Experiments to explore this topic may be laboratory or field based. Although carried out by social scientists they have their roots in the natural and physical sciences and involve the formal testing of hypotheses about causation. Quasi experiments are those in which the experimenter has no control over the 'independent variable' (e.g. a natural disaster), but can choose how, when and for whom 'dependent variables' (e.g. features of behaviour indicating stress) are measured.
Potential causal relationships can be tested. True experiments: The independent variables can be controlled. 'Subjects' can be randomly assigned to various experimental 'conditions'. Lab experiments: Features of phenomena can be isolated. Relatively convenient to organize. Field experiments: Natural setting gives greater 'external validity'. Less experimenter influence where subjects unaware of experiment.Quasi experiments: Enable study of unmanipulable variables. Some phenomena can only be studied this way. Considerable external validity.
Disadvantages
True experiments: Ethical issues of deception.
Lab experiments: Artificial setting may threaten their 'external
validity'. Tendency to use convenient subjects (e.g. students) who may not
be representative. Possible influence of experimenters.
Field experiments: Variables more difficult to isolate. Little
control over events. Less convenient to organize.
Quasi experiments:
Independent variable uncontrollable. Additional uncontrolled variables may
affect dependent variables studied. Random assignment of subjects to
conditions more difficult. Unpredictable phenomena limit preparation time.
Interpretation has to be more cautious.
Investigator more reactive than purposely manipulative. Main role that of an observer (typically 'participant observer'). Needs to be 'reflexive': to reveal and reflect on own biases.
Not so artificial. In depth study over time. Large amount of rich data. May
generate useful insights.
Disadvantages
Usually limited group of people involved. Systematic observation and typification difficult. Cannot generalise beyond group observed. Very time consuming and labour intensive. Lack of control, so difficult to use this strategy for exploring causation. Possible influence of investigator.
Direct questioning of people about topic. Requires systematic sampling
procedures to enable generalization.
Can be very specific. Questionnaires Permit large scale studies and greater generalizability. Easy, economical, more anonymous. Interviews can be sensitive to understanding of question and to nuances of response.
Disadvantages
Self report can be inaccurate (poor recall or bias) by comparison with observation. Good question design demanding. Pre testing necessary. Questionnaires Less investigator control. Questions may be misunderstood. Interviews Require careful handling, expensive and susceptible to questioner bias.
More open ended, less directive interview style. Often used as part of ethnographic field studies but not necessarily.
Detailed information with subtle nuances. Open to interviewee's relevant but unanticipated concerns which may lead on to further lines of enquiry. Interviewee can pursue more specific answers.
Disadvantages
Some potential respondents are not comfortable with probing interviews. Requires very experienced and sensitive interviewer. Difficult to sustain at length. Transcription from tape to written record very time consuming. Coding problematic. Possibility of data overload. Unreliability of self report.
Involves typifying and measuring recurrent features in television programmes (e.g. stereotypical representations of gender roles or ethnic groups).
Useful in revealing common stereotypes on TV and prevailing values of programme makers. Useful baseline data on formal (presentational) features such as incidence of editing techniques. Can be relatively inexpensive. Unobtrusive. Quantification can facilitate replication.
Disadvantages
Can tend to emphasize an objectively definable 'content' rather than the importance of the process of negotiating meanings. Makes no allowance for viewing choices or individual styles of interpretation. Danger of assuming and inflating power of content to shape attitudes and behaviour of viewers. Typification/definition of units measured, and selection of sample material is problematic.
Daniel Chandler
Experiments
Advantages
Ethnographic Field Studies
Advantages
Questionnaire Surveys and Structured Interviews
Advantages
'Unstructured' (Case Study) Interviews
Advantages
Content Analysis
Advantages
Further reading
November 1994