Designed a web application (Polly) that allows students to learn effective + accurate punctuation practices by speaking aloud (using rhetorical grammar and prosodic cues).

For the last decade, punctuation misuse has consistently ranked top #20 most common issues in college students' writing (Lunsford & Lunsford 2008).
"Because there's a lot of grammar rules, it gets confusing."
-- university student, user survey response
You are a university student. You believe that correctly and effectively using punctuation means you need to memorize grammar rules. But rules are plenty, so punctuation feels unintuitive. You're discouraged from revising your writing at all, preferring to delegate this task to AI tools.
“I rely too much on auto-correct features and ChatGPT.”
-- university student, user survey response
But these tools don't help you learn: Grammarly, Google Docs, or ChatGPT simply correct your writing or cite rules (agh!) to explain their corrections. You rarely understand why the change is better.
But... punctuation is just a matter of grammar rules, right? Not quite!
The punctuation of a ‘good’ writer is meaning-dependent, not grammar-dependent.
-- Dawkins (1995)
Students want to revise independently and meaningfully, but lack tools that support that kind of learning. An intuitive tool for practicing revision can empower students to revise their writing without being overburdened by grammar rules.
Furthermore, this meaning-dependent revision process can be made embodied.
User reads (performs) their sentence aloud.
.png)
Polly generates a graph of the user's intonation pattern. User reviews significant changes in intonation, which denote potential information unit boundaries.
.png)
Polly guides users to revise their sentence using meaning-based grammar, with their intonation pattern as cues.
.png)
// Polly's approach to revision and grammar is informed by extensive literature review and SME interviews that are not captured on this site - contact me for full project details :-)
Using our MVP, we conducted a within-subject experiment with 9 participants (university students in Rochester, a range of class years and years of instruction in English). We measure success based on punctuation accuracy and confidence compared to baseline (prior to using Polly).
After using Polly, participants are able to identify more punctuation issues (+26%) and of more types; all attempts to revise issues are correct (+25%). Participants revising more than 4 sentences using Polly showed the clearest improvement.
After using Polly, participants are more confident about identifying and correcting punctuation issues (self-rated 3.9/5, +0.5).
Polly encourages participants to develop concrete strategies to identify punctuation issues. Before using Polly, 5/9 participants expressed reliance on “intuition” and 4/9 on grammar rules. After using Polly, 7/9 participants read aloud to revise sentences, integrating speech patterns with punctuation placement for all punctuations.
“I used to only read aloud very complicated texts, but it helps with revising.”
—D. A, '24 (Participant)
“Incorporating more vocal patterns made revising easier.”
—N. M., '25 (Participant)
While not statistically significant, Polly demonstrates promising results.
// contact me for full project details
I would like to enhance Polly's voice recording capabilities by adopting a baseline approach, where users establish a baseline speech profile by recording a sample. This baseline would then inform the analysis of subsequent recordings, ensuring personalized and accurate feedback tailored to each user's natural speech patterns.
I would also like to conduct longitudinal testing to assess whether Polly effectively supports and sustains the desired writing behaviors over time.
my resume ---- my LinkedIn ---- my e-mail