What kind of data are produced by surveys and what are the strengths and weaknesses of such data? What research design issues are implicated in the use of survey methods? What are their advantages and disadvantages? What are the options for methods of administering surveys? Definition
A systematic collection and analysis of
data relating to the attitudes, living conditions, opinions, etc. of a population, usually taken from a representative sample of the latter (Oxford Dictionary) Survey = Questionnaire? Structured observation Diaries Content analysis Often are questionnaire-based Less structured interviews? Defining through nature of data? Units of analysis: commonly individuals, but may be households, countries etc. Compare units of analysis, or sub-groups of them (e.g. gender, class) Quantitative data (how many?): usually fixed responses coded numerically But, like other data, survey data are created rather than simply collected Strengths of survey data Suited to large-scale research Generalizable to populations Relatively easy to prepare for analysis Suited to complex statistical investigation Easily replicable Transparent Convincing to non-experts Weaknesses of survey data
Better at providing breadth than depth
Better at answering What? than Why? Does not explore meaning or significance of response categories Decontextualizes: responses might vary under different conditions Less flexible if problems or new questions emerge: hence importance of pre-testing Surveys in research design May be preceded by a more qualitative research phase: raise significant issues and provide foundation for questions May be succeeded by a more qualitative research phase: follow up interesting findings and further questions in more depth Both approaches may help mitigate weaknesses of survey data Different survey designs Cross-sectional: one-off snapshot; do not assess change over time Longitudinal (or panel): examine same respondents at two or more time points; aim to assess change; may have experimental aspect; disadvantages related to resources and attrition Repeated cross-section: same question(s), different time points, but different respondents; amenable to retrospective or ad-hoc assessment of change but less confidence that change has really occurred Operationalization to answer research questions Concept(s) Indicator(s) Question(s) Fieldwork Analysis Findings (Defining) Concept Indicator(s): what needs to be measured? Questions: Existing? New? How many? Piloting and Pre-testing
Pre-administer survey to small number of
respondents from same sample/population Utilize expertise (specific or general) Especially important for new questions and/or new contexts How many and who? Piloting: objectives? Highlight redundancies or fundamental deficiencies Are questions/wording relevant to all? Assess comprehension Any unease or offence? Distribution of response? An interactive and reflective process Cognitive interviewing/testing
Potential to highlight differences in
understanding In context of survey research, is used to study the manner in which targeted audiences understand, mentally process, and respond to the materials [e.g. survey questionnaires] we present (G. Willis, 2004: 3)