Data & Analysis

We ranked universities based on students’ coding skills—and the results aren’t what you might expect.

2021 CodeSignal University Ranking Report

01.

Introduction

CodeSignal’s mission is to help companies go beyond resumes and pedigrees to identify and hire candidates who demonstrate real on-the-job skills. One area where companies tend to rely on pedigrees the most is in university recruiting. Unfortunately, this ends up reinforcing inequality—the students who go to the “best” schools are the ones who get access to the prestigious internships and new grad jobs that lead to even more career opportunities later on.

However, the talent landscape is changing. Competition is hotter than ever for technical talent, and with remote work, companies are able to consider more diverse candidates based all over the US. Furthermore, two thirds of engineers are self-taught and the traditional educational institutions are no longer the sole source of technical talent. But our recruiting practices haven’t caught up. Everyone is still targeting the same so-called “top engineering schools”—when what’s really needed is information that helps recruiters find the students with the most demonstrated and proven skills, regardless of where they went to school (or whether they have a college degree at all for that matter).

We developed the CodeSignal University Ranking Report to make early talent technical skills signals more widely available to companies gearing up for the upcoming university recruiting season. This annual report ranks universities based on their students’ performance on CodeSignal’s General Coding Assessment (GCA), a standardized skills assessment measuring the fundamental computer science and programming ability with interactive code writing tasks.

02.

Our Top 10 Universities

We analyzed a subset of CodeSignal’s General Coding Assessment results data and ranked universities based on the percentage of their students whose GCA score is in the 84th percentile on the General Coding Assessment. 
Here is our top 10.

Rank
University
#1
Massachusetts Institute of Technology
#2
Miami University
#3
Carnegie Mellon University
#4
State University of NY at Stony Brook
#5
Columbia University
#6
University of Chicago
#7
Rice University
#8
University of Oklahoma
#9
Drexel University
#10
California State University, Sacramento

Highlighted universities don’t appear on the US News & World 
Report’s top 10 Best Undergraduate Engineering Programs.

03.

Key Highlights

Only 2 out of the 10 schools (MIT and Carnegie Mellon) from our list were in the top 10 US News & World Report Best Undergraduate Engineering Program Rankings.

Miami University and California State University, Sacramento aren’t included at all in the US News & World Report’s ranking (which has over 200 schools).

SUNY is ranked #65 by US News & World Report.

Drexel is ranked #94.

UC Berkeley and Stanford, tied for #2 in the US News & World Report ranking, don’t make our top 10.

04.

Conclusions

We hope highlighting the discrepancy between the traditional engineering school ranking and the ranking based on students’ coding skills will help recruiters and hiring managers question the efficacy of using pedigree as proxy for skills and consider directly utilizing objective skills assessment in their hiring process.

In the future we hope to gather more in-depth educational background information to include skills data of technical talent that don’t have a traditional computer science college degree. We also hope to append demographic info such as gender, race, and socio economic background to understand the GCA performance data in various dimensions beyond schools.

05.

Additional Information

What is General Coding Assessment (GCA)?

General Coding Assessment is a standardized test for core programming and computer science knowledge that is shared by most software engineers and is taught in most computer science undergraduate programs in the US.

GCA consists of 4 code writing tasks in the cloud integrated development environment and is language agnostic. This means the task takers can choose a computer language that they feel most comfortable in. GCA measures the task taker’s code-writing skills, problem-solving skills, and ability to write clean code at a reasonable speed.

What is the methodology?

CodeSignal’s GCA is used by over 150 top tier companies such as Facebook, Asana, Uber, and eBay, as their primary technical assessment for early talent. GCA has been completed by over 95,000 students who are seeking new grad software engineering opportunities in their senior year or summer internship software engineering opportunities in their junior year. More than 50% of undergraduate students completing their CS programs in the US take the GCA, with over 9,000 students taking the test per week at the height of the recruiting season.

For this analysis, we’ve inferred the name of educational institutions from the email domains associated with the GCA takers. To support valid comparisons, we’ve adopted U.S. News & World Report’s grouping of colleges and used the “National University” group to come up with CodeSignal University Rankings.

What does a top score (over 800) mean?

Similar to an SAT, the GCA has many variations of questions but produces comparable test scores. The GCA score ranges between 600-849.

SCORE RANGE
Percentile
Description
600-649
0th-6th
The test taker can write simple code to do some operations.
650-699
6th-25th
The test taker is familiar with programming and can work with some common data structures.
700-749
25th-59th
The test taker has solid implementation skills and can solve some algorithmic tasks.
750-799
59th-84th
The test taker has good problem-solving skills, is familiar with algorithms, and can implement ideas that don’t require too much coding.
800-849
84th-100th
The test taker has excellent algorithmic and problem-solving skills and can develop large, complex applications.
Designed by our in-house Software Engineers, Computer Science experts, and PhDs in I/O Psychology, the GCA measures test taker’s computer science fundamentals on topics like algorithms and data structures. Those who demonstrate a great CS foundation and strong ability to implement scalable applications receive a GCA score in the highest range: 800-849 or 84th percentile.
We use cookies to improve the interaction with our website. By continuing to use this site, you are giving us your consent to use cookies. Learn more