Section 12.5. Recommendations and Open Research Questions

12.5. Recommendations and Open Research Questions

What lessons have we learned about the usability of cryptographic smart cards? How many of these lessons can be generalized? We believe the following recommendations hold for any security device:

The system as a whole matters

Good behaviors, as well as dangerous ones, arise from the interaction of social, human, and technical components. Every evaluation or design activity must take into account the system as a whole.

Devices that require a reader affect mobility

Ease of mobility affects security. If the user must carry around a reader, we must ensure that the reader, host, and security device all connect without potential for gross error. Further, we must devise countermeasures to limit these potential errors.

Add value to the security device^[29]

If the devices have multiple values to users, users will care about them morefor example, they won't forget them. Mass storage or RFID technology for building access are examples of how to increase value.

Simple is secure

Users prefer simple devices to those that are more complex or those that have more parts. Users will also make fewer security errors with them.

Tune and adapt the software

Doing so helps to avoid or limit usability and security issues.

Conduct experimental usability studies, even small ones

Such studies might catch use-time simple issues that could cause global usability and security problems.

This chapter has reported our efforts in attempting to cover the lack of studies on the usability of security devices. The following is a short list of actions that we believe deserve further research efforts:

• Evaluate the effects on usability and security of the interaction between software applications and security devices. Doing so may lead to identifying integrated criteria for system design.

• Provide general abstractions, guidelines, checklists, etc., to educate software developers in implementing software that minimizes usability issues with security devices.

• Provide quantitative usability evaluations for a wider spectrum of security devices.

• Investigate which complementary technologies enhance usability while maintaining comparable or increased levels of security. For example, are USB tokens with fingerprint sensors both more usable and at least as secure as normal USB tokens? Can we keep integrating additional technologies without impairing usability?