Adult Education Data Intern @ IRC
Internship Overview
- Position: Adult Education Data Intern
- Organization: International Rescue Committee, Spring 2025
- Location: IRC Elizabeth Office, NJ
- Supervisor: Vera Liang (vera.liang@rescue.org)
- Focus: Attendance data management, Power BI dashboards, system integration
- Skills: Power BI, Excel, Power Query, ETO, Client Track
Quick Overview
I worked as an Adult Education Data Intern at the International Rescue Committee’s Elizabeth, New Jersey office in spring 2025.
I maintained attendance and student-progress data, prepared records for upload into the program’s internal systems, and built Power BI dashboards covering fiscal years 2022 through 2025.
The Program
IRC’s Adult Education program provides English-language instruction to immigrants and refugees.
The team tracks attendance, assessments, course completion, and changes in English proficiency to understand participation and improve its programs. Much of this information was maintained across spreadsheets and internal case-management systems.
The data supported recurring operational and reporting needs, but inconsistencies between files made it difficult to prepare reliable records and compare results across classes and semesters.
The Data Problem
Attendance records came from different cohorts and instructors, which meant that spreadsheets did not always follow the same structure.
Common issues included missing entries, inconsistent formatting, duplicate or mismatched records, and differences between attendance files and the information stored in ETO and Client Track.
Before the data could be uploaded or analyzed, these discrepancies needed to be reviewed rather than automatically overwritten.
Attendance Data and Quality Control
I created structured Excel attendance trackers used for batch uploads into ETO.
The trackers calculated attendance rates, recorded absences, and included consistency checks across students and cohorts. I used Power Query to clean, normalize, and transform the source spreadsheets into an upload-ready format.
When records did not match, I traced the issue back to the source file and worked with team leads to determine which value should be treated as authoritative.
This process taught me that data cleaning is not simply a matter of making columns look consistent. A technically valid transformation can still be wrong if it changes the meaning of the underlying record.
Power BI Dashboards
I built Power BI dashboards covering Adult Education data from FY2022 through FY2025.
The dashboards included:
- Student demographics, including age, country of origin, gender, and educational background
- Attendance rates, absences, and course completion
- Changes in English-language proficiency
- Comparisons across semesters, classes, and course levels
I added filters and slicers so staff could explore specific cohorts without editing formulas or rebuilding charts.
The dashboards were designed for colleagues who understood the education programs but did not necessarily work with data tools every day. This influenced how I organized the pages, labeled the metrics, and explained the underlying calculations. Below are some examples:
- Biodata: gender, origin, age, educational background
- Attendance trends: rates, absences, class completion
- Literacy progress: ESL level transitions, course difficulty comparisons
Working With the Team
I participated in weekly team meetings and provided updates on attendance processing and batch uploads.
I also explained spreadsheet formulas, Power BI controls, and data-quality issues to non-technical colleagues. These conversations often revealed contextual information that was not visible in the datasets themselves.
Working directly with program staff helped me understand that data quality depended on communication between the people collecting, maintaining, and using the records.
Outcome
The structured attendance trackers made recurring batch uploads more consistent and easier to validate.
The Power BI dashboards gave the Adult Education team a centralized way to review participation, student demographics, and learning progress across multiple fiscal years.
The project also made discrepancies between spreadsheets and internal systems easier to identify before they affected reporting.
What I Learned
This internship gave me a different perspective on operational data.
At IRC, a missing or incorrect value could affect how a student’s attendance or progress was represented. That made validation more than a technical requirement.
I became more careful about tracing data back to its source, documenting assumptions, and asking program staff for context before deciding how a record should be corrected.
I also learned that a useful dashboard is not the one with the most charts. It is the one that helps its intended users answer recurring questions confidently.