Skip to main content

Research Plays


facebook icon   twitter icon   print icon

The Paperwork Reduction Act (PRA) of 1995 requires that U.S. federal government agencies get Office of Management and Budget (OMB) approval before requesting or collecting most types of information from the public. The goal of the PRA is to keep federal agencies from overburdening the public with federally sponsored data collections.

How to Get PRA Approval to Conduct Research


1

  Determine if You Need to Get OMB Approval Before Collecting Data
According to the PRA, agencies need to get OMB approval prior to collecting federally sponsored data in any situation where:
  • There are 10 or more respondents within a 12-month period.
  • The questions are standardized—whether they’re asked in-person, on the phone, or online.

2

  Identify Your “Respondents”

A respondent includes individuals; partnerships; associations; corporations; business trusts; legal representatives; organized groups of individuals; and State, territory, tribal or local governments.

One exception: if you’re soliciting feedback from Federal employees, you do not need PRA clearance (though you may need approval from the relevant unions).

3

  Identify Your Clearance Process Requirements and Options

There are two processes for obtaining PRA clearance:

  • Traditional Clearance (requires a 60-day Federal Register Notice)

  • Fast-Track Clearance (does not require a 60-day Federal Register Notice and includes customer experience activities such as comment cards, focus groups, usability studies, and some surveys. However, you can’t publish or publicly report your findings. Click the link to see the full Fast-Track Clearance requirements.)

The process for obtaining a PRA Clearance varies depending on the subject matter you’re collecting data about. The processes can differ from agency to agency, both on level of detail and platform for submission.

4

  Consult Your Agency’s PRA Clearance Officer

Your agency’s PRA clearance officer will have information and specific guidance on your submission requirements and process. Contact your PRA clearance officer as soon as you think you might be doing a survey or other user research, so you’ll have enough time to get the clearances you need for your project. If you don’t know your PRA officer, please contact USDA’s PRA Officer, Ruth Brown (ruth.brown@usda.gov).

 

Toolkit Resources

Additional Resources

User research lets you learn more about your target audience’s goals and needs. Surveys measure user satisfaction with your site or app and help you figure out what works well and what falls flat.

Rule of Thumb: The best surveys feel like conversations with people you’d like to get to know.

When you create a good feedback survey, you’re getting concrete information on which elements of design, content, and functionality you need to help and support your users.
 

How to Create a Feedback Survey


1

  Target the Appropriate Audience
  • If you have a highly targeted website or application, chances are you’ve already established relationships with your users. Perhaps they’re federal employees, stakeholder organizations, or members of practitioner communities. If so, those groups will be your best contacts for surveys (and any other user research).

  • If your website or application is for the general public, you may need to use other methods to find survey participants, such as promoting surveys in email newsletters or on your website.

2

  Determine if You Need OMB Approval for Collecting Data

3

  Make the Survey Easy to Take
  • Use online surveys wherever possible, and make them easy to get to, via short web links or QR codes that participants can scan with their phones.
  • Wherever possible, make sure your survey tool that has these features:
    • Progress bars that show respondents how far along they are in the survey process.

    • Conditional logic, so participants automatically skip past questions they don’t need to answer.

  • Format your questions so they can be answered easily or quickly; try using multiple choice or sliding scales.

  • Don’t use too many free text comment fields

4

  Make Sure Every Question Is Actionable and Answerable
  • Before you ask a question, consider how you’ll use the responses. Don’t waste the respondent’s time and energy.
  • Group related questions together.
  • Use language and terminology familiar to your respondents. For specialized audiences, make sure your language is clear and precise.
  • Avoid jargon and spell out acronyms.
  • Ask one question at a time.
  • Don’t ask leading or biased questions.
  • When you use response scales (whether numbers or “highly unlikely” to “highly likely”), make sure those scales are balanced and suitable to the question you’re asking.

5

  Keep These Questions in Mind:
  • How will we administer the survey?
  • Who is the audience for this survey?
  • Will we need additional permission to do this survey?
  • What are we going to do with the responses?
  • What will be our next steps after the survey?

Guidelines

  • The Voice of the Customer DR recognizes that the OMB uses High Impact Service Providers (HISP) as a term to outline a public-facing service. USDA uses a similar term, customer-facing program or service, but it is inclusive of HISPs and broadens it to all public services. All customer-facing programs are encouraged to conduct an A-11 Survey (PDF, 93.3 KB) and should refer to the guidance in the OMB Circular A-11 Section 280 on question wording, response architecture and branching logic.
  • Standardization of customer experience measurement allows for aggregation of data across USDA agencies and programs. The Office of Customer Experience at USDA created standardized guidelines (PDF, 371 KB) that are A-11 compliant and are intended for all customer-facing programs.

Tools (Downloadable Templates and Resources)

Additional Resources

There are many methods you can use to test your existing website or application with users. Your research goal should dictate what methods you use—you might:

  • Conduct surveys to determine general user satisfaction

  • Perform card sorts or tree tests to help you assess how your information is organized (information architecture)

  • Hold usability testing sessions to watch actual users as they interact with a live site or prototype from start to finish

You might use some or all of these testing methods; just be sure your research methods support your research goals.

How to Test Your Site


1

  Determine your goals or research objective for testing.
What do you want to learn and how will you use the information you gain through testing?

2

  Identify top tasks, known pain points, or research questions you would like to explore during testing.

3

  Determine the primary users of your site and recruit those users for testing.
If you don’t have access to typical users, explore ways you can market your research opportunities, or use a recruiter to help find people to take part in your user research.

4

  Select a research team to support testing and analysis.
The team should work together to refine the methodology, create data collection documentation and mechanisms, create a test plan, and support test logistics.

5

  Select the best research methods for your data gathering.
This can range from Google Analytics reports to individual interviews for a more targeted view of user behaviors.

6

  Document the findings about user goals, needs, behaviors, pain points and preferences.

7

  Share findings with your team.

8

  Use findings to prioritize updates that will help users get through tasks.


Tools (Downloadable Templates and Resources)

Additional Resources

User research methods can help you learn why users need a product or service, what works well, and what needs to be fixed or changed. You shouldn’t make any changes without first learning from the people who will benefit most from your work. 

Some research methods you might use include:

  • Surveys to determine general user satisfaction

  • Focus groups, stakeholder interviews, or observing and interacting with people in their real-life environments (ethnographic research) to understand the day-to-day interactions your users are having with your site

  • Card sorts or tree tests to assess your information architecture

  • Usability testing to watch actual users as they work with a live site or prototype and observe how they interact with it from start to finish
     

How to Conduct Research to Identify User Goals, Tasks, and Pain Points


1

  Identify and document the primary goals of your site.

2

  Identify your target user population(s), or the people your project is meant to help or attract.

3

  Come up with a plan for connecting with people who represent your typical users.
Employees, newsletter subscribers, and members of relevant organizations can be good resources to pull from. If you don’t have access to typical users, explore ways you can market your research opportunities, or use a recruiter to help find people to take part in your user research.

4

  Select the best research methods for gathering data to help you understand your users.
These methods can range from using Google Analytics reports for a broad view of user behaviors, to individual interviews for a more targeted view.

5

  Keep user experience in mind
Regardless of what research method you choose, keep these questions at top of mind as you collect and review data:

  • Bright Spots: What were users’ positive experiences?

  • Pain Points: What were users’ negative experiences?

  • User Goals: What is important to your customers? What do they hope to accomplish?

  • Behaviors: What are your users seeing, hearing, and doing? What actions are they avoiding or favoring?

  • Share findings with your team.

  • Use findings to prioritize updates that will help users get through tasks.


6

  Share findings with your team.

7

  Use findings to prioritize updates that will help users get through tasks.


Tools (Downloadable Templates and Resources)

Additional Resources

The term “PII” (as defined in OMB Memorandum M-07-1616) (PDF, 227 KB) refers to information that can be used to determine a person’s identity, either alone or when combined with other information.

To define PII, you need to assess (on a case-by-case basis) the specific risk that it could be used to identify a person. Furthermore, you need to recognize that it’s possible for non-PII to become PII if it’s combined with other information that becomes publicly available.

The following are some examples of data which, when combined with a person’s name, constitute PII. For a decision on other data elements not on this list, you should contact the USDA Chief Privacy Officer.

  • Social Security number

  • Place of birth

  • Date of birth

  • Mother’s maiden name

  • Biometric record (such as fingerprint, iris scan, DNA)

  • Medical history information (including medical conditions and metric information, e.g. weight, height, blood pressure)

  • Criminal history

  • Employment information to include ratings, disciplinary actions, performance elements and standards

  • Financial information

  • Credit card numbers

  • Bank account numbers
     

To report a PII Incident, call: 1-877-744-2968

At USDA, we do not collect PII without consent. Refer to our Privacy Policy for more details on how we protect the privacy of the public who visits our sites.
 

For additional information please reference:

Tools (Downloadable Templates and Resources)

Additional Resources

 

Previous: Research   Next: Analytics

This page was last updated February 15, 2024.