More than a Dashboard
A Curriculum design research project that concretely shapes
the future of HCI education @CMU
Product
Data Cleaning, Analysis & Visualization
Company
Human Computer Interaction Insititute
My Role
Design Engineer
Overview
This project is more than a dashboard.
It showcases two core strengths that underpin my approach to design:

Data Cleaning and Data Analysis Expert
I transformed messy, poorly-structured survey data into a well-organized, interactive dashboard, enabling clear insights and actionable results.

User-Centric Design Solutions
I focus on understanding stakeholder needs and feedback to design thoughtful dashboard features, such as filters and correlation tools, that enhance functionality and usability.
Goal
Effectiveness? Innovation?
How is CMU education doing?
HCII needs to know how well their education is helping their students achieving their post graduation role.

In order to obtains the most updated design trend and market requirement, my supervisor conduct extensive research involving multiple stakeholders such as students, alumni, hiring manager, and other professionals.  
Is a Master in HCI worth it right now?
Honestly just looking for insight from those in the industry or in a program about how much they think the value of the degree has held up and/or helped them career-wise...
The Impact of Artificial Intelligence on Product Design and the Future of UI/UX
The Human-Computer Interaction Institute offers multidisciplinary undergraduate and graduate educational programs that emphasize technology for the benefit of people and society.
Context
Turn data into actions
Previous they've designed and collected a bunch of data, both qualitative and quantitative, from current students, alumni, industry partners, hiring managers, and other affiliated professions.

Now they need someone to turn those data into actions
My Deliverable
More than a presentation,
More than a dashboard
Blablablablabla
My Process
Sanitize, Analyze, Visualize
Sanitize: Make it Clean
The survey structure is a mess on qualtrics which causes a lot of columns in the output csv file to have repeated naming, which complicates the data entry process.
The data structure is complicated and hard to understand at glance, I created a visual question key mapping out the survey flow.
In that case, anyone with the both the key mapping and survey csv file would be able to find which column in the csv file correspond to which question answered by which group.
Analyze: What pops?
I did significance test to see if there are actual difference of the ratings across different groups.
In other words, on a skill of 0 to 5, if all currents students rate "animation" as 2.4 while hiring managers give a 2.7, could we treat them as same number in one category, or they are actually different numbers
Visualize: Make it Clear
For the dashboard, the three main principles that I need to follow includes:
1. Comprehensiveness 
2.
3. Empowering users
I really learned
Create a cleaner survey saves a lot of time wasted.

I did not involve in the beginning process when survey was created, which complicates my whole analysis process and delays final deliverable. So if I do user/market research in the future where a survey is needed, I'll make sure that I tie the survey structure closely to our research objectives and format the survey in a way that facilitates data analysis.