Personalized and Adaptive Course

Project Goal: Deliver a personalized, adaptive learning experience and insights that reduce time to proficiency

PLEASE NOTE: Due to confidentiality, detailed images are note used on this page. Alana will speak to the details of this project live. Images speak to the project overview and Alana’s workflow process. 

OVERVIEW

The who: The Learning Genome (LG) is Amazon’s mechanism for data-driven acceleration of knowledge and capability development.  The LG value proposition is that Amazon can improve and accelerate learning—not only through better learning design grounded in the science of learning, but also through data-driven feedback and actionable insights that support human decision-making. 

The why: Personalized and adaptive learning at a high level provides differentiated experiences to each learner, based on what that learner needs. Personalized learning is important because learners benefit from it, business and program owners want it, and the LG needs it to raise the bar on data-driven human-in-the-loop decision making for learning and learning design. 

The project: Deliver a personalized, adaptive learning experience and insights that reduce time to proficiency.

11

PROCESS

11

Research and hypothesis – First I needed to create an ecosystem of all the we know, need to know and is outstanding. This included information gathering, data mining, creating personas, competitive research, insights, hypothesis and logistics.

11

Design Sprint – I built and facilitated a remote week-long team design sprint to generate ideas from all angles using a diverse group of team members. By the end of the week, we developed a working prototype to test and build from. Sprint exercises included:

  • Hypothesis and risks
  • Lightening Demos and voting
  • 4-step sketch and storyboarding
  • Building a working prototype
  • Testing 
11

Testing – Tested prototype with 10 unique users. Gathering data at this stage consisted of breaking up the user experience of the prototype and asking pointed questions to our users as they went through the prototype. The result from the testing was:

  • Qualitative/quantitative data
  • Data-driven decisions
  • A foundation to build from
  • More things to consider
11

Discovery and findings – Through our user research channel, competitive analysis, review of research and data, I was able to make solid decisions about the product to funnel and target ideas. Detailed research included:

  • Filtering findings
  • Categorizing trends
  • Finding patterns
  • Making a list of focus points
  • Gather insights from all directions
11

Wireframe – The prototype was turned into solid designs using the research and data to make informed decisions. User flows and stories are apart of this phase with close attention to accessibility and sustainability.

11

Design – Once the wireframes went through a series of bug bashes and solid user stories layouts, I was ready to incorporate visual elements and match style guide and accessibility standards.

11

Handoff – With solid designs in place, I worked closely with the tech team to build this product and deliver it to the users. Detailed notes were included to confirm communication was solid.

Test and Gather – Once the tech team completed their build, we were able to deliver the product to the user. We ran two versions of the course, the classic version and the adaptive version (approx. 2k users per month running for four months)

Results

  • Created a trustworthy partnership with the client to create more courses
  • Created user trust within their experience
  • Users in the adaptive course finished with higher percentage in proficiency and time

Highlights

  • Proficiency model – The products system was a way to deliver a course that adapts selection and sequencing of educational activities automatically to the user instead of a one-size fits all approach
  • Pretest that informs the system of their prior knowledge state using the proficiency model
  • Intelligent Coach that informs and supports the user about how the recommendations are leading them to proficiency
  • Path page navigation – shows user learning goals and no sequence to break the barrier of traditional learning
  • Personalization and adaptivity throughout the course experience

User experience feedback

  • Users indicated a positive experience overall, indicated the user interface was clean and modern, had high trust in the recommendations, and felt like it was a good use of their time
  • They found the reference materials to be valuable and recommended navigation very helpful
  • Users found elements of the Intelligent Coach feature useful. They really liked the feedback presented upon answering an assessment, and generally like the guidance and support

Follow-up

  • Run experiments to evaluate whether the adaptive course is effective to reduce the time to proficiency (to improve business outcomes) compared to the classic course. Collecting more data is important.