View this document as: a single page | multiple pages.

Usability Considerations

This section is informative.

In order to align with the standard terminology of user-centered design and usability, the term “user” is used throughout this section to refer to the human party. In most cases, the user in question will be the subject (in the role of applicant, claimant, or subscriber) as described elsewhere in these guidelines.

Ergonomic of Human-System Interaction — Part 11: Usability: Definitions and Concepts [ISO/IEC9241-11] defines usability as the “extent to which a system, product or service can be used by specified users to achieve specified goals with effectiveness, efficiency and satisfaction in a specified context of use.” This definition focuses on users, goals, and context of use as key elements necessary for achieving effectiveness, efficiency and satisfaction. A holistic approach considering these key elements is necessary to achieve usability.

From the usability perspective, one of the major potential benefits of federated identity systems is to address the problem of user fatigue associated with managing multiple authenticators. While this has historically been a problem with usernames and passwords, the increasing need for users to manage many authenticators — whether physical or digital — presents a usability challenge.

As stated in Sec. 8 of [SP800-63A] and Sec. 8 of [SP800-63B], overall user experience is critical to the success of digital identity systems. This is especially true for federated identity systems, as federation is a less familiar user interaction paradigm for many users. Users’ prior authentication experiences may influence their expectations.

The overall user experience with federated identity systems should be as smooth and easy as possible. This can be accomplished by following usability standards (such as the ISO 25060 series of standards) and established best practices for user interaction design.

Guidelines and considerations are described from the users’ perspective.

Section 508 of the Rehabilitation Act of 1973 [Section508] was enacted to eliminate barriers in information technology and require federal agencies to make electronic and information technology accessible to people with disabilities. While these guidelines do not directly assert requirements from Section 508, identity service providers are expected to comply with Section 508 provisions. Beyond compliance with Section 508, Federal Agencies and their service providers are generally expected to design services and systems with the experiences of people with disabilities in mind to ensure that accessibility is prioritized throughout identity system lifecycles.

General Usability Considerations

Federated identity systems should:

Specific Usability Considerations

This section addresses the specific usability considerations that have been identified with federated identity systems. This section does not attempt to present exhaustive coverage of all usability factors related to federated identity systems. Rather, it is focused on the larger, more pervasive themes in the usability literature, primarily users’ perspectives on identity, user adoption, trust, and perceptions of federated identity space. In some cases, implementation examples are provided. However, specific solutions are not prescribed. The implementations mentioned are examples to encourage innovative technological approaches to address specific usability needs. See standards for system design and coding, specifications, APIs, and current best practices (such as OpenID and OAuth) for additional examples. Implementations are sensitive to many factors that prevent a one-size-fits-all solution.

User Perspectives on Online Identity

Even when users are familiar with federated identity systems, there are different approaches to federated identity (especially in terms of privacy and the sharing of information) that make it necessary to establish reliable expectations for how users’ data are treated. Users and implementers have different concepts of identity. Users think of identity as logging in and gaining access to their own private space. Implementers think of identity in terms of authenticators and assertions, assurance levels, and the necessary set of identity attributes to provide a service. Given this disconnect between users’ and implementers’ concepts of identity, it is essential to help users form an accurate concept of identity as it applies to federated identity systems. A good model of identity provides users a foundation for understanding the benefits and risks of federated systems and encourage user adoption and trust of these systems.

To minimize the personal information collected and protect privacy, IdPs ought to provide users with pseudonymous options for providing data to RPs, where possible, and inform users of the benefits and drawbacks of pseudonymous identification. Likewise, RPs ought to request pseudonymous options for users when pseudonymity is possible for the RP’s policy. Both IdPs and RPs need to seek to minimize unnecessary data transmission and inform users of which information is transmitted and for what purpose.

Many properties of identity have implications for how users manage identities, both within and among federations. Just as users manage multiple identities based on context outside of cyberspace, users must learn to manage their identity in a federated environment. Therefore, it must be clear to users how identity and context are used. The following factors should be considered:

User Perspectives of Trust and Benefits

Many factors can influence user adoption of federated identity systems. As with any technology, users may value some factors more than others. Users often weigh perceived benefits versus risks before making technology adoption decisions. It is critical that IdPs and RPs provide users with sufficient information to enable them to make informed decisions. The concepts of trust and tiers of trust — fundamental principles in federated identity systems — can drive user adoption. Finally, a positive user experience may also result in increased user demand for federation, triggering increased adoption by RPs.

This sub-section is focused primarily on user trust and user perceptions of benefits versus risks.

To encourage user adoption, IdPs and RPs need to establish and build trust with users and provide them with an understanding of the benefits and risks of adoption. The following factors should be considered:

User concern over risk can negatively influence willingness to adopt federated identity systems. Users may have trust concerns, privacy concerns, security concerns, and single-point-of-failure concerns. For example, users may be fearful of losing access to multiple RPs if a single IdP is unavailable, either temporarily or permanently. Additionally, users may be concerned or confused about learning a new authentication process. In order to foster the adoption of federated identity systems, the perceived benefits must outweigh the perceived risks.

User Mental Models and Beliefs

Users’ beliefs and perceptions predispose them to expect certain results and to behave in certain ways. Such beliefs, perceptions, and predispositions are referred to in the social sciences as mental models. For example, people have a mental model of dining out that guides their behavior and expectations at each establishment, such as fast food restaurants, cafeterias, and more formal restaurants. Thus, it is not necessary to be familiar with every establishment to understand how to interact appropriately at each one.

Assisting users in establishing good and complete mental models of federation allows users to generalize beyond a single specific implementation. If federated identity systems are not designed from users’ perspectives, users may form incorrect or incomplete mental models that impact their willingness to adopt these systems. The following factors should be considered: