Designing an inclusive learning discovery pathway for disenfranchised learners

Te Pūkenga - Vocational education - New Zealand

When vulnerable learners consider furthering their education, there are critical journey points where they can't or won't commit without sufficient support. We targeted these key moments in our AI-powered information way-finding journey, as well as offering a seamless escalation to a real human when needed.

Background: Te Pūkenga is the name of the amalgamation of 25 of New Zealand’s work-based training organisations, institutes of technology and polytechnics, into one organisation.

Visual way-finding to support low literacy

We conceptualised how we might help learners with variable literacy levels and complex needs to feel more confident in identifying the best learning pathway for them. We learnt that using a friendly human avatar, plain English, and pairing key options with clear visuals, meant learners would self-serve for longer before needing to book a call.

We moved beyond simple web accessibility, to focus on inclusive and equity-centred design. Considering to apply for higher learning is a brave topic for vulnerable learners; with issues like self-esteem, low income, and a history of failure in the traditional education system coming to the fore. It was important that the digital experience on Te Pūkenga’s website helped them feel comfortable and empowered to take the next step forward.

Upon the public release of OpenAI's GPT-3, Ghost's UX leads jumped at the opportunity to leverage this new technology within an intelligent search experience across 16 New Zealand Institutes of Technology and Polytechnics, and 9 Industry Training Organisations.

For some users, the standard chatbot entry point felt too abstract, and the bounce rate was higher than desired. Our design above features audience-specific use cases before the chatbot entry, as an alternate, more human way to access the most relevant chatbot section for people’s needs.

While the pairing of visuals with way-finding text is helpful for those with low-literacy levels, it’s less helpful for sight-impaired users as screen readers  read out captions for every image, meaning a double up of content. UserWay’s accessibility toolbar provides a simple option to hide all images, so the experience works for everyone.

Customer care research

In almost any organisation, the customer care team hold a gold mine of insights, especially around audience needs that are not being served well online.

In November 2021, we interviewed 7 customer service representatives across Te Pūkenga providers, and then conducted an online survey and card-sort activity with 33 representatives covering 13 of the 25 providers. We learned what value the chatbot could bring to help take pressure off call centre staff, and identified gaps where particularly disenfranchised Māori and Pasifika learners may be falling through the cracks in the current ecosystem. These ‘topic pillars’ formed the roadmap for evolving the conversational AI’s areas of expertise.

Highlights of key insights from interviews:

  • Not fully understanding the time commitment required to complete a programme can lead to drop outs later.
  • Where you live has a big impact on which programmes suit you. Moving to a new place for education can be a huge barrier for people.
  • Emotional support / helping people feel confident is a huge part of stopping people falling through gaps.
  • There is nothing to help users find foundational courses across institutions in the network. Customer service representatives tread lightly on this topic - it's embarrassing for people to hear they need to start with the foundations first.

Excerpt of one of our questions around a particularly prominent gap:

Key takeaway

Through collaboration with our client's service and support teams, and focusing our research on the most vulnerable learners, we paired visual way-finding and a sense of real care with the intelligent search experience.

Download
SHARE ARTICLE

More case studies