Questionnaires Flashcards
Practical ++++
✓ Quick & cheap way of collecting large quantities of data from large numbers of people who can be widely spread geographically (especially if a postal questionnaire is used). There is no need to train interviewers or observers – respondents simply complete & return the questionnaire themselves.
✓ Easy to quantify – pre-coded, closed-ended questions are easy to count. Can be processed quickly by computer to show relationships between different variables.
✓ Hypothesis testing to establish cause & effect: Relationships can be tested between 2 different variables. E.g. a questionnaire may find a link between material deprivation & criminal activity. This allows us to make statements about the possible causes of criminal behaviour & predictions about who is more likely to commit crime.
✓ Before & after comparisons: Questionnaires allow for comparisons to be made. E.g. you may want to compare peoples’ attitudes to homosexual behaviour in the past & present, & generalisations can be made by comparing the before & after.
Practical —–
× Limited information/insight gained: Questionnaires need to be brief since most respondents are unlikely to complete a long, time-consuming questionnaire. This limits the amount of information & insight that can be gained.
× Researcher cannot be sure: (i) Whether the potential respondent has actually received the questionnaire. (ii) Whether a returned questionnaire was actually completed by the person to whom it was addressed.
× Inflexibility: Once the questionnaire has been finalised, the researcher is stuck with the questions they have decided to ask & cannot explore any new areas of interest should they come up during the course of the research.
ethics
✓ Pose fewer ethical problems than other research methods. Questionnaires may ask intrusive or sensitive questions, but respondents are generally under no obligation to answer them. Even so, researchers should gain informed consent, guarantee anonymity & confidentiality, & make it clear that the respondent has a right not to answer any of the questions that they do not wish to.
Reliability
✓ High in reliability: If repeated by another researcher, a questionnaire should give similar results to those gained by the first researcher. This is due to having closed, set, standardised questions & often pre-set answers which either reduces or eliminates (as with postal questionnaires) the possible influence of the researcher to answer in a certain way.
✓ Allows comparisons over time & between societies or areas & by asking the same questions we can compare results.
validity +++
✓ Detachment & objectivity: Questionnaires are said by positivists to be unbiased because they are a detached & objective form of research, where the sociologist’s personal involvement with their respondents is kept to a minimum. They argue that this allows the method to gain more truthful data about the respondent’s opinions, etc.
validity —
× Detachment: Interpretivists argue that because questionnaires involve little or no contact between researcher & respondent, questions & answers cannot be clarified or misunderstandings cleared up. There is also no way of knowing whether the respondent & researcher both interpret the questions & answers in the same way (a particular problem if there are cultural or language differences between researcher & respondent) which can reduce the validity of the results gained.
× Lying, forgetting & ‘right answers’: Respondents may lie, forgot, not know, not understand, or try to please or annoy the researcher. Some may give socially desirable answers they feel they ought to give, rather than tell the truth. When respondents give answers that are not full or honest, this undermines the validity of the data.
• J.W.B Douglas ‘The Home and the School’ (1964) Based on questionnaires given to over 5000 parents, Douglas concluded that the degree of parents’ interest in their children’s education was the single most important factor affecting attainment.
Representativeness
✓ Increased chance of a representative sample: Collect information from a large number of people, widely spread geographically (especially if a postal questionnaire is used). This means that they are more likely to represent the feelings/opinions, etc. of the wider population & results can then be generalised.
× Low response rates: Very low response rates are a major problem, especially with postal questionnaires (25% would be an excellent return rate). Few of those who receive a questionnaire bother to complete & return it which will reduce the likelihood of a representative sample. Complex language or bad design reduces response rates. A higher response rate can be obtained if follow-up questionnaires are sent & if questionnaires are collected by hand but this obviously adds to cost & time.
× Those that return the questionnaire may be different from those who don’t: Busy people in full-time employment or with young children may fail to respond, whereas the unemployed, retired or socially isolated people with time on their hands may be more likely to. Those with strong views on a subject are also more likely to respond compared to people who have little knowledge or interest in it. If the respondents are different from non-respondents, this will produce distorted & unrepresentative results from which no accurate generalisations can be made.
Theory
✓ Favoured by positivists: Questionnaires achieve the main positivist goals of reliability, generalisability & representativeness. Standardised (set) questions & answers produce reliable data because other researchers can replicate the questionnaire. Pre-coded responses allow us to produce quantitative data, identify & measure behaviour patterns, & establish cause-&-effect relationships. As they are often large scale, they are often more representative of the wider research population.
× Rejected by interpretivists: As they impose the researcher’s ideas & meanings on respondents through the questions they chose to ask. The researcher chooses which questions to ask & decides the response categories, while closed-ended questions limit the answers respondents can give. Interpretivists say that this tells us little about the meanings held by social actors (i.e. how people truly feel). Questionnaires fail to achieve the main interpretivist goal of validity.
