there – assembling data collected by others, often for reasons unconnected with those of the research project. Archival methods are often used to make comparisons between different cultures or nations regarding things such as suicide, mental health or childrearing strategies. Archival research is of course not subject to demand characteristics, but can be unreliable because the researcher has no control over the primary data collection, which might be biased or unreliable in other ways (e.g. missing vital data).
Case studies are another nonexperimental method. They involve indepth anal ysis of a single case (a person, a group or an event) and are well suited to the study of unusual or rare phenomena that could not be created in the laboratory: for instance, mass murderers, bizarre cults or terrible disasters. Case studies employ a variety of data collection and analysis techniques involving structured and open ended interviews and questionnaires, and the observation of behaviour. They are useful as a source of hypotheses, but findings can be compromised by the bias of the researcher, who is not blind to the hypothesis. Another issue is participant bias, as well as evaluation apprehension that people feel when they are the object of attention. Finally, findings of a case study may not easily be generalised to other cases.
A variant of the case study is discourse analysis, an approach that is popular in Australia and New Zealand and is associated with Margaret Wetherell and other Case studies are another nonexperimental method. They involve indepth anal ysis of a single case (a person, a group or an event) and are well suited to the study of unusual or rare phenomena that could not be created in the laboratory: for instance, mass murderers, bizarre cults or terrible disasters. Case studies employ a variety of data collection and analysis techniques involving structured and open ended interviews and questionnaires, and the observation of behaviour. They are useful as a source of hypotheses, but findings can be compromised by the bias of the researcher, who is not blind to the hypothesis. Another issue is participant bias, as well as evaluation apprehension that people feel when they are the object of attention. Finally, findings of a case study may not easily be generalised to other cases.
A variant of the case study is discourse analysis, an approach that is popular in Australia and New Zealand and is associated with Margaret Wetherell and other
Survey research Method in which a large and representative sample of people answer direct questions about their attitudes or behaviour.
what is a confounding variable?
Where two or more independent variables covary in such a way that it is impossible to know which has caused the effect.
Your text identifies five non-experimental methods: (a) archival research; (b) case studies; (c) discourse analysis; (d) survey research; (e) field studies. For each method, ensure that you can: (i) describe the method; (ii) provide a research example that might use it; (iii) list the advantages of using the method; (iv) list the disadvantages of using the method.
Archival research is a nonexperimental method that is useful for investigating
largescale, widely occurring phenomena that may be remote in time – for example a stock market crash, or a disastrous decision made by a government, such as groupthink, an idea explored by Irving Janis (1972; see Chapter 6). The researcher has to make do with whatever is there – assembling data collected by others, often for reasons unconnected with those of the research project. Archival methods are often used to make comparisons between different cultures or nations regarding things such as suicide, mental health or childrearing strategies. Archival research is of course not subject to demand characteristics, but can be unreliable because the researcher has no control over the primary data collection, which might be biased or unreliable in other ways (e.g. missing vital data).
Case studies are another nonexperimental method. They involve indepth anal ysis of a single case (a person, a group or an event) and are well suited to the study of unusual or rare phenomena that could not be created in the laboratory: for instance, mass murderers, bizarre cults or terrible disasters. Case studies employ a variety of data collection and analysis techniques involving structured and open ended interviews and questionnaires, and the observation of behaviour. They are useful as a source of hypotheses, but findings can be compromised by the bias of the researcher, who is not blind to the hypothesis. Another issue is participant bias, as well as evaluation apprehension that people feel when they are the object of attention. Finally, findings of a case study may not easily be generalised to other cases.
Discourse analysis:
A variant of the case study is discourse analysis, an approach that is popular in Australia and New Zealand and is associated with Margaret Wetherell and other researchers (Wetherell, Taylor & Yates, 2001). Here the focus is on what people actually say in naturally occurring conversation or ‘discourse’ and what is behind the mere words to detect underlying discursive themes. Effective discourse analysis requires a great deal of skill and expertise and is prone to subjectivity – the interpretation of the discourse resting heavily on the perspective and expertise of the researcher. However, when well executed it can be very effective at detecting attitudes and feelings that people are careful to hide – for example racist attitudes in western societies (van Dijk, 1987).
Survey research Method in which a large and representative sample of people answer direct questions about their attitudes or behaviour. Surveys can obtain a lot of data from a large sample of participants that is representative of the general population as a whole. Anonymous and confidential surveys that are well designed can measure people’s true attitudes and feelings.
field study, in which a nonintrusive and ‘invisible’ researcher simply observes, records and codes naturally occurring behaviour. Field studies are excel lent for investigating spontaneous action sequences in a natural context. However, they are prone to observer bias and to distortions by the unintended impact of the researcher on the people being investigated. Field studies also lack objectivity and make for poor generalisations.
What sorts of conclusions can be drawn from these methods? Be sure that you understand the distinction between correlation and causation.
correlations are associations between two concepts but you cannot determine which concept effects one or the other.
causation is when one concept effects the other.
what is a confounding variable?
Where two or more independent variables covary in such a way that it is impossible to know which has caused the effect.
What are the five ethical principles that have received the most attention?
Physical welfare of participants:
Clearly, it is unethical to expose people to physical harm. For example, the use of electric shocks that cause visible burning would be difficult to justify. However, in most cases it is also difficult to establish whether nontrivial harm is involved and, if so, what its magnitude is, and whether debriefing (see below) deals with it. For instance, telling experimental participants that they have done badly on a word association task may have long term effects on self-esteem and could therefore be considered harmful. On the other hand, the effects may be so minor and transitory as to be insignificant.
Respect for privacy Social psychological research often involves invasion of privacy. Participants can be asked intimate questions, can be observed without their knowledge, and can have their moods, perceptions and behaviour manipulated. It is sometimes dif ficult to decide whether the research topic justifies invasion of privacy. At other times, it is more straightforward – for example, intimate questions about sexual practices are essential for research into behaviour that may put people at risk of contracting HIV and developing AIDS. Concern about privacy is usually satisfied by ensuring that data obtained from individuals are entirely confidential: that is, only the researcher knows who said or did what. Personal identification is removed from data (rendering them anonymous), research findings are reported as statistical means for large groups of people, and data no longer useful are usually destroyed.
Use of deception
Laboratory experiments, as we have seen, involve the manipulation of people’s cognition, feelings or behaviour in order to investigate the spontaneous, natural and nonreactive effect of independent variables. Although some experiments have used an amount of deception that seems excessive, in practice the deception used in the overwhelming majority of social psychology experiments is trivial. For example, an experiment may be introduced as a study of group decision making when in fact it is part of a program of research into prejudice and stereotyping. In addition, there has been no evidence of longterm negative consequences of the use of deception in social psychology experiments (Elms, 1982), and experimental participants themselves tend to be impressed, rather than upset or angered, by cleverly executed deceptions, viewing deception as a necessary withholding of information or a necessary ruse
Informed consent
A way to safeguard participants’ rights in experiments is to obtain their informed consent to participate. In principle, people should give their consent freely (preferably in writing) to participate on the basis of full information about what they are consenting to take part in, and they must be entirely free to withdraw without penalty from the research whenever they wish. Researchers cannot lie or withhold information in order to induce people to participate; nor can they make it ‘difficult’ to say ‘no’ or to withdraw (i.e. via social pressure or by exercise of personal or institutionalized power). In practice, however, terms such as ‘full information’ are difficult to define, and, as we have just seen, experiments often require some deception in order that participants remain naïve.
Debriefing is designed to make sure that people leave the laboratory with an increased respect for and understanding of social psychology. More specific ally, debriefing involves a detailed explanation of the experiment and its broader theoretical and applied context. Any deceptions are explained and justified to the satisfaction of all participants, and care is taken to make sure that the effects of manipulations have been undone. However, strong critics of deception (e.g. Baumrind, 1985) believe that no amount of debriefing puts right what they consider to be the fundamental wrong of deception that undermines basic human trust.
