We are almost there, it’s time to finalize our survey before sending it out. Now is the time to think about other outside factors. 

• Select a sample that is representative of your target population. The larger the sample, the more confidence and power you have in claiming the results of your analysis. We recommend at least 100 people, or 10% of the population.
• Take in consideration of the response rate. This depends on:
  • survey length
  • the time the survey is sent out
  • subject pertinence
  • research affiliation
  • presence of incentives
  • means of delivery

Good response rates for specific survey mediums:

• Mail = 60%
• Phone = 80%
• Email = 50%
• Online = 30%
• Classroom (on paper >=50%
• Face-to-face = 80%

Sample Size considerations:

Use a Larger sample size if:

• Time and money are available
• Important to have accurate results
• If you need to split the data into many subgroups, or if there are a lot of covariates
• If people will vary a lot in response

Use a Smaller sample size if:

• Limited time or money
• Results can be a rough estimate, or you plan to estimate for a larger survey later
• You do not need to split data into more than just a few subgroups, or ther are only a few covariates
• If people will not vary much in their answers