Data3Sixty Govern (“Govern”) has a feature to create surveys. Surveys are helpful to obtain feedback from Govern users about a specific topic. The feedback is gathered and presented via dashboards from Data3Sixty DQ+. Below are 3 steps of analysis to consider before creating an effective survey in Govern:
1. Understand high level requirements of a survey
- What is the goal of the survey?
- Who is the audience for the survey?
- Where should the survey be located in Govern?
- How long can users take the survey?
- How often can users take the same survey?
2. Keep survey questions to a minimum of between 3 – 5 questions
3. Use clear concise questions so responses can be easily evaluated
- Ensure the question and response options are framed unbiasedly. For example,
- A leading/biased question: Do you find our magnificent data governance program helpful?
- A neutral/unbiased question: Do you find the data governance program helpful?
- Ensure your question is straightforward. For example, don’t use double negatives.
- Use terminology and words within your survey that all users will understand.
- Be careful not to assume that your audience is data governance experts.
- Try not to use acronyms.
- Provide concrete and finite answer options that pertain to the goal of the survey.
- Multi-choice options for a response are preferred over free text comments.
- Review Yes/No questions to see if providing other options would give you more insight into the user’s response.
- Do not include multiple topics in 1 survey question. For example,
- Unhelpful question: Do you understand why the company is investing in data governance both at the enterprise and the business unit level?
- Instead, break this out into 2 questions
- Do you understand why the company is investing in enterprise data governance
- Do you understand why the department is investing in data governance?
- Create Incentives! Remember the users are taking your survey to help you. Therefore, consider offering a reward for completing the survey within a designated timeframe.