Please enable JavaScript.
Coggle requires JavaScript to display documents.
Research Methods Chpt 9 (Probability samples (Cluster samples (may include…
Research Methods Chpt 9
How a Questionnaire is Designed
Determine Types of Qns
Open-ended qns
Disadvantages
Harder for less articulate respondents to answer
Harder to code data-->data analysis messy job
Advantages
More useful for smaller & preliminary studies--> can get a better sense of the range of likely answers before giving close-ended format
Close-ended qns
Advantages
Easy for less articulate respondents to answer
Easier to code and analyze, fewer off-the-wall responses
Disadvantages
respondent may not agree with the options
tend to put words in mouths of respondents-->suggest alternatives that respondents might never come up with themselves
Issue of study may be too complex to be reduced to few options
Respondent may misinterpret qn-->no way to discover the fact
Clerical error when coding the data--> no way to discover the fact
Write the Items
Options must be
Mutually Exclusive
&
Exhaustive
Beware of
Social Desirability
tendency
Word qns such that each alternative appears equally socially desirable
Include a
verification key
eg. lie scale
Avoid Bias (the way you phrase the question)
Beware of
Acquiescence
Address a Single Issue per Item (cannot be double-barreled)
Determine the Format of the Item eg.
Visual Analogue Scale
,
Likert Scale
& include
Branching Items
Sequence the Items
Determine how Data will be Analysed
Determine Purpose
Determine how Data will be Analysed
Administering the Questionnaire
Computerised Administration
Adv
computer is absolutely consistent
computer can check for invalid responses, prompt interviewer to check implausible answers
impersonal, social desirability reduced
available 24/7, high level of participation possible at low cost
Disadv
due to anonymity, ppl may lie abt age or gender
diff to obtain truly random sample since respondents are self-selected
diff to obtain true measures of response rates
ppl who don't own a computer won't be aware of the survey,can't participate
illiterate participants may be embarrassed to admit problem & answer qns they don't understand
security of responding is vital
storage of survey info must be ensured
Telephone Administration
adv
able to do a computer-assisted interview
can be conducted from a central location where interviewers can be supervised
low cost, fast, no need to travel
Disadv
person with cell phone can be located anywhere-->sampling appropriately from within a specific location extremely diff
cell phone users pay by minutes, less inclined to waste minutes on a questionnaire-->harder to calculate response rate
less anonymous-->introduce interviewer bias
only ppl with telephone and willing to do survey can be sampled
Harder to establish rapport, cannot use visual aids
Written Responses
Types
Drop-off administration
Mail administration
Group administration
adv
low cost
other than group adminstration, respondents can complete at leisure, greater anonymity, reducing interview bias
disadv
Low response rate
unable to clarify qns that may be misunderstood
illiterate participants may be embarrassed to admit problem & answer qns they don't understand
impossible to determine if respondents took survey seriously
Face-to-Face
Disadv
much more expensive compared to other methods
safety of interviewers cos interviews usually at night
presence of interviewer creates social sit that may result in biased response
interviewer may fake data to save travelling & interview time
adv
able to clarify misunderstandings
interviewers can guarantee the order in which qns are administered
interviewers can establish rapport
Visual aids can be presented to clarify a survey qn that may not be possible in other methods of administration
Sampling
Types
Purposive samples
Convenience samples
Haphazard samples
Probability samples
Probability samples
Systematic samples
Simple random samples
sampling frame
Stratified Random sample
Cluster samples
may include
multistage sampling
pg 236 for steps for sampling