Engineering school failure rates (esp. UW engineering)

[1138]

I was reading that thread in the SW vs. ST forum where the guy was bluffing about his education. I remember Mike mentioning that a lot of people failed out of UW engineering, with more than half failing first year. I graduated from UW engineering last year and I don't think we lost more than 10 people in a class of 80 and of those, several just got delayed to the next graduating year. And I don't remember a lot of people failing in first year.

I'm curious about this. Has the program gotten easier or have they relaxed the promotion rules? I remember that the dean made a big deal during frosh week or some other orientation session about reassuring us that we wouldn't fail out easily. I know we have special rules for 1A students which allows them to be promoted to 1B even with a failed course or two.

What's the situation like at other schools? I've heard the situation at U of T's Engineering Science program and they really do try to weed the students out.

[Stark]

In general this wouldn't surprise me. Talking to people who did my course in the early 90s, it was far more complex and indepth. When I was studying (in the late 90s) the practice of 'weeder years', where they gave you terrible lecturers and intentionally steep learning curves to get rid of the ambivalent students, was still in force.

Maybe Australia is just behind the times.

[Darth Wong]

I was reading that thread in the SW vs. ST forum where the guy was bluffing about his education. I remember Mike mentioning that a lot of people failed out of UW engineering, with more than half failing first year. I graduated from UW engineering last year and I don't think we lost more than 10 people in a class of 80 and of those, several just got delayed to the next graduating year. And I don't remember a lot of people failing in first year.

Don't tell me they're neutering my beloved Waterloo engineering program. The mass slaughter was half the fun: at the beginning of every term, we would look around the class and try to count off the people that we figured wouldn't make it. Maybe I should make a call to Waterloo and try to confirm this rather dramatic change in policy. They had a reputation for 30 years; it seems strange that they would suddenly change, unless you're talking about one of the easier faculties, like geo eng.

[aerius]

As of 2002 when my friend graduated from UW, the weeding out program for Waterloo engineering students was still in full effect. In her first year they loaded them up with 35 hours a week of lectures, plus labs, and wiped out nearly a third of the class in the first semester alone. By the time they graduated less than a quarter of the class was left.

[Darth Wong]

As of 2002 when my friend graduated from UW, the weeding out program for Waterloo engineering students was still in full effect. In her first year they loaded them up with 35 hours a week of lectures, plus labs, and wiped out nearly a third of the class in the first semester alone. By the time they graduated less than a quarter of the class was left.

I'm pretty skeptical about UW changing from a ruthless weeding-out school (one whose reputation was built over decades) into a damned tea party. I'll have to contact the faculty to verify this claim before taking it seriously.

[1138]

As of 2002 when my friend graduated from UW, the weeding out program for Waterloo engineering students was still in full effect. In her first year they loaded them up with 35 hours a week of lectures, plus labs, and wiped out nearly a third of the class in the first semester alone. By the time they graduated less than a quarter of the class was left.

35 hours of lectures? In first year, we had 6 courses, so 18 hours of lectures a week (3 hours each). What program was that? I don't think the 1st year Mechs (who had the most lectures in first year) had that many in 2002. We were told that we should put in 60 to 70 hours of work per week in total (incl. lectures, labs and homework) to be successful.

I'm pretty skeptical about UW changing from a ruthless weeding-out school (one whose reputation was built over decades) into a damned tea party. I'll have to contact the faculty to verify this claim before taking it seriously.

It could be department differences. I was in Systems. But I didn't hear anything about the ruthless weeding out from friends in ECE or ME. People complained about failing courses but not terms.

Were the majority of failures in first year? If so, it may be that you didn't have the same promotion rules for first term. In 1A, they're allowed to fail 2 courses and still proceed to 1B. I don't know when those rules came into effect but the undergraduate calendar had special rules for 1A starting in 1997 - 1998.

[aerius]

35 hours of lectures? In first year, we had 6 courses, so 18 hours of lectures a week (3 hours each). What program was that? I don't think the 1st year Mechs (who had the most lectures in first year) had that many in 2002.

Chemical engineering, she started in '97 and completed the 5 year co-op program. In 1st year most of the engineering programs had common courses, and all of them as far as she knows were loaded up with at least 30 hours/week of lectures, plus labs & tutorials. Quite a few people were toast by Thanksgiving, and had dropped out or transferred to easier programs.

[1138]

Chemical engineering, she started in '97 and completed the 5 year co-op program. In 1st year most of the engineering programs had common courses, and all of them as far as she knows were loaded up with at least 30 hours/week of lectures, plus labs & tutorials. Quite a few people were toast by Thanksgiving, and had dropped out or transferred to easier programs.

I was in one of the programs that didn't have the common set of courses but I don't think my ECE or ME friends had as many hours of actual lectures when we did first year in 2002. I certainly did not. I should ask some of my profs what their schedules were like back in the 90s. It really seems like they changed things around.

I'd really hate it if it ends up being that my program was easier than 10 years ago. I went to UW because people said it was hard (I turned down UofT's eng sci because they didn't have co-op).

[The Spartan]

I don't know the exact statistics or what they are currently (I've heard they're trying to be even more stringent) but for what it's worth, when I got to my Senior Design Project at UT/Arlington our professor told us to be proud for making it that far because, statistically, only a third of freshman in the Mechanical and Aerospace Engineering Department graduated with one of those degrees. The rest switch majors or drop out.

For us the two hurdles were freshman year and then junior year. Freshman year weeded out anyone who shouldn't really be in an engineering curriculum, period and junior year weeded out those who just didn't quite have what it took.

[The Dude]

35 hours of lectures? In first year, we had 6 courses, so 18 hours of lectures a week (3 hours each). What program was that? I don't think the 1st year Mechs (who had the most lectures in first year) had that many in 2002.

Chemical engineering, she started in '97 and completed the 5 year co-op program. In 1st year most of the engineering programs had common courses, and all of them as far as she knows were loaded up with at least 30 hours/week of lectures, plus labs & tutorials. Quite a few people were toast by Thanksgiving, and had dropped out or transferred to easier programs.

30 hours per week would have to be including labs and tutorials. A full-time course is 3 hours/week of lecture. Even in engineering, it is rare to take more than 7 full-time equivalents per semester and 10 is unthinkable. Some engineering schools did trim their curricula in the mid to late 90s (including Waterloo and UBC), but not by that much.

To the OP's point, the rate of fail-outs in engineering schools has been vastly exaggerated. Waterloo engineering has had a graduation rate of over 80% since at least 1990 (compare this to Ryerson, which is in the 50's %). In fact, the first-tier engineering schools consistently have significantly higher graduation rates than lesser schools, largely because they admit smarter students. Further, engineering programs are answerable to the CEAB and to their respective provincial ministries for their graduation performance; low graduation rates are not taken as a positive sign.

Some references on the graduation rates:
http://www.analysis.uwaterloo.ca/docs/pi/kpi1999.html
http://www.edu.gov.on.ca/eng/general/po ... grgrad.pdf

[Mayabird]

Dunno about UW, but I remember the graduation rate at Georgia Tech being somewhere between 66-75%, but that's including all majors, not just engineering. I don't know the individual make-up for all schools (I suspect management's graduation rate was significantly higher than the average), but for aerospace engineering, which I started in, about 50% of the people who came in eventually graduated with a BS in AE. The other half flunked out, switched schools, changed majors, or something like that. Granted, less than half of the people who got the degree ended up actually getting a job in aerospace, but there you go.

But a lot of the general education courses like Physics II and the more advanced chemistries really were weed-out courses. I remember one physics test where, of the thousand or so of us who were taking the course (different professors, different times, same material) only 26 got the correct answer on a question. No, I was not one of them. And I knew of one chemistry professor who was in trouble because too many people passed his classes. I also had a math TA who had taught the class the previous semester, but again, too many people passed, so he got demoted back to being a TA.

[Adrian Laguna]

And I knew of one chemistry professor who was in trouble because too many people passed his classes. I also had a math TA who had taught the class the previous semester, but again, too many people passed, so he got demoted back to being a TA.

Where they in trouble specifically because too many people passed the class? Or did the high passing rate lead to an investigation of the standards being used to assign grades, and said standards where found wanting?

[phongn]

Hell, even at my school, I saw plenty of people in the initial two CS weeder courses (known as the GATEs) that didn't appear at the next sequence, and from there more people dropped out.

The bio/pre-professional degrees are also pretty draconian. I recall the average test grades in my sister's Anatomy and Physiology course being around 40% and graded on a ten-point scale.

[Darth Wong]

First-year failure rates were the biggest by far. We started with 160 students, and had 80 at the end of first year. It seems a bit strange that a class would start with 80, since I don't think the faculty has shrunk or lost popularity since I was there. After that, the failure rate was pretty steady: we would lose roughly 10-15 guys each term, but of those, quite a few would just drop back a term, so they wouldn't be entirely out of the program (you could only do that once, though).

[General Trelane (Retired)]

First-year failure rates were the biggest by far. We started with 160 students, and had 80 at the end of first year. It seems a bit strange that a class would start with 80, since I don't think the faculty has shrunk or lost popularity since I was there. After that, the failure rate was pretty steady: we would lose roughly 10-15 guys each term, but of those, quite a few would just drop back a term, so they wouldn't be entirely out of the program (you could only do that once, though).

That pretty much agrees with my memories of engineering at the U of Alberta in the mid-to-late '80s. They even expected about half the first year class to drop out. In the very first common class, I recall the prof saying something like, "Look at the person on your left. Now look at the person on your right. Chances are that next year, one of them won't be here anymore."

[Darth Wong]

Well let's face it, there are a lot of people who show up for first year and are really not cut out for it. And you can't identify these people by simply looking at high-school grades. The only way to change this policy would be to either make the program vastly easier (something I seriously doubt has happened) or to have some kind of near-psychic method of determining which students are cut out for the job before the first year begins.

Unless, of course, 1138 has simply forgotten how many students showed up at the beginning of first year. Class sizes tend to settle down after the first wave of flunkings.

[General Trelane (Retired)]

Unless, of course, 1138 has simply forgotten how many students showed up at the beginning of first year.

That's very likely. I have to admit that I can't remember how many students were in my first year. There was one common class (an orientation to the engineering profession, held in the Tory Turtle, that was a simple pass/fail--if you showed up, you passed), but for all 'real' classes, we were split into groups of 40-50, so it's hard to recall the total number (200 comes to mind, but I can't think of why).

In 1138's case, he admits that his program wasn't part of the common engineering class at UW and that he gets his ancedote from his friends. That's not particularly reliable.

Personally, I'd find it surprising that such a policy has changed.

[1138]

First-year failure rates were the biggest by far. We started with 160 students, and had 80 at the end of first year. It seems a bit strange that a class would start with 80, since I don't think the faculty has shrunk or lost popularity since I was there. After that, the failure rate was pretty steady: we would lose roughly 10-15 guys each term, but of those, quite a few would just drop back a term, so they wouldn't be entirely out of the program (you could only do that once, though).

The class sizes are limited by the number of seats in the class room. Systems Design only has room for I think 92 people (and that's a hard limit, there's literally only space for that many chairs), so that's why the classes are so small. 80 is just a rough figure; it may have been 85 people. In my year there were two comp. eng classes that started at 125 each, I believe.

I asked around (grad students who graduated a few years before me) today at the school and here's the story I got:

For ECE students, about 1/3 of the class washes out in first year and then a few get lost per term. So it's possible that half of the students remained by the end. Technically, ECE doesn't either but they do it anyway.

Maybe I'll ask the undergrad chair if I see him.

In 1138's case, he admits that his program wasn't part of the common engineering class at UW and that he gets his ancedote from his friends. That's not particularly reliable.

Yeah, I wonder why my friends never mentioned that to me before. I discussed the UofT situation with them at length but they never said UW does it too. I always assumed it wasn't any different. My program did share the Calculus and D.E. courses with the ECE students and there weren't a lot of failures from my class for those (I think 25% failed the Calc 1 midterm but they got their act together). We may just have had the stronger students or those are precisely the courses that didn't try to weed people out.

[Elessar]

I was in one of the programs that didn't have the common set of courses but I don't think my ECE or ME friends had as many hours of actual lectures when we did first year in 2002. I certainly did not. I should ask some of my profs what their schedules were like back in the 90s. It really seems like they changed things around.

I'd really hate it if it ends up being that my program was easier than 10 years ago. I went to UW because people said it was hard (I turned down UofT's eng sci because they didn't have co-op).

Greetings. I am Class of 2007, Computer Engineering (putting the C in ECE). That would mean my first year was Fall 2002.

In 1A and 1B, there were easily 40+ hours of lectures and labs, in addition to required tutorials. This did not drop down significantly until 3A, and while the class size had diminished only to 3/4 of the original, we had plenty of transfers in and out. This was after they implemented the 1A 1-failure forgiveness.

I do believe that the average quality of students may have risen, because certainly course-load and curriculum has not taken any noticeable dive.

And don't worry about those EngSci smucks. PEY is awful compared to Engineering Co-op, just take a look at all the employers flying into YYZ just to interview UW students.

[1138]

Greetings. I am Class of 2007, Computer Engineering (putting the C in ECE). That would mean my first year was Fall 2002.

Yep, that would be the same year as me.

In 1A and 1B, there were easily 40+ hours of lectures and labs, in addition to required tutorials. This did not drop down significantly until 3A, and while the class size had diminished only to 3/4 of the original, we had plenty of transfers in and out. This was after they implemented the 1A 1-failure forgiveness.

I'm confused about that 40 hour figure. I thought you had 6 courses (Chem, Calc, Alg, Phys, Intro to ECE and Programming), only two of which had labs (Intro to ECE and Programming). 6 courses * 3 hours / week = 18 hours per week. Add 3 hours of labs for each lab course and I get only 24 hours of lectures and labs, which, coincidentally, is the same number of hours as my schedule. It would be 25 hours, because the "tutorial" for the ECE (and Systems) version of Calc that year was actually a lab where we had to work on assignments for 1 hour and then hand it in. Am I missing any classes or labs?

I do believe that the average quality of students may have risen, because certainly course-load and curriculum has not taken any noticeable dive.

I think I'm convinced that the curriculum hasn't gotten easier, especially since it all has to be accredited.

And don't worry about those EngSci smucks. PEY is awful compared to Engineering Co-op, just take a look at all the employers flying into YYZ just to interview UW students.

Co-op has been a very positive experience for me, except that applying to jobs took me a lot of time.

[Darth Wong]

So ... the workload hasn't changed too much, but they've shrunk class sizes and are flunking fewer first-year students? Do they have some improved method of determining which students are likely to wash out before they sign up?

PS. Actually, I did some quickie research. Interestingly enough, it appears that the "doting parent" phenomenon in the current generation has been strongly correlated to a vastly improved work ethic among students, in direct contrast to the widespread opinion that these kids who grew up with highly structured lives and parental oversight would be incapable of functioning on their own (I wonder if this will silence the usual suspects who love to rant about the benefits of the "unstructured play time" they had when they were kids). Far less rowdy partying goes on, provided one can successfully identify this type of student in the admissions process (which is actually not difficult, since a questionnaire can easily target kids who were active in large numbers of scheduled extracurricular activities). This has some interesting ramifications for the widespread ranting about how the latest generation of kids is going down the tubes. It may be true in some respects, but apparently not for this subset. Mind you, I don't know how well that applies to schools like MIT, which are notorious for their wild parties. UW was always a notoriously strait-laced "geek school" rather than a "party school", but after having spoken to a few people, it's become even more strait-laced in the 15 years since I left.

[Terralthra]

Interestingly, the school where I first went to college (SJSU) had the same sort of rep UW did for the CompE program I went into. Larger enrollment, but same sort of washout rate. At least 1 out of 2 students who entered as freshmen were gone by sophomore year. The curriculum was excellent: most of the faculty were part-time professors, part-time engineers at Intel, AMD, nVidia, etc.

When the dot com bubble burst, I left school because the company that had been giving me a scholarship went belly-up, leaving me with quite the bag of debt to hold. By the time I'd found gainful employment with half of a CompEng degree in the IT industry, the SJSU CompE department had been thoroughly raped. Most of the former professors couldn't afford to teach part-time anymore. The professors they'd brought in to replace them were mostly the ones that had been getting pink slips and could not find new jobs, ie, the mediocre ones. In order to maintain a stable budget with better enrollment, the rigor and the washout rate were both decreased, with a predictable effect on the quality of education and reputation the program had.

Thus, my switch to a different school and a different field: linguistics.

[Elessar]

I'm confused about that 40 hour figure. I thought you had 6 courses (Chem, Calc, Alg, Phys, Intro to ECE and Programming), only two of which had labs (Intro to ECE and Programming). 6 courses * 3 hours / week = 18 hours per week. Add 3 hours of labs for each lab course and I get only 24 hours of lectures and labs, which, coincidentally, is the same number of hours as my schedule. It would be 25 hours, because the "tutorial" for the ECE (and Systems) version of Calc that year was actually a lab where we had to work on assignments for 1 hour and then hand it in. Am I missing any classes or labs?

I took a look at the undergraduate calendar and it appears you are correct. In terms of required hours @ UW, it was rarely more than 30 hours... of course, this was before all the work necessary to simply keep up with the weekly assignments. I swear I remember waking for 8:30 tutorials and being at the school until 4:30, which was how I derived my original 40 hour estimate.

I guess I included 1 hour lunch.

UW was always a notoriously strait-laced "geek school" rather than a "party school", but after having spoken to a few people, it's become even more strait-laced in the 15 years since I left.

That wouldn't surprise me. Demographics have shifted considerably during that time. Think '97 Handover.

Edit: Didn't notice this the first time around.

So ... the workload hasn't changed too much, but they've shrunk class sizes and are flunking fewer first-year students? Do they have some improved method of determining which students are likely to wash out before they sign up?

Class sizes aren't intentionally smaller; CE 2007 was initially 200-strong. Our 'class' was actually two distinct sections that didn't intermingle much until 2nd year.

However, I've heard that UW has an 'adjustment' chart that they apply to applicant OAC averages based on the High School they came from. I don't recall the source though, so it's probably just a rumour.

[Spin Echo]

So ... the workload hasn't changed too much, but they've shrunk class sizes and are flunking fewer first-year students? Do they have some improved method of determining which students are likely to wash out before they sign up?

PS. Actually, I did some quickie research. Interestingly enough, it appears that the "doting parent" phenomenon in the current generation has been strongly correlated to a vastly improved work ethic among students, in direct contrast to the widespread opinion that these kids who grew up with highly structured lives and parental oversight would be incapable of functioning on their own (I wonder if this will silence the usual suspects who love to rant about the benefits of the "unstructured play time" they had when they were kids). Far less rowdy partying goes on, provided one can successfully identify this type of student in the admissions process (which is actually not difficult, since a questionnaire can easily target kids who were active in large numbers of scheduled extracurricular activities). This has some interesting ramifications for the widespread ranting about how the latest generation of kids is going down the tubes.

What research did you look at to determine this? There had been some studies a while ago that were showing the opposite. The kids whose parents filled all their free time with French lessons and Soccer lessons and Oboe lessons and so forth were ending up burnt out as opposed to well rounded. While participating in some activities is obviously good for children, unstructured play time has been shown to be beneficial for the children's development and mental well being. (Perhaps you're hyperbolising and I'm just missing it)

It may be true in some respects, but apparently not for this subset. Mind you, I don't know how well that applies to schools like MIT, which are notorious for their wild parties.

We are? Occasionally the Frats would get up to something stupid, but on the whole people seemed too busy to get off their arses and organise that sort of thing.

[Ericxihn]

So ... the workload hasn't changed too much, but they've shrunk class sizes and are flunking fewer first-year students? Do they have some improved method of determining which students are likely to wash out before they sign up?

PS. Actually, I did some quickie research. Interestingly enough, it appears that the "doting parent" phenomenon in the current generation has been strongly correlated to a vastly improved work ethic among students, in direct contrast to the widespread opinion that these kids who grew up with highly structured lives and parental oversight would be incapable of functioning on their own (I wonder if this will silence the usual suspects who love to rant about the benefits of the "unstructured play time" they had when they were kids). Far less rowdy partying goes on, provided one can successfully identify this type of student in the admissions process (which is actually not difficult, since a questionnaire can easily target kids who were active in large numbers of scheduled extracurricular activities). This has some interesting ramifications for the widespread ranting about how the latest generation of kids is going down the tubes. It may be true in some respects, but apparently not for this subset. Mind you, I don't know how well that applies to schools like MIT, which are notorious for their wild parties. UW was always a notoriously strait-laced "geek school" rather than a "party school", but after having spoken to a few people, it's become even more strait-laced in the 15 years since I left.

Just because MIT has fraternities doesn't mean it has wild parties, at least compared to Ivy league/Stanford.

My take on the situation from not getting into MIT and knowing a couple people who did is that just getting an A in high school calculus and a 5 on the AP test doesn't prove anything because sheer hard work can get though it, but since some many people get those accomplishments, they're not nearly enough to get into MIT or Caltech. Generally you'll need to prove engineering/science/math aptitude by participating in various math and science competitions such as science bowl or AMC and doing well in them. Living close to a major university helps as taking university math and science classes early helps a lot(you'll need to take calculus early which is also a sign of actual aptitude) or going far in the Intel or Westinghouse research contests.