Carney Curve

Not affiliated with Boston College

Carney Curve is an independent, unofficial project built by students. It is not affiliated with, endorsed by, sponsored by, or operated by Boston College, and it uses none of the university's names, logos, or trademarks to suggest otherwise. Nothing here is an official university publication.

Where the numbers come from

The evaluation scores on this site come from Boston College's published course evaluations — the same official questionnaire students fill in at the end of each semester. Each record is one section in one semester, and every average on the site weights each section record equally.

Where a question was not reported for a section, it is shown as blank rather than as a zero. A high average built on one or two sections is not strong evidence, so the number of evaluation records behind every score is always displayed next to it.

Student reviews, where they appear, are written by people who verified a @bc.edu address. They are opinions, not evaluation data, and they are shown separately for that reason.

Some professors also have ratings on RateMyProfessors. Where we have them, they are labelled as such, shown with the number of ratings behind them, and linked back to the page they came from. Anyone can post there — there is no @bc.edu check and no enrolment check — so they are never presented as evaluation data.

The three numbers we lead with

Every instructor page opens with the same three figures, because they are what people actually came for: an overall rating, how challenging the classes were, and how much work they took.

Two of the three are shown as a rank rather than a score, and that is deliberate. Boston College's questionnaire is bunched hard at the top: across every published record the median answer to almost every question is about 4.5 out of 5, which is 90 on the scale this site prints. A bare 90 therefore means “completely typical”, and a reader who takes it for “outstanding” — or, on a question about how demanding a course is, for “brutal” — has been misled by the number rather than informed by it. So we rank: “more challenging than 68% of BC classes” is a thing you can actually use. The raw score is never hidden; it sits in the breakdown beside the rank.

There is no “easiness” or “difficulty” field in Boston College's instrument, and we have not invented one. What we call Challenge is their own “intellectually challenging” question — and that question is one students answer positively about courses that stretched them. It measures intellectual challenge, not grading and not workload, and the site says so wherever it appears. RateMyProfessors' difficulty rating IS a genuine difficulty measure, but it is about a person across everything they teach rather than about any one class, so it is shown alongside as its own labelled figure and is never folded into the challenge number.

Workload is “workload” from the same questionnaire. Its field name in the underlying data reads like a count of hours and it is not one: every reported value is a whole number from 1 to 5, so it is one of five bands. Boston College does not publish what the bands mean in hours, so this site will not translate it into hours — an earlier version of this page did, and it was wrong.

Higher challenge means more challenging and higher workload means more work. Neither means worse. Nothing on this site tints either as good or bad, marks one instructor's workload as “best”, or ranks anybody on either — a demanding professor is a demanding professor, and what you want from that is your call.

Why the leaderboard is per class

A ranking of every instructor at the university would compare people who teach nothing in common, so the board that matters here is a class's: open a course and you get everyone who has taught it, ranked on their record in that course and nothing else. Someone strong across their other classes but weak in this one ranks on this one.

Everyone who taught the class is on its board, however thin their record, because hiding the person you are deciding about would be worse than showing their number with a caveat. Rows resting on fewer than 3 evaluation records are marked provisional, and the record count is printed in every row so you can see what a rank is standing on.

Department standings are a browsing aid rather than a comparison: they average a person across the different classes they teach, which is not like-for-like, so they at least require 4 records before anybody appears. An instructor with no rating at all is never ranked in either direction — “we have no reading” is not a zero.

How the score is worked out

Scores are shown out of 100. The underlying questionnaire is a 1-to-5 agreement scale, and a 5 is shown as 100, a 4 as 80, and so on — the same whole numbers the original site used. Workload is the exception: it is one of five bands rather than an agreement score, so it is shown as a rank against the rest of the university and never as a score or a number of hours.

The large number at the top of a professor's page is a blend of every source we have for them: BC's evaluations, student reviews posted here, and RateMyProfessors. Reviews somebody chose to write count for more per response than a form the whole class filled in at the end of term, so BC's evaluations are weighted the lowest of the three. That is our judgement, not a fact.

The blend also accounts for how many records sit behind each source, so a handful of reviews nudges the number rather than replacing it, and only earns real influence as more people post. Where only one source has anything to say, the headline is simply that source and no weighting is applied.

Because the weighting is a judgement you might not share, every page showing a blended number also lists each source with its own score and its own count, directly underneath. You can always see what moved the number.

Credit

This project is a rebuild of EagleEval, the BC course-evaluation site created and maintained by its original student authors until they graduated. Their work is released under the MIT License, and the structure of the data on this site follows theirs directly. Full credit to them for the original: github.com/AndrewBoessen/EagleEval.

Carney Curve is a working name, not a final one. It is neither a continuation of EagleEval nor endorsed by its authors.

Corrections and removals

Placeholder: contact route to be added before launch. If you are an instructor and something here is wrong, there will be a way to say so and have it looked at.