Planning for a
survey
SIMPLE SURVEY
CONSIDERATIONS
What data do you need?
Qualitative?
Specific comments / Statements
Feelings / Opinions
Quantitative?
Observable trends?
Numbers?
Sales figures / participation rates / etc.
PURPOSE FOR THE SURVEY
What is the total amount of
people you have access to?
How many people are you
prepared to administer your
survey to?
Do you want your survey to be
representative?
SAMPLE SIZE
Systemic sampling
Setting a prefix indicator
Time / Specific number
Every 3rd/5th/etc person, we will survey
Every time the indicator comes up, you
conduct the survey
Random sampling
Random selection of any potential respondent
that crosses your survey area.
SAMPLING
Keep questions short and simple.
Each question should only
address 1 area.
Do not direct the respondent’s
answer.
Do not introduce your own
biases.
KEY CONSIDERATIONS
Personal Details
Yes / No
Range indicator
Data entry / open-ended
option
QUESTION TYPES
Necessary to spot potential trends.
Demographic trends can affect survey results.
Not a bias you built in but a resultant one
nonetheless.
Track the following
Age
Sex
Education level
Income group
PERSONAL DATA
Structured data
Simple contrasts
Provide choices that you want to
track
Inaccuracy due to option that is not
present
Choose the next best option.
CLOSED ENDED QUESTIONS
Allows for measuring degree
How much / How little
Useful for comparing differences in your
respondent groups.
People above 15 prefer … People 15 and
below prefer
Use an odd number scale.
5 point scale would be easier to analyse.
RANGE / SCALE
Providing a final option for
respondents to insert their own
answer.
Enhances accuracy
Potentially useful only as a sidenote.
Numerically too little to be of
significant value.
OPEN-ENDED OPTION
Sensitive question
What is your monthly income?
Does the weather affect your work performance?
Irrelevant question
How often do you skip lunch? (Survey is on food stalls
in school canteen)
Biased question
Why is Coke a better beverage than Pepsi?
Double-barrelled question
Do you agree that the lack of amenities is a problem
with this theme park and that the number of visitors
should be limited?
SOME BAD QUESTIONS
Any questions?
HOPE THIS HELPS!

Dept ppt sharing simple survey considerations

  • 1.
    Planning for a survey SIMPLESURVEY CONSIDERATIONS
  • 2.
    What data doyou need? Qualitative? Specific comments / Statements Feelings / Opinions Quantitative? Observable trends? Numbers? Sales figures / participation rates / etc. PURPOSE FOR THE SURVEY
  • 3.
    What is thetotal amount of people you have access to? How many people are you prepared to administer your survey to? Do you want your survey to be representative? SAMPLE SIZE
  • 4.
    Systemic sampling Setting aprefix indicator Time / Specific number Every 3rd/5th/etc person, we will survey Every time the indicator comes up, you conduct the survey Random sampling Random selection of any potential respondent that crosses your survey area. SAMPLING
  • 5.
    Keep questions shortand simple. Each question should only address 1 area. Do not direct the respondent’s answer. Do not introduce your own biases. KEY CONSIDERATIONS
  • 6.
    Personal Details Yes /No Range indicator Data entry / open-ended option QUESTION TYPES
  • 7.
    Necessary to spotpotential trends. Demographic trends can affect survey results. Not a bias you built in but a resultant one nonetheless. Track the following Age Sex Education level Income group PERSONAL DATA
  • 8.
    Structured data Simple contrasts Providechoices that you want to track Inaccuracy due to option that is not present Choose the next best option. CLOSED ENDED QUESTIONS
  • 9.
    Allows for measuringdegree How much / How little Useful for comparing differences in your respondent groups. People above 15 prefer … People 15 and below prefer Use an odd number scale. 5 point scale would be easier to analyse. RANGE / SCALE
  • 10.
    Providing a finaloption for respondents to insert their own answer. Enhances accuracy Potentially useful only as a sidenote. Numerically too little to be of significant value. OPEN-ENDED OPTION
  • 11.
    Sensitive question What isyour monthly income? Does the weather affect your work performance? Irrelevant question How often do you skip lunch? (Survey is on food stalls in school canteen) Biased question Why is Coke a better beverage than Pepsi? Double-barrelled question Do you agree that the lack of amenities is a problem with this theme park and that the number of visitors should be limited? SOME BAD QUESTIONS
  • 12.