3.3.2: Sources of bias and assessing the value of different methods

bias linked to research design and conduct

• one or more RM chosen - may be subject to bias

• non-random sampling technique = sampling bias

  • due to own judgement

  • not representative

  • difficult to generalise

• sampling bias linked to non-response

  • systematic bias

• sampling bias - inadequate sampling frame

• researcher’s presence - interview bias, experimenter bias, observer effect

• social desirability effect - invalidates findings

• researcher asking leading questions - impacts participant answers

• coding errors for questionnaires

researchers values

• study something they consider important

• influences by values, beliefs, morals

• may take sides when analysing interview data

funding

• funded by public bodies

  • if they identify a skill shortage they may try to address it by funding more studies based on a certain method

  • have research priorities - competitive funding

reflexivity

• redfield and lewis

• hollis

• reflecting on yourself and how you as a person might have affected findings and conclusions

assessing the value of different research methods

concepts of validity, reliability, objectivity, representativeness and ethics

validity

• concerns whether the method used to gather data actually measures what it is intended to measure

• true and accurate data

• reduced by social desirability effect

reliability

• concerns whether the method produces the same or consistent results the second time round

• qualitative - cannot be easily replicated

objectivity

• whether the method produces impartial and unbiased data

• whether the researcher’s values intrude

representativeness

• concerns whether the sample is typical of the population
  that it is drawn from

• generalisations made from a representative sample

ethics

• concern whether the researcher conducted the research

• used the methods in ways that comply with ethical guidelines