CDC Health Portal

Companies frequently speak their own language, using unfamiliar content and acronyms that do not communicate in the language of their customers. CDC joined with User Insight to perform various methods of user research including one on one interviews, remote user testing, prototyping, iterative testing through JiTR, surveys, card sorting and usability testing, resulting in improved communication with all six of their diverse consumer groups.

Centers for Disease Control and Prevention (CDC) engaged User Insight to design a major data portal for a variety of user types. This large, multi-phase project began with concept research, proceeded with prototyping and iterative user testing, and ended with User Insight’s team consulting throughout CDC’s development of the final application.From concept to coding, User Insight helped CDC ensure the application design met user needs and employed usability best practices.

Client Goals

  • Exploration of user needs and design concepts.
  • Deeper understanding of user types.
  • Well-tested portal design.
  • Assistance with design implementation.

Insights Gained

Phase 1: Research and Requirements

  • Searching (via Google) was the prominent method for accessing health related information.
  • CDC was viewed as a primary source of credible information on the topic.
  • All groups desired to “see themselves” in the data.

Phase 2: Task-Driven Exploration

  • Users wanted information to be pre-populated into queries based on information currently being researched.
  • Geographic results were expected to be associated with interactive maps.
  • In-depth scientific information was expected upon deeper probing of content.
  • Expectations of access to content, data analysis and publishing privileges varied according to User group.

Phase 3: Card Sort and Clustering

  • The initial groupings identified in Phase 1 were upheld as top level clusters.
  • Clustering via Natural Associations
    • Health Condition remained the most stable category.
    • Hazard was the least stable category of those defined in previous Phases of research.
  • Tenuous relationships clustered the majority of indicators.
  • Top 5 terms were confirmed by 75% of Users in 3 of the 5 categories:
    • Health Effects
    • About Tracking Network
    • Location
  • Exposure was the least substantiated category with no more than 64% of Users placing any one term into the folder.
  • Top 15 terms for each category had little cross-over with the exception of air related terms.

Phase 4: Iterative Testing during Development

  • Users had no clear preference for positive versus negative terminology and imagery.
  • Unconventional action sequencing was confusing.
  • Query Forms appeared at unexpected points within the navigation process.
  • Users desired consistency between pages of equal depth within the site.

Our Approach

Phase 1: Research and Requirements

  • 42 user assessments – a mix of one-on-one interviews, contextual interviews, and remote testing – within a four week period to gain feedback on user preferences and recommendations for the portal’s look and navigation.
  • Competitive analysis on sites with similar functionalities to gather insight for user preferences and expectations.
  • Requirements gathering based on user feedback.
  • User profiles to distinguish how each user type would want to interact with the portal.
  • Information architecture to drive a prototype for testing the design concept, including page layout and functionality.

Phase 2: Task-Driven Exploration

  • Prototyping of a mock portal to test with users.
  • One-on-one interviews, contextual interviews, and remote testing to gather feedback about content expectations, as well as layout and navigation preferences for the tracking portal.
  • Usability evaluation by studying the paths users took through the prototype.
  • Refinement of requirements and user profiles to reflect different interaction styles across user groups.

Phase 3: Card Sort and Clustering

  • Web-based card sort activity to understand how well proposed information categories matched user expectations and needs.
  • Paper card sort with NEHA members to help define portal nomenclature.

Phase 4: Iterative Testing during Development

  • One-on-one interviews to pinpoint user preferences for specific functionalities and cross-referencing opportunities.
  • Design recommendations to address the ways users prefer to interact with each version of the portal.

Who We Talked To

For each of the phases listed below, User Insight talked to a mix of the following users:

  • NAHDO Conference Attendees
  • Non-Expert Level Health Care
  • Business and Industry Health Care
  • Not for Profits
  • General Public
  • Politicos and Media invoved in Health Care

Phase 1: Research and Requirements

42 one on one in lab and contextual interviews in California, Atlanta and Washington, D.C.

Phase 2: Task-Driven Exploration

70 one on one in lab, contextual interviews and remote technology sessions in 13 states.

Phase 3: Card Sort and Clustering

272 online card sorting surveys and 15 experts in group sessions in Washington, D.C.

Phase 4: Iterative Testing during Development

15 User Assessments and JiTR sessions in Atlanta, GA