#9

UI/UX Design: Show less personal information on interfaces – and show it at later stages

Designing fair and inclusive interfaces

Interfaces serve as the accessible entry – the visible face – of systems that are otherwise complex and opaque to most users. Job searches and applications should be a straightforward experience that is accessible to all, regardless of background or ability. Hence, designing inclusive interfaces is essential to reduce bias and to ensure equitable access. Inclusive design includes, amongst many aspects, thoughtful presentation of information that is tailored to different stakeholders using the system and that supports fairness and clarity throughout the process.

Recruiters’ search behaviour

As recruiters are tasked with searching and finding the best candidate for a given job description, their search behaviour might be different from the typical behaviour of an individual when searching and traversing a list of items. Recruiters are more likely to examine more results from the ranked list.1 As such, the ranking shown to the recruiter can have an impact on the fairness of the selection.2 For comprehensive exploration, the recruiter is more interested in viewing as many results as possible. Therefore, the amount of attention spent on the profile of the candidate might be lower as opposed to targeted refinement where the recruiter is more interested in checking the profile of the candidate for the required specialised skills or background, thus, spending more attention on a specific candidate’s profile.

Information – display less and show it later

The information displayed about each candidate influences the first impression of recruiters and their likelihood to click on the profile or contact the candidate. To avoid highlighting information that may lead to discrimination, the interface should display only information that is relevant to the corresponding job, such as work experience or skills. Revealing the name of the candidate should happen as late as possible in the recruitment pipeline to ensure that the initial shortlisting is based on relevant experience.3 Thus, it is also essential to avoid displaying information (e.g. name, photo, phone number, address, location/name of university/company etc.) that could serve as a proxy for protected characteristics such as gender, nationality, or other data. However, it is important to keep in mind that even without obvious proxies, details like work experience and education can still hint at protected characteristics (e.g. the school one went to can serve as a proxy for religion).–> see recommendations in chapter #6

Inclusive and Accessible UI/UX – welcoming by design

If users struggle to navigate a recruitment platform, their chances of finding a suitable job decrease. In order to identify potential problems of the User Interface/User Experience (UI/UX) one could create a focus group composed of people from various backgrounds. FINDHR research4 has shown that many users, especially users with migrant backgrounds, might be using smartphones to access recruitment platforms. Therefore, the UI and all features should be designed fully mobile-friendly. Moreover, many of the existing job application websites pose unnecessary impediments for people with visual impairments. Thus, the UI should be accessible to users with visual and/or auditory as well as other types of impairment.5

→ for more information and guidelines on inclusive interface design, consult the FINDHR Software Development Guide, section 2.5.

Recommendations for inclusive interface design
  • Be aware of position bias and other biases that can be enforced by the UI/UX.
  • Collaborate with recruiters to understand their needs in professional search.
  • Design interfaces to be inclusive and accessible with respect to various user needs, including users with impairments.
    • Design UI in a fully mobile-compatible way, since many users, especially migrants or those with fewer technical skills or resources, access recruitment platforms via smartphones.
    • Ensure that the UI is inclusive and easy to follow especially for people with less technical literacy.

 

  1. Kim et al. (2011)
  2. Alvarez et al. (2023)
  3. Lin, J. (2023)
  4. César et al. ( 2023)
  5. see Reuschel et al. (2023).

Privacy Preference Center