Introduction To Survey Methods – Topic 1
What is a survey?
- Collection of standardised information from a specific population / sample
- Small amount of info from large number of people.
- Vs : case study research : detailed info collected from relatively few people.
- Definition: in general, surveys are methods of data collection in which information is gathered thru oral / written questioning.
Why conduct a survey?
- Recording info about a sample to make inferences about the population
- Primary data
- Secondary data sources :-
o Documents, eg, academic articles
o Secondary data : survey data collected for other research purpose (eg, ABS)
o Admin records, eg hospital records.
What is a questionaire
- A questionaire is a document containing one / more questions that is administered personally to the respondents by mail or personally by the researcher. The main characteristic is that data are offered by the respondents, with limited interference on the part of research personnnel (Sarantakos, 1998, p 223)
- Wikipedia :-
o Research instrument
o Series of questions
o And other prompts
o For purpose of gathering information
- Verbal / written questioning
- Options
o Structured form
o Structured interview schedule
o Unstructured interview schedule
- Questions re questionaire
o Is questionaire most appropiate method of data collection
o What type of questions should be considered
o How many questions should be included
o In standardised questions, what type of responses should be considerd
o How will the language of the questionaire be chosen
o What are the aspects of the study that must be covered by questionaire
o What is the specific unit of analysis that must be addressed by questionaire
o What elements will be included in the insytructions and cover letter of the questionaire
o How long should questionaire be to do justice to its subject.
o How will ethics and objectivity be observed in questionire.
Planning a survey project
- What is research question
- Who are subjects of research
- What is already known
o Secondary data
o Literature search
- Is the information adequate (if so, stop)
- What resources are required / available
o Time
o Money
o Labour
o Other
- What is best data collection method
o Experiment – establish causes
o Survey – investigate relationships
- Target population – complete group you are interested in
- Meaningful context : lit review / existing data, etc
- If info not adequate – precisely define any voids in info, so new info collected is not just repitition.
Summary of considerations
- Clearly define aims of reserach – clear statement of investigation
- How will proposed research contribute to knowledge
o Replicate previous study in new context
o Different design / new methodology
- Research design
o How will you collect data
o How will you analyse data
- Planning survey
o Work out info required to meet reserach objective
§ Types of info
§ Create questions that are ultimately variables
o How to analyse
o Classify info required into categories. Determine whether you require info that represents
§ Attributes
§ Behaviour
§ Attitudes
§ Beliefs
o Who can provide you with info required
o How will you collect info
Reviewing literature
- Increase awareness of subject
- Determine if there is genuine need for research
- Identify previous / other research
- Identify gaps in knowledge
- Define and focus research topic
- Provide theoretical framework
EBSCOhost
What to look for in the literature
- What is main message
o Presenting or discussing theory
o Primary report on research conducted by author
o Evaluation of other person's work
- What is author's perspective
o Their opinion
o Opinion of other author who they have referenced
o Theoretical perspective?
- How is article relevant to your own area of interest
- What are paper's strengths
o Sample design
o Original methodology
o Analysis methods
- What are paper's weaknesses
o Sample?
o Limited / inapprop analysis methods
o Biased opinions?
- How can you build on this work?
o Approach topic differently
o Followup on any citations
- Checklist for reviewing survey reports
o Citation info
o Theoretical perspectives
o Data source/s
o Methodology
o Analysis methods
o Conclusions
Value of surveys
- Specific hypotheses can be tested and questions asked
- Results available quickly.
- Cost
- Less respondent burdon
Issues with surveys
- Sample surveys are subject to error à sampling error
o Non-response
- Sample may not be representative of target population.
o Ie, homeless people, itinerants
- Expense
o Target pop difficult to locate / not available by phone
o Large sample m/be required for power
o Low response rate
- Labour intensive
Other Issues
- Dating of results
- Validity and reliability
o Measurements are said to be valid if what is measured is correct or believed to be correct
§ Example : in a survey of use of library facilities, assume that we found that 70% of the sample visited the library in the last 12 months. This measure is valid if 70% is really the proportion of students who visited the library
· GW : how do we ever know 'real' proportion.
· GW : is this refering to systemic error
o A measurement is said to be reliable if we consistently get same results on each measurement
· GW : presumably measuring variability.
· GW : is this measuring random error
- Options
o Neither valid nor reliable : we have not measured correct rate of library visits, and each survey produces a difference result.
o Reliable only : consistently obtain result known to be incorrect
o Valid & reliable : consitently obtain correct result.
- Always getting an answer
o Even if questions don't make sense / respondent doesn't understand
- Over-surveying
- Confusion with tele-marketing
- Ethics
o Voluntary
o Informed consent
o Not harmed
o Anonymity
o Confidentiality
o Privacy
What is qualitative research
- Exploratory research
- Informants / participants – vs respondants
- Less structured
- Recraeting reality by darwing on peoples' experiences, thru joint exploration of idaes and attitudes
- Context of discussion important : non verbal, etc
GW Questions / Comments
- If questionaire is not most appropiate data collection method, what are the alternatives?
- Locate reference on market reserach undertaken in Britain during WW2 to assess public morale : Mass Observation
- Re validity & reliability – refer Measurement book
- Look at SMS pup questionaire
- Look at Murdoch University questionaires : http://scored.murdoch.edu.au/index.html
No comments:
Post a Comment