Fall Career & Internship Fair — Analysis README

Project Summary

Analysis of the University of Pittsburgh Fall Career & Internship Fair (Arts, Sciences, Government & Non-Profit session, September 18, 2025) for a stakeholder presentation to Career Center leadership and the Division of Student Affairs data team.

---

Data Sources

File	Records	Description
FCF_Student_Attendees.xlsx	1,288	Student registration and check-in records
FCF_Student_Survey_Responses.xlsx	2,493 (273 fair-specific)	Post-event survey across all Career Center events
FCF_Employer_Registration.xlsx	182 (68 at A&S session)	Employer booth registrations across all fair days
FCF_Employer_Survey_Responses.xlsx	33 (31 usable)	Post-event employer survey. Row 1 is header text, "test" entry excluded

---

R Script: data-validation.Rmd

Purpose

Independently reproduces every statistic used in the presentation and Tableau dashboard from the raw Excel files.

Requirements

• R with packages: readxl, dplyr, stringr
• All four Excel files in the working directory

What It Does

• Loads all four raw datasets
• Computes attendance metrics (attended, registered, walk-ins, show rate)
• Breaks down attendees by class year and major (using CONTAINS matching for double majors)
• Filters student survey to fair-specific responses and computes NPS, satisfaction, recruiter alignment, marketing channels, and next engagement
• Removes header row from employer survey and computes NPS, overall ratings, return likelihood, interview plans, and student competency ratings
• Computes international student access metrics from employer registration
• Prints all results organized by presentation slide number for easy comparison

Key Filtering Notes

• Student survey is filtered to Event = "Fall Career & Internship Fair - September 16-18"
• Employer registration A&S filter uses Assigned Booth Options containing "(A&S)" per data provider guidance
• Employer survey excludes row 1 (question header text) and the "test" entry
• Major grouping uses CONTAINS logic — a student majoring in "Psychology, Spanish" counts under Psychology

---

Tableau Dashboard: Student Experience

Data Source Setup

Four separate data sources loaded independently — not joined or related:

1. FCF_Student_Attendees.xlsx
2. FCF_Student_Survey_Responses.xlsx
3. FCF_Employer_Registration.xlsx
4. FCF_Employer_Survey_Responses.xlsx

Column Renaming

Student survey columns renamed with S_ prefix. Employer survey columns renamed with E_ prefix. Key renames documented in PROJECT_DOCUMENTATION.md.

Calculated Fields Created

Student Attendees:

• Attended_Flag — 1 if checked in, else 0
• Registered_Flag — 1 if registered, else 0
• Walk_In — attended but not registered
• Year_Group — collapses Masters/Doctorate/MBA into "Graduate"
• Major_Group — groups top 10 majors using CONTAINS logic, remainder as "Other"
• College_Group — Dietrich School vs. Other Colleges
• Attendance_Category — Registered & Attended / Walk-In / Registered No-Show

Student Survey:

• S_Is_Fall_Fair — 1 if Fall Career Fair, else 0
• S_NPS_Calc — computes NPS score from promoters and detractors
• S_Learned_Pct — percentage who learned something valuable
• S_Engage_Again_Pct — percentage who plan to engage with Career Center again
• S_How_Heard_Grouped — groups low-frequency awareness channels into "Other"
• S_Next_Engage_Short — shortened labels for next engagement plans

Employer Survey:

• E_Interview_Pct — percentage of employers planning to interview students

Filters Applied

• Student Attendees charts: Attended_Flag = 1
• Student Survey charts: S_Is_Fall_Fair = 1, nulls excluded from relevant fields
• Employer Survey: header row excluded via Q1_1, "test" entry excluded

Tooltips

• Bar charts include count and percent of total on hover
• "Other" bar in Awareness Method lists the grouped channels
• Treemap uses viz-in-tooltip: hovering a major shows a mini bar chart of class year breakdown for that major

Known Caveats

1. Student attendees file and student survey both cover all three fair days (Sep 16-18), not exclusively the A&S session. There is no field to distinguish which day a student attended.
2. Employer A&S filter uses Column I (Assigned Booth Options) containing "(A&S)" — confirmed by data provider email.
3. Employer survey sample is small (n=31 after removing header and test entry). Results are directional.
4. No shared identifier between student registration and student survey — individual-level funnel analysis is not possible.
5. Major grouping uses CONTAINS matching. Double majors counted once under their first matching group in the priority order.
6. International student metrics (work auth, OPT/CPT, sponsorship) are calculated across all 182 employers, not filtered to A&S only.