Online Poll and Survey Tools

Index

  1. Review Your Project
  2. Accessibility Considerations
  3. Relevant Vocabulary
  4. Select a Tool


  1. Review Your Project

    1. What is the scope of your project?
      Will you require a simple poll, a simple survey, or a complex survey? This will determine how powerful and customizable your tool needs to be.
    2. What are your goals?
      What types of reports do you plan to generate? What questions do you need to ask to get the information that you want? This will determine which question types you should look for and whether you need advanced features like response cross-tabulation.
    3. Who are your respondents?
      Will they be from inside the department, the campus, or from the outside community? How many respondents will be involved? This will determine the size and accessibility of your survey.
    4. Will you need to request private data?
      Will your survey require student ID or social security numbers? Will the respondents be providing you with personal information? This will determine the level of security required.
    5. Will you need to ask different questions to different groups of respondents?
      Are there any specific questions targeted to a subset of your audience: like faculty, industry or students? This will determine whether you need advanced features like skip logic and piping.
    6. Who will be administering it?
      Will there just be one person or many people administering your survey? How technically sophisticated are your survey administrators? How long will you need to keep the results? How much do you have to spend? This will help determine the impact on your unit and the level of administrative control required by the tool.
    7. Additional considerations:
      • If you only need to ask one question, then you should conduct a poll rather than a survey. Your needs will be much simpler. See the poll tools chart below.
      • If you are creating a complex survey and have some money to spend it might be worth investing in one of the priced versions.

    Design Tips



  2. Accessibility Considerations

    Disabled users employ a variety of technologies to access content on the web. Each technology offers different accessibility challenges. The 3 basic user types are:

    Form-Specific Suggestions

    When designing your survey, there are a few things which should be avoided if possible:

    General Good Practices for Accesssibility



  3. Relevant Vocabulary

    # types of questions
    There are 4 main groups of question types:
    1. open ended : Allows respondents to reply in their own words (e.g. text boxes)
    2. close ended : Limits respondents' answers to only those provided by the survey. Includes true/false and multiple choice questions (e.g. radio buttons and check boxes)
    3. scale : Respondents are offered an array of choices. These include rating and ranking scales (e.g. rank your experience on a scale from 1 to 5), and complex ratio scales.
    4. unfolding (also split, filter, or contingency) : Begins by asking respondents a general question then follows up with clarifying questions based on their answers. (e.g. Did you attend the event? _yes _no
      [if yes] Which lectures did you attend?       ). If you wish to do this, it is recommended that you employ skip logic to avoid unnecessary confusion.
    # answer options
    The number of times a respondent may reply when answering a single question. (e.g. Select all of the times you will be available to attend the meeting). This is usually only an issue when selecting poll software.
    508 compliant
    Said to meet most accessibility standards.
    response validation
    Verifies that a respondent has filled out the survey correctly before submitting the reply to the server.
    cross-tabulation
    Allows you to see the data relationships between different sets of questions or do complex statistical analyses.
    skip/branch logic
    Allows respondents to skip whole sections of questions that don't apply to them. Instead of having to create a phrase like "Those who answer ‘no’ to question 10 should skip to question 15." you can design the survey so that, when someone selects ‘no,’ all of the intervening questions before question 15 disappear.
    piping
    Allows you to pull answers from one part of a survey and plug them into another part. For example, in response to question 5 "Where do you live?", the respondent writes ‘New York.’ Question 6 then becomes "What do you like best about New York?"
    randomization
    Allows you to mix up the questions randomly to prevent bias (mainly used for marketing).
    filtering/segmentation
    Allows you to filter the results for patterns (mainly used for marketing)

  4. Select a Tool



maggie@eecs
Main study conducted 04-may-2009
Notes about accessibility added 02-mar-2010

EECS Administrative Computing Group | acg@eecs.berkeley.edu
University of California, Berkeley