Production, Circulation, and Drainage

Cerebrospinal Fluid:

An interactive tool for undergraduate medical students

Master's Research Project:

learning biostatistics

Background & Rationale

The undergraduate medical program at the University of Toronto has recently incorporated a two-year Health Science Research (HSR) course into their pre-clerkship (Foundations) curriculum. This course aims to help students understand and use research to contribute to the improvement of individual and population health. Currently, the HSR is taught using an asynchronous “blended” classroom approach, which requires students to learn course content outside of class via e-modules and dedicates class time to student-centred learning activities. Among the course’s diverse subject matter, biostatistics is a topic for which students require more support. This suggests that the teaching strategy employed for this conceptually-abstract content may not be optimal for learning. Since a conceptual understanding of biostatistics is essential for critical appraisal of medical literature and the application of research to clinical scenarios, optimization of the learning environment for this material is desirable.

Goals​

  • Create an interactive multimodal* e-learning environment for biostatistics

  • Contextualize content in relevant clinical cases

  • Draw upon principles of problem- and case-based learning to promote active learning of content

  • Improve interest and learning outcomes in biostatistics as taught within the context of HSR course

  • Employ a user-centred design approach

         *utilizing both semantic (verbal) and visual (non-verbal) modalities

Committee

Dr. Shelley Wall, AOCAD, MScBMC, PhD, FAMI

Primary supervisor, Biomedical Communications

Dr. Derek Ng,  MScBMC, PhD

Voting member, Biomedical Communications

Dr. Tim Guimond, MD, FRCPC, PhD 

Content expert, Faculty of Medicine

Dr. Debra Katzman, MD, FRCPC

Voting member, Faculty of Medicine

Audience

Pre-clerkship undergraduate medical students enrolled in Health Science Research course  (Foundations Years 1 & 2, University of Toronto)

Format

Interactive, responsive, web-based application

Date

May 2018 - July 2019

iMac landpage mockup.jpg
Project Development

Overview:

User personas  &

context scenarios

Identify communication

challenge

Project

proposal

Content

development

Content mapping & wireframing

Define problem space, project scope, & timeline

Prototyping &
user testing

Proof-of-concept tool

Step 1:

Identify a communication challenge

This project began with my interest in scientific research and my curiosity about whether/how clinicians used findings in the research literature to inform their clinical practice. I decided to investigate further by speaking to practicing clinicians as well as medical students at the University of Toronto. From the interviews with medical students, I learned of a relatively new course on evidence-based medicine (Health Science Research, or HSR), which prepares students to be consumers of medical research. Students identified this course as one which could benefit from improvements to the mandatory e-learning tools.

 

Based on this information, I further investigated the HSR e-modules to identify specific knowledge gaps that might be overcome with visual solutions. I learned that HSR's biostatistics modules presented the most learning challenges for students - as evidenced by students' increased need for support in this subject area relative to other HSR subject areas. In reviewing the biostatistics e-modules in particular, I observed that they were poor in visualizations and afforded few opportunities for active engagement with the content. This was a perfect opportunity for the use of interactive visualizations to support learning of this complex, abstract material.

Step 2:

Define the problem space, project scope, & timeline

With a challenge identified, I sought to discover who would use or be impacted by a visual, interactive learning tool for biostatistics, why they would use such a tool, and what other tools exist to achieve similar goals.

Who 

To determine the who, I identified and met with stakeholders- HSR course director, Dr. Debra Katzman, and HSR tutorial leader, Dr. Tim Guimond- who then joined the project as supervisors and subject matter experts (SMEs). Meetings with these stakeholders helped me identify the target audience: pre-clerkship undergraduate medical students.

Why

To determine the why, I conducted preliminary needs assessments to determine the needs, goals, and motivations of the target audience. Methods included informational interviews with medical students and their proxies (HSR course representatives), analysis of HSR course evaluation data, and analysis of HSR exit survey data.

What

To determine the what, I researched existing solutions to the communication challenge, including literature review and media audit. Key questions I sought to answer through my researched included:

  • What strategies are employed for teaching biostatistics in undergraduate medical curricula, and what are their theoretical underpinnings?

  • How do different teaching strategies affect learning outcomes or attitudes towards learning and/or the content itself?

  • What visual or interaction-centred tools exist for teaching biostatistics, and what are their relative merits?

Upon completion of this preliminary research, the project's scope was determined in consultation with project supervisors and stakeholders, and a project timeline was created:

GANTT chart 2018-2019 for committee meet

Step 3:

Project Proposal

Upon completion of the preliminary research stage, I drafted and submitted a formal project proposal for approval by my project committee. You can read my full project proposal here.

In brief, I proposed to develop a series of interactive e-learning modules using a user-centred design approach.

Step 4:

Pre-production

I. User personas & context scenarios

With my project proposal approved, I developed user personas and context scenarios based on information gathered in informational interviews with medical students and discussions with stakeholders. This helped me better understand the needs and goals my tool's end-users were likely to have and how they might use the tool to achieve their goals. These exercises guided my initial design of the tool's functionalities and behaviours.

 Multiple iterations of the personas and scenarios were created in response to feedback from peers, project supervisors,  and stakeholders. Increased understanding of the target audience and tool usage context over time also spurred the iterative process.

Take a look at an early iteration here and a later iteration here.

II. Content mapping & wireframing

User personas and context scenarios provided a basis upon which I began mapping the structure and flow of the content within the e-modules. Initially, in the absence of module content, I developed a framework for the categories of information I envisioned would be included. As the content was eventually developed, I iterated the information architecture of the modules to accommodate the information that would be included.

MRP structure_v001.png
MRP structure_v002.png
MRP structure_v003.png

At the same time as content mapping, I also began brainstorming how module pages would appear (UI design), focusing on user interface elements critical for navigation and interactive functioning on each page. I captured these ideas in low-fidelity wireframes.

wires for MRP structure_v003_3rd iterati
wires for MRP structure_v003.png

III. Content development

Module content was developed in parallel to content mapping and wireframing. As the content was not covered in the course at the time, it needed to be created from scratch. To do this, I worked closely with the SME in a content co-creation process:

  • Designer requests specific content or resources to create content

  • Content supervisor provides content or resources to create content

  • Designer processes content or resources, and translates them into content appropriate for the target audience

  • Designer verifies accuracy of new content with SME and adjustments are made
    (if necessary) 

In this way, we developed content for 1 variable type and 2 distributions types using a clinical research case from psychiatry. This addition of this content into the tool drove further iteration and refinement of the information architecture and wireframes.

IV. Prototyping & user testing

Information architecture and wireframes were translated into a moderate-fidelity interactive prototype in Adobe XD, and final content was added as it became available. This prototype was tested with current HSR students (end-users) and their proxies to assess the usability of the design.

 

Based on the feedback collected during user testing sessions, the prototype was refined and its fidelity elevated before retesting with a new group of end-users. We iterated this prototyping process for a total of three rounds of user testing (12 participants total). The final prototype was web-based, built using HTML, CSS, JavaScript, and the d3.js library.

prototype1_single screen.png
prototype1_all screens.png
prototype2_single screen.png
prototype2_all screens.png
prototype3_single screen.png
prototype3_all screens.png

Step 5:

Production

The project culminated in the development of a web-based, proof-of-concept prototype of MedStats informed by 3 rounds of user testing with target end-users (undergraduate medical students) and their proxies (MScBMC students).

Demo coming soon!

While beyond the scope of the Master's Research Project per se, further development and refinement of the MedStats e-modules is necessary to address all end-user needs revealed during user testing, and additional user testing will be required to evaluate any proposed design changes. This process is currently ongoing