ORIGINALITY
MY ROLES
LED USER RESEARCH to identify problem spaces in academic integrity, and validate the usability of our analytics tools
COLLABORATED WITH MACHINE LEARNING team to discover new opportunities for decades of academic data
LED USER RESEARCH to identify problem spaces in academic integrity, and validate the usability of our analytics tools
COLLABORATED WITH MACHINE LEARNING team to discover new opportunities for decades of academic data
PROTOTYPED multiple iterations of our product from proof of concept to enterprise release.
IMPACT
+2000%
increase in contract cheating detection at University New South Wales16,000
active institutions using Originality product① PROBLEM
In 2018, we developed a brand new product to identify student assignments written by a different student or essay mill (known as contract cheating). The product itself was simply a very long report of machine intelligence - determined factors that may be evidence of contract cheating.
Beta testing revealed that non-subject matter experts had no idea what the data signified, or where to start.
Beta testing revealed that non-subject matter experts had no idea what the data signified, or where to start.
② RESEARCH
INSIGHT
Academic integrity is very serious, and accusations can be very stressful on students.
Academic integrity is very serious, and accusations can be very stressful on students.
Certain types of evidence are more effective in making a contract cheating case.
Factors that imply cheating are often interconnected.
Investigations can take a long time.
OPPORTUNITY
1. Move that evidence to the top of the report.
2. Clarify what all data means and why it is important. Suggest when to investigate but never accuse a student of cheating.
3. Display visible changes in a student’s work over time.
4. Provide more tools for taking notes and reviewing at a later time.
1. Move that evidence to the top of the report.
2. Clarify what all data means and why it is important. Suggest when to investigate but never accuse a student of cheating.
3. Display visible changes in a student’s work over time.
4. Provide more tools for taking notes and reviewing at a later time.
③ CONCEPTS
④ DESIGNS
1. BE CONCISE
I condensed much of the report’s data into logical groups. Users now had more flexibility with sorting and filters and customizing table columns.
I condensed much of the report’s data into logical groups. Users now had more flexibility with sorting and filters and customizing table columns.
2. PRIORITIZE
I added a fixed report summary that directly exposes the most important issues in a student’s work, and jumps to that section of the report.
I added a fixed report summary that directly exposes the most important issues in a student’s work, and jumps to that section of the report.
3. SHOW TRENDS
I organized much of the data into interrelated graphs. This would help users recognize patterns in a student’s work over time.
I organized much of the data into interrelated graphs. This would help users recognize patterns in a student’s work over time.
4. PROVIDE TOOLS
In the previous designs, users only had one place to leave notes. I added the ability to tag and leave comments on individual assignments.
In the previous designs, users only had one place to leave notes. I added the ability to tag and leave comments on individual assignments.
⑤ LESSONS LEARNED
1. JUST ENOUGH DATA
When you have a lot of data and an eager machine intelligence team, it’s tempting to show as much of that data as possible. Our users, however, have limited time and resources to find patterns in a mass of information. Once we learned that a certain subset of clues was most effective in making a cheating case, I pushed for that data to be the most visible.
When you have a lot of data and an eager machine intelligence team, it’s tempting to show as much of that data as possible. Our users, however, have limited time and resources to find patterns in a mass of information. Once we learned that a certain subset of clues was most effective in making a cheating case, I pushed for that data to be the most visible.
2. EMPATHY EXTENDS BEYOND USERS
Contract cheating is often symptomatic of a student who needs academic assistance or is having personal difficulties. By making a more intuitive tool, not only were we helping administrators build a case faster, but also providing students an intervention before facing expulsion.3. SOLVING AN UNKNOWN PROBLEM
Contract cheating is still new in the academic world. Our superusers are very skilled and very vocal, so it was difficult to design a solution for others. I realized that new users needed much more guidance, with clear instruction on what to look for and why.TEAM
ProductMark Ricksen - Product Manager
Ryan Crews - Front End Engineer
- Role
- Senior UX Designer
ProductMark Ricksen - Product Manager
Ryan Crews - Front End Engineer