What's Your Best Method to Collect the Data You Need?

  • There are multiple data collection methods; the most common are surveys, interviews and focus groups. This Decision Tree will lead you through the steps to determine the best method to use to collect the information you need.


  • Do you need in-depth information about people's experiences or perspectives?

    Posted by:

    Surveys give you general information – breadth, not depth. Interviews and focus groups are better methods to use when the subject matter is complex and you want to fully understand the opinions of your subjects.

    --Individual Interviews are useful when you're concerned about maintaining the participants' anonymity, or want to make sure that they feel free to express unpopular ideas.

    --A Focus Group is a carefully planned discussion designed to obtain perceptions on a defined area of interest in a permissive, nonthreatening environment. It is conducted with 6 - 10 people who are selected because they have certain characteristics in common that are related to the topic. Participants simply state what they believe and are not expected to work toward consensus or rethink their views. Many participants find the interaction stimulating and mention things they would not have thought of individually.

    Comments (-1)
  • Do you need information only from a few people?

    Posted by:

    You need at least 15 respondents to assure anonymity and a minimum of 30 respondents to conduct statistical analyses on your results. Consider interviews or focus groups for small samples.

    --Individual Interviews are useful when you're concerned about maintaining the participants' anonymity, or want to make sure that they feel free to express unpopular ideas.

    --A Focus Group is a carefully planned discussion designed to obtain perceptions on a defined area of interest in a permissive, nonthreatening environment. It is conducted with 6 - 10 people who are selected because they have certain characteristics in common that are related to the topic. Participants simply state what they believe and are not expected to work toward consensus or rethink their views. Many participants find the interaction stimulating and mention things they would not have thought of individually.

    Comments (-1)
  • Do you need information from people who may have limited literacy skills?

    Posted by:

    A telephone survey with translators available is a good method to use. Alternatively, interviews and focus groups with translators or bilingual facilitators should be considered.

    --Individual Interviews are useful when you're concerned about maintaining the participants' anonymity, or want to make sure that they feel free to express unpopular ideas.

    --A Focus Group is a carefully planned discussion designed to obtain perceptions on a defined area of interest in a permissive, nonthreatening environment. It is conducted with 6 - 10 people who are selected because they have certain characteristics in common that are related to the topic. Participants simply state what they believe and are not expected to work toward consensus or rethink their views. Many participants find the interaction stimulating and mention things they would not have thought of individually.

    Comments (-1)
  • Do any of your respondents represent cultural communities who may be unfamiliar with a survey?

    Posted by:

    Do any of your respondents represent cultural communities who may be unfamiliar or uncomfortable with a paper or electronic survey?

    Adequate responses from all portions of the community you intend to survey is necessary if you want to know the perspectives of the entire community. Interviews and focus groups are a better method for participants unfamiliar with surveys.

    Comments (-1)
  • Has a survey already been conducted in SPPS which addresses your topic?

    Posted by:

     If yes, it can save you time or it can inform your data collection (You have the benefit of other people’s experience and you may be able to compare past results with your own). Consider using the existing survey rather than designing your own unique instrument.

    Comments (-1)
  • Do you have the time and resources to design and write a survey?

    Posted by:

    Well-constructed surveys with good reliability and validity usually take from 4 – 6 months from initial planning to the final reporting, including time for item construction, translation, peer review/revision, pilot testing, response time and analysis.

    Consider hiring a program evaluator to help you. Contact Stacey Gray-Akyea for further guidance. Alternatively, consider a smaller scale project such as conducting informal interviews with key individuals.

    Comments (-1)