DUOLINGO
Re-designing the UI
This project includes a usability study of the Duolingo interface, a new prototype and a heuristics evaluation of the design proposal.
Problem statement
Role
Duolingo relies on translation as the main method of teaching a language. Users encounter problems keeping engaged in the interface beyond basic level exercises. I redesign part of the UI to address this problem.
I conducted a usability study of the interface, redesigned part of the UI and performed a heuristics evaluation of the prototype.
Insights
Duolingo engages learners of a foreign language through gamification, visual vocabulary, repetition and visual positive reinforcement. Through practice and repletion it facilitates the channeling of learning and remembering. However, for intermediate users, repetition poses a burden on language learning. Reliance on translation interferes with engaged past intermediate level.
USABILITY STUDY
Background
Duolingo is a language learning platform and mobile app that people can use for free. The UI guides users through a set of language learning exercises, which are sequentially structured, adding levels of difficulty depending on the level of proficiency.
Participants
Some experience using the UI, having used the interface more than once before the study.
Basic level (A1) to intermediate level (B1) in the target language.
Methodology
Questionnaire
I administered a pre-interview demographics and usability questionnaire which included usage habits and basic demographic questions.
Interviews
First, participants answered questions about their experience with learning a foreign language, and the app. Next, participants were guided through a set of questions in the task-based interview where the researcher took interview and observational notes as participants moved through the unit.
Tasks
Participants were guided to open the UI and take the language placement exercises. Once the system identified their proficiency level they were guided through the translation based unit.
Participants were asked to complete a translation based exercise provided by the UI from start to finish.
​
Results
Experience with UI
All participants rated their experience 4/6 ranked from poor to excellent in Likert scale.
Time spent on the UI
All participants reported they used the service 10 minutes once a week
Gamification
Gamification was the most effective aspect of the UI in keeping participants engaged
Correcting mistakes
All participants wanted to correct their mistakes but were not given paths for error correction.
Raking of features usefulness in Likert scale
Users ranked the usefulness of features from least useful to most useful on a 1-6 Likert scale. Written translation, Audio translation and Vocabulary were ranked as the most useful.
Tactical level pain points
Participants had a hard time understanding word boxes could be rearranged. No visual cues of affordances of features. Resulted in participants with the autonomy to fix their errors
Tactical level pain points
Participants were confused when the word boxes didn't give all cues for the task it was expecting for them to complete. For instance, lack of question marks left participants confused if they should answer the question or translate it
Overuse of translation hindered usability
The usability of Duolingo tis jeopardized by its reliance on translation between mother language and target language. This created confusion in participants' learning.
" You can't think in the language you are learning. This delays the learning process. I mean, you are constantly going back to your own language" (P1)
"At first it is great, you repeat things and it is helpful. But when you are beyond beginner level, you repeat way too much. It comes to the point in which it becomes tiring and it stops being fun" (P1)
Repetion hindered learning
Participants in the intermediate level of target language, expressed frustration when prompted to repeat exercises from prior modules without having options to skip these.
Conclusion
Based on the findings of the usability study, Duolingo needs to re-design the interface having into consideration the learning process for all levels of language proficiency and not rely on purely translation based exercises.
Additionally, design choices must give users options to correct their mistakes. It could also reinforce learning through repetition for beginner learners and offer intermediate learners options to move forward.
At a tactical level, usability could be improved by highlighting the affordances of buttons and text boxes.
Prototyping
The redesign of the UI addresses the main findings from the usability study and proposes two solutions.
1
Engaging the user in language learning through other paths beyond translation
​
2
Incorporating a pathway for correcting their mistakes
Wireframing
Happy path
Original Duolingo Translation based Exercise
Error Correction
Original Duolingo Translation based Exercise
Refined prototype
Duolingo Happy Path
Duolingo error correction path
Heuristics Evaluation
1
Visibility of System Status
The prototype includes a more detailed stepper bar to signal the user where they are
2
Match between system and real world
The prototype keeps the use of colloquial language to match real wold conversations. At a tactical level, it keeps icons that represent objects of the real world
3
User control and freedom
The prototype allows the user to skip to the next exercise or correct their mistakes before moving to the next step.
4
Consistency and standards
The prototype keeps the consistency of button and overall visual language and organization patterns Duolingo employs in the UI.
5
Error prevention
This prototype does not address any error prevention affordances in the system.
6
Recognition rather than recall
The prototype provides multiple visual cues that instructs the learner what to do and where in the process they are. This prototype added a stepper to see progress.
7
Flexibility and efficiency of use
The prototype added drop and drag action for text boxes to improve efficiency.
8
Aesthetic and minimalist design
Duolingo uses a minimal approach to their visual language. The prototype keeps the original aesthetics of the UI.
9
Help users recognize, diagnose and recover from errors
This prototype did not address this aspect
10
Help and documentation
The prototype included the addition of a help sidebar for correcting mistakes