Wk9- Surveys & Questionnaires Flashcards
Questionnaire Surveys: Quote + 2 Crucial Errors
“I know, I’ll do a questionnaire” (Bridge, 2003: 230)
1) Planning before fieldwork begins
- Constructing a meaningful questionnaire
2) Amount of work they entail
- Obtaining sample: Public tolerance for surveys low?
Types of Questionnaire Data (3)
1) CLASSIFICATION/PROFILE
- classifies people, their circumstances and env
E.g. age, income, housing tenure
2) BEHAVIOURAL
- Relates to the behaviour of people
e. g. where do they shop? What sort of work do they do?
3) PERCEPTION
- Relates to people’s attitudes, opinions and beliefs
- Typically the most difficult to collect
- Potential for patterned response, insincerity and response errors
Administering a Questionnaire
- TYPE OF QUESTIONNAIRE
1) Researcher- led
(e. g. f2f, phone)
2) Self- Completion
(e. g. 3rd person distrib and collection, postal, online)
3) Online Options: Email/social media/internet questionnaire website e.g. SurveyMonkey
- TIMING AND LOCATION
Importance of a Pilot Study (5)
1) Have you asked the right qus?
2) Are qus in the right order?
3) Have all qus been understood?
4) Logistical issues
- Allows skipped qus to be identified
- Gives indication of timescale for completion
- Allows typo corrections
- Provides researcher experience and confidence
5) Always discuss questionnaire with participants- then refine qus
What to do with Qu Data?
- Managing data with Excel, SPSS, or other software
- Qualitative vs Quantitative Responses
- Cross tabulations
Ensuring sufficient collection of contextual data (age, gender, income/class, hometown) - Presentational Issues
Questionnaires
- Strengths (3)
- Weaknesses (2)
Strengths:
- Affords an overview of a pop’s characteristics
- Permits analytical exploration of key relationships
- May be effective in combination with other methods
Limitations:
- May lack explanatory power/qualitative depth and generally does not enable further probing of responses (unlike interviews)
- Challenging to work out variable data
Two Crucial Errors of Qu Surveys
Bridge (2003)
1) Only defined an area of research rather than a SPECIFIC RESEARCH QU
2) Assumption that qu is the BEST AVAILABLE AVAILABLE
(because it has come to be associated with the so- called ‘hard social science’ involving large surveys and statistical analysis
- Considered the ‘cure all’ for the problem of doing fieldwork
- Only effective if it is the most appropriate method of providing the info needed to address a well- defined research problem
Success/Failure of a Questionnaire depends on…
Bridge (2003)
1) Sampling Theory
2) Qu. Design (Wording of the qu, layout of the qu)
3) Analysis and Interpretation of the results
Survey Definition
Fink and Kosecoff, 1985
‘A survey is a method of collecting info directly from people about their feelings, motivations, plans, beliefs and personal, educational and financial background’
Aim of Sampling Theory
- Avoid bias and ensure that sample is REPRESENTATIVE of the total pop of interest as possible
- Determines the most appropriate way of selecting households or individuals from your sampling frame
Sampling Size
- Sampling SIZE also important: must allow for non- responses, especially in postal questionnaires
- Minimum of 30 and max often determined on PRACTICALITY (cost of stamps, postage etc)
‘Bigger, the better)
Analysis and Interpretation
ANALYSIS is critical- poor analysis has the potential to ruin a wonderfully designed questionnaire with high response rates
- there are statistical tests that conduct anything you can think of but don’t necessary need to use most sophisticated software: problem is INTERPRETATION i.e. knowing WHAT you are looking for rather than knowing about the stats
What determines the form of questionnaire?
1) NATURE of the study
2) PRACTICAL LIMITATIONS
e. g. if potential respondents are scattered all over the country = postal qu best
Postal Questionnaire:
Pros (4) and Cons (6)
PROS:
- Reduces travelling time and legwork
- No interview bias
- Useful for geographically- dispersed population sample set where it just wouldn’t be practical to interview all face to face
- Time to reflect and gives respondents time to answer difficult qus
- Avoid f2f contact if a shy individual…
- Good for embarrassing and personal qus
CONS:
- Qus must be easily understandable and unambiguous as answers cannot be rechecked with the respondent
- BIAS: Have an insight into the line of questioning = tailor responses to fit your reasoning
- Don’t know who is answering or if instructions are followed
- To get the min 30 sample size- with a response rate of about 30% need to send out over 100!
- Expensive- postage, stamps and paper as must explain academic position/research
- Supplementary rich observation data is not available- could provide useful contextual data for survey and help further probing via interview
Well- Crafted Questionnaire is the product of… (4)
1) Clearly defined research objective
2) Sensitivity to the potential respondents
3) Trial and error (using a pilot survey or if not possible, passing the qu around friends, family and colleagues to ensure they understand the qus in the same way you do)
4) COMMON SENSE
- most valuable commodity!
