Topic 1 - Quantitative Methods

• Positivism


  Positivists argue that society is an objective social reality that shapes our behaviour. They seek to discover laws of cause and effect that explain these patterns. They believe that sociological research should follow the model of natural sciences to produce objective knowledge about society (facts about society).
• Quantitative methods


  Research methods that produce quantitative data, such as experiments, structured interviews, questionnaires, official statistics.
• Lab experiments


  Divide a set of identical research subjects into two groups: Experimental Group (exposed to the independent variable to test its effect) and Control Group (not exposed to the independent variable)
• Lab experiments have many practical, ethical and theoretical limitations
• Practical issues of lab experiments


  • Individuals are complex and therefore it is difficult to 'match' members of control and experimental groups exactly
  • Expectancy Effect: Investigator bias as to what they expect the outcome of the study to be
  • More likely to be a small sample making it difficult to investigate large-scale social phenomena
  • Society is an open system where a variety of factors interact with each other in complex ways
  • Hawthorne Effect as the environment is artificial
  • Unable to study the past
• Practical issues of field experiments: Whilst this method does overcome the issues surrounding Lab experiments as they are more natural, there is an issue with the lower levels of control a researcher will have. This means we cannot be certain if we've identified the true cause.
• Practical issues of questionnaires


  • Quick and cheap way to gather large data sets and data is easy to quantify, especially where pre-coded questions are used and can be computer-processed
  • Data is often limited and superficial
  • Very low response rate
  • Researchers must already have some knowledge of the subject, therefore it makes it less suitable for investigating unfamiliar topics
  • No need to recruit and train interviewers
  • May be necessary to offer incentives
  • With postal and emailed versions, we can never be sure if it was ever received by the respondent
• Practical issues of structured interviews


  • Suitable for gathering straightforward factual information and results can be easily quantified because of the use of closed-ended questions with pre-coded answers
  • Training for the researchers is relatively straightforward; however, this does make it more costly than other methods
  • Inflexible, because the schedule is drawn up in advance and the researcher must stick to it rigidly
  • Can cover large numbers of people because they are quick and fairly cheap to do
  • Only a snapshot taken at one moment in time, so they fail to capture the dynamic nature of social life
• Ethical issues


  • Confidentiality
  • Psychological Harm
  • Informed Consent
  • Deception
• Reliability


  Repeatable in every detail by other researchers to obtain the same results
• Representativeness


  Whether our sample is a true cross-section of our target population
• Validity


  Whether our research is a true picture of what something is really like. Whether our research is measuring what it sets out to measure.
• Positivists value reliability, representativeness and validity in their research as it enables them to draw wider conclusions that findings are true, make large-scale generalisations about how wider social structure shapes individuals behaviour, and discover true meanings that underlie actions and create a social reality.
• Reliability, representativeness and validity of lab experiments


  • Lab experiments are reliable as the original experimenter can control conditions entirely, produces quantitative data (easy to compare), very detached and objective
  • Lab experiments are not representative as they can only study small samples
  • Lab experiments are not valid as there is a lack of generalisability from the artificial setting thus they lack external validity, and they may also lack internal validity (measuring what they are meant to be) due to the Hawthorne Effect
• Reliability, representativeness and validity of questionnaires


  • Questionnaires are reliable as they are wide scale and standardised, yielding quantitative data that allows correlations to be established between variables
  • Questionnaires are representative as they are large scale and standardised, using sophisticated sampling methods
  • Questionnaires are not valid as they are the most detached of the primary methods, posing risks of lying, forgetting, social desirability, and imposed meanings
• Reliability, representativeness and validity of structured interviews


  • Structured interviews are reliable as they are standardised measuring instruments with pre-coded questions
  • Structured interviews are representative as they are relatively quick and easy to undertake so large numbers can be sampled, although those with time to complete the interview may be unrepresentative
  • Structured interviews are not valid as they use closed-ended questions which force participants to choose limited answers, giving interviewers little freedom to clarify or explore, and there is still the issue of imposed meanings and lying
• Reliability, representativeness and validity of official statistics


  • Official statistics are reliable as they are compiled by trained staff using standardised techniques, although untrained individuals could make errors or omit information
  • Official statistics are representative as they are very large scale, often covering the whole population, with great care taken with the sampling method
  • Official statistics are not valid as Interpretivists would state that they merely represent labels that society constructs to measure the labels that officials give to people
• Feminism


  Argues that the survey method treats all individuals as being equal units and therefore does not reflect the patriarchal society in which the data are gathered. Shalumit Reinharz (1983) goes as far as saying the interview process is 'research as rape'.
• Interpretivism and statistics


  Statistics purely represent what society constructs to measure the labels that officials give to people. Statistics purely symbolise 'decision gates' taken by agencies e.g. doctor diagnosis or arrest.
• Marxism and statistics


  Unlike Interpretivists, Marxists do not see official statistics as outcomes of labels. Instead they serve the interests of capitalism. The function of official statistics is to conceal or distort reality and maintain the power of the capitalist class. They are part of the 'ideological state apparatus' – a set of institutions that produce ruling-class ideology.