1 of 7

Ray Opoku

Cover

Data Reduction with

Pandas

2 of 7

OUTCOME

SOLUTION

PROBLEM

One

3 of 7

1. WHERE DID THEY GO?

3. CAN WE FIND OUT?

2. WHY DON’T WE KNOW?

4. NSC STUDENT TRACKER

Two

For a little bit of context, Universities use information on their students to make critical decisions about enrollment, retention, grants, state appropriations, etc. so just like big tech, big education tracks its most valuable commodities - Students past, present and future. When a student leaves the college, it is in the interest of the college to know why they left and where they went. Just like a jealous boy friend or girl friend. They ask question such:

1. Why did my student leave?

2. What happens when UF Online students transfer out or stop-out (we like to call them stop outs not drop-outs? Here lies our problem?

3. Do some transfer to a 4-year university?

Do they pursue a graduate degree?

Do they enroll at another 2-year institution?

Did they enroll in a vocational program?

Did they not enroll anywhere else?

As big as a university is, you’ll think we should know right? But we don’t and here’s why

Students are not required to inform us

Some Immediately transfer when they meet requirements for their choice school

Some stop enrolling – Fall off the face of the each

Yes! we can find out. How do schools get this data?

Send out Post-graduation surveys

Send out surveys 1-year, 3-years, 5-years, and 10-years after

Send you mails

Stalk you on Facebook, twitter and LinkedIn

Sometimes we cold call you and to ask

Technically these methods don’t work well as you can imagine. The most efficient way we track our students is through NSC database. Higher education institutions that receive or accept Federal Financial Aid

98% of all students in public and private U.S. institutions. Provide enrollment information to Department of Education's National Student Loan Data System (NSLDS) and education finance industry

Meet the requirements for deferment or forbearance

Report that tells you students’ enrollment history

Matches students records by birthdate and name

The problem is how messy NSC data is.

4 of 7

WHAT WE WANT

WHAT WE RECIEVE

Three

5 of 7

Pandas To the Rescue

Options

Four

6 of 7

FINAL RESULTS

Five

ID	College	Term Start Date	Term End Date	College Sequence	Earliest Enrollment	Latest Enrollment
00001	Evergreen	9/1/2015	12/12/2015	1	9/1/2015	12/12/2016
	Evergreen	1/1/2016	5/25/2016	1	9/1/2015	12/12/2016
	Evergreen	9/1/2016	12/12/2016	1	9/1/2015	12/12/2016
00001	Santa Fe	1/25/2016	6/3/2016	2	1/25/2016	5/24/2018
	Santa Fe	9/5/2016	12/15/2017	2	1/25/2016	5/24/2018
	Santa Fe	1/24/2018	5/24/2018	2	1/25/2016	5/24/2018
000002	FIU	9/1/2015	12/12/2015	1	9/1/2015	5/25/2016
	FIU	1/1/2016	5/25/2016	1	9/1/2015	5/25/2016
000002	Embry Riddle	9/1/2016	12/12/2016	2	9/1/2016	12/12/2016
000002	Perdue	1/25/2016	6/3/2016	3	1/25/2016	12/15/2017
	Perdue	9/5/2016	12/15/2017	3	1/25/2016	12/15/2017

7 of 7

THANKS FOR YOUR TIME

CONNECT

TWITTER: @RAYOPK

GITHUB: RAOPK

RAYOPOKU.COM

End