Jack Smith: a case study in college selection

Applying to College
81

My post is quoted below. Note that the post in no way suggests that all employers meet the description. Instead it implies that there is not uniform behavior. Employers can and do emphasize different criteria in hiring, particularly at smaller companies. When I graduated from college, i applied to a company that seemed to strongly favor hiring grads who shared the same first name as the founder. The portion of employees who had that first name was far too high to be random. . However, this is not a typical hiring preference for new grads. Having variation between different employers does not mean that one cannot evaluate trends or what types of behaviors are most common.

You mentioned education and healthcare specifically. These were 2 of the industries that were including as groupings in the previously linked survey, which showed a similar type of pattern to the overall averages, as did all the other listed industry groupings.

However, the same is not true in most other fields. Employers often use GPA as basic screen, rather than focusing on pinnacle of highest GPA with 3.9x cum laude vs a more common 3.8. Once you cross the GPA screen, employers tend to focus on non-stat criteria like relevant experience, major, skill set, etc.

1 Like
82

The contrary is also true. Many don’t realize that most MIT grads will be average engineers. Most Ivy grads will have average careers. We tend to think in absolutes when the record is clear, there are lots of shades of gray.

7 Likes
83

Okay. Got it.

84

You’re exactly right. That’s, in fact, one of the big reason I do spend a lot of time counseling students/families to think beyond prestige. Being average or below average at MIT or any other prestigious school isn’t going to open special doors…and 50% of students at those schools are going to fall in that category by virtue of math.

6 Likes
85

If GPA/LSAT is supposedly predictive of success as a lawyer (or at least of success in law school) I wonder how predictive GPA is of success in other jobs. S’s smallish consulting employer is very data driven and uses college GPA as their primary screen for new grads because it is so correlated with previous successful hires. In view of the downturn they’ve just raised the cutoff (for an initial interview) from 3.7ish to 3.9! However when I talk to friends at FAANG companies, they say that their hiring practices are subject to close legal scrutiny and they are forbidden from using GPA cutoffs or any other hard and fast rules to disqualify applicants.

86

College GPA has a non-zero correlation with job performance, but studies usually find that correlation is small. The VP of People Operations at Google wrote the following. Note that a biased and limited sample contributes to Google’s experiences. They are not comparing a random 2.0 GPA kid to a random 4.0 GPA id. They are instead looking at the small minority of applicants who were hired and passed extensive additional considerations, including things like writing code during interviews, and comparing GPA within that limited group.

One of the things we’ve seen from all our data crunching is that G.P.A.’s are worthless as a criteria for hiring, and test scores are worthless — no correlation at all except for brand-new college grads, where there’s a slight correlation.

Suppose you work at Google and want to hire a software engineer for your team. Which of the following new grads sounds most promising, based on details below? I’d expect most persons in hiring positions would favor student A, in spite of having by far the lowest GPA. Student A had a successful internship at the company where the team got to meet him and work with him. The team knows him and knows he does well in a Google environment, similar to the job position. That’s more correlated with job success than average GPA during college, in isolation.

Student C has a by far the highest GPA, but I’d expect him to be least likely to be hired for the position due to lack of relevant experience – both in college and in past employment/internships. My point is there are many relevant and often more influential factors to consider in hiring decisions besides average college GPA or prestige of college attended for undergrad.

Student A – 3.5 GPA, Attended SUNY Oneonta, CS Major, Received key CS award at SUNY and has great recommendations, Excellent code portfolio, Has experience in desired coding languages, Received high ratings during internship at Google and is well liked by team, Personality/character fits well with Google’s mission, Does well on interview testing

Student B – 3.8 GPA, Attended Stanford, CS Major, Interned as SV company as SW engineer, Quality recommendations, Great coding portfolio and wrote successful app as hobby, Has CS experience, but does not have experience in all desired coding languages and desired skill sets, Does well on interview testing

Student C – 4.0 GPA (summa cum laude), Attended Yale, Economics major, Interned at an elite finance company, Took an intro CS class during freshman year of college, Does poorly on interview testing due to lack of programming experience

1 Like
87

But a job like software engineer with a specific “code writing” test is not at all representative of most hiring. And it’s all very well to say you hire based on a successful internship, but you still have to hire the interns in the first place. Hopefully that’s through a transparent process.

My question is how predictive higher GPA is of performance in other jobs without a specific pre-existing required skill. For example, the rationale at S’s company is that they really care about precision and attention to detail (spreadsheets without errors, presentations with consistent formatting, no spelling and grammar mistakes) and they believe the focus necessary to get almost all As in college is helpful in that regard. And note this is a prescreen before interview, so having a lower cutoff would mean either more interviews (and billable time wasted) or using some other aspect of the resume as a filter (eg directly relevant course work/job experience/college clubs, test scores, etc)

3 Likes
88

I think you’re right that it all depends on the kind of job/profession. An employer looking to fill a specific technical job that relies on certain skills may have one set of requirements (and/or their own specific test/technical interview), while another employer looking for more general cognitive ability may rely more on a different set of requirements (and/or more general tests/interviews).

89

A random cutoff is very common in engineering. They’re not typically stratospheric numbers though, as engineering GPAs tend to be lower than other majors. A 3.0 tends to be the “magic” number, but I’ve seen 3.5 and as high as 3.7. Certainly MANY engineers get jobs with lower GPAs.

The thing that I find interesting about that tactic, is that it assumes that every institution grades the same. They don’t, far from it in fact. You have the Stanfords and Browns where like Lake Woebegone, everyone is above average, the median being superstar level at most other schools, right at 3.70. Then you have schools like Purdue and especially HMC and Cal Poly, where the mean is roughly whole point below the grade inflating schools. (reported in various places as 2.5-2.7).

The year my son graduated from CP, another student, a Software Engineer, finished with a 4.0. It made all the local news because it had been a decade since the last student walked with a 4.0. Mudd is quite similar, once a decade or so for a 4.0.

In engineering it tends to be a automated screen because some positions get more applications than they have humans to read. Past that though, it becomes “show me what ya got” pretty quickly.

In my son’s specific case, he knew he wanted to work for an early startup of all industry veterans. They had no new grads, but his coursework, senior project, thesis, and work experience all aligned with their project, so they talked to him. He had several longish interviews, and still feared they wouldn’t take the risk on someone without an industry track record. Magna Cum Laude BS and MS With Distinction in hand, he pitched hiring him as an intern. Six weeks after his internship started, he was converted to full time.

He was recruited to his second job. Still it was a 4 interview process, the last of which was an 8 hour, sequential panel technical interview.

The TL;DR: GPA, as least for technical engineering, will get a foot in the door, but then you have to demonstrate skill to move up the chain. That level of skill has little correlation to undergraduate institution, but rather curiosity, horsepower, and drive.

It’s a little different for non-technical fields because it is harder to objectively determine a persons strength. That probably why law and medicine rely on metrics assumed to be accurate proxies for knowledge and skill.

1 Like
90

It’s different for law and medicine because they are admitting students to a largely academic classroom based program. Students who get high grades in undergrad school are generally also likely to get high grades in grad school. Grades are largely based on listening to lectures, doing well on exams that test the content of those lectures and little beyond that material, etc. GPA is often one metric students use to compare different grad schools and is one factor in USNWR ranking. Colleges do not want to tank that publicly reported measure. Colleges also have a limited number of available criteria to consider and may lack superior alternatives to grades.

Predicting success in employment is different. The kid with the highest average GPA during undergrad is not necessarily the one who will have the most automotive sales. Success in sales emphasizes different criteria from listing to academic lectures and doing well on exams. Similar statements could be said about most other professions. Average GPA in college doesn’t necessarily correlate well to success in teaching a classroom of children, working will with your team, figuring out a creative solution to a problem that was not explained in college classes. Even in jobs for which work tasks are well correlated with material covered in college, it’s often limited to a small subset of those college classes rather than average GPA. Software engineering performance is probably more correlated with grades in a few key CS classes than grades in English or history classes.

I am not saying that there is zero correlation between job success and college GPA. There have been many studies on this topic, and they consistently find a non-zero correlation. However, that correlation is usually small.

3 Likes
91

Software engineering was an arbitrary example. As noted in my previous 2 posts, studies usually find a small, but non-zero correlation. The exact value varies depending on specific job, what job performance criteria they are using, and sample group bias.

It sounds like your daughter works in consulting and you are interested in that field. I haven’t seen a study specifically on consulting, but I wouldn’t be surprised if consulting performance is more correlated with average college GPA than typical, with job requiring research of materials and presentation, and putting in the effort to work long hours to get project done when necessary . However, there are also a variety of other areas that are important for success in consulting that are not as well correlated with average GPA, such as having good people skills and working well with team as part of group projects. I’d expect a small correlation between consulting performance and average undergraduate GPA, notably more than zero, but also notably less than correlation with grad school grades.

This “small” correlation does not mean it’s a good hiring practice to choose the 3.9x grad over the 3.9y grad, but there could be some justification for using GPA as a filter to reduce number of applicants, before spending time and expenses on more detailed reviews of applicants and interviews. This is more true in fields for which there are less additional relevant criteria on resume to consider.

92

I wonder if it follows from this that companies with plenty of applicants should implement more rigorous GPA screens for those types of jobs where success is likely to be more correlated with GPA. But do they? What minimum GPA is expected for potential MBB consultants?

93

I agree. I asked my son what distinguished a 2.75 grad from a 3.75 grad. His answer “Work ethic. They’re all smart if they get that far.” The differential in drive and work ethic should correlate to job performance and may be the thing that drives the non-zero correlation between GPA and success.

1 Like
94

This. An elite education is fantastic if 1) you are accepted and 2) you can afford it without taking on insane debt (personally, I’d take on modest debt for some elite schools, but that is me). Unfortunately, because the number of great students far exceeds the number of spots at elite schools, a lot of kids are going to be disappointed - that disappointment shouldn’t be amplified by letting them believe that they won’t enjoy success just because they couldn’t get into Harvard.

6 Likes
95

There is plenty of anecdotal evidence though that “elite” may not correlate to the quality of education.

For fear of a flame war, I won’t name names, but plenty of “elite” publics have massive class sizes. One many might consider the most elite public is so inundated with bureaucratic red tape that a a student told me it was like going to school at the DMV. There was major controversy at a bluest of blue blood Ivy institutions, because undergrads were leading recitations.

I think the network and prestige associated with said institutions have power, but as one should do with any school, understanding what the experience will be like will there, determining “fit” if you will, is still paramount.

3 Likes
96

Agree 100%

1 Like
97

This is brilliant! Perfectly describes the large state school I went to. I used to tell myself it prepared me well for real life, for dealing with, well, the DMV, and the IRS, and any company’s circular recorded phone menu that will get you everywhere but where you need to go.

That said, I did get a good education. But it’s not for the faint of heart.

2 Likes
98

Not to mention thise who take 5 or 6 years to graduate because they can’t get into the courses they need.

1 Like
99

I think taking 5-6 years is not the problem of the schools, even big universities, so much as students failing to organize their pre-reqs or changing majors too often, or leaving their core/required classes too late.

Most kids I know who went to big universities graduated on time or even early.

4 Likes
100

I think this argument went a bit too far. An average, or slightly below average, student at MIT (or a few similar institutions) does enjoy some advantages:

  • They’re more likely to be surrounded by peers who can challenge and push them to do better

  • They’re more likely to be able to interact with distinguished professors who can impart special knowledge and greater insights in certain subjects

  • They may be able to do meaningful research starting from their freshman years

  • They may be more likely to have made friends or acquaintances of people who may be in position later to help their careers

  • They may be able to impress some people later in their careers (often in different professions) with just the names of their colleges :stuck_out_tongue_winking_eye:

On the other hand, being in the bottom quartile or quintile in any school isn’t fun (and may even be mentally stressful at a couple of these schools).

1 Like