Sapper PE LS
Aspiring Couch Potato
- Joined
- Apr 25, 2006
- Messages
- 5,261
- Reaction score
- 910
I carpool with a fella who just took the PE. He's also a member of this board, but I won't dime him out here. Anyway, during our 45 minute car rides together we often have time to talk about a great many things and earlier this week on our way in to work, of course the topic of the exam results wait time came up. Having now been a member of this board for 8 years, I've seen all the complaints, I've made the complaints myself, but lets walk through why we wait, just so it's here, and so everybody knows the situation. But first, a couple of interesting facts that you may not have known.
Bar Exam results take longer. From what I can tell, the Bar exam is generally administered in July each year (any lawyers who want to correct me feel free) and results aren't usually released until November. That's worse than we have it, for sure.
Medical Board results take about three months. This is also currently longer than PE exam results. A few years ago, PE results took about 3 months, but now it's down to about 6 weeks. Medical professionals wait longer.
So, okay, so others have it worse than we do, sucks to be them, we're not lawyers or doctors, why can't we get our results faster? The reason is simple, because the results really matter. If you were taking a cosmetology test, it'd be pretty ridiculous to wait for two months for the results. Your practice of cosmetology will not result in dead people, most likely, if you do something wrong. However, as engineers, what we design and calculate has real meaning and real implications in the real world with regard to health, safety, and welfare. This may not be true for all engineering professions, but it certainly is for the majority of them. While we can easily fall back on the old adage of "well, my work gets reviewed by so many people anyway, it's never going to get built wrong" we know that is simply a cop out. When those other people retire and move on, if you haven't been thoroughly tested, you won't be able to review the younger folks. The entire profession will decline. So, a testing procedure that is taken very seriously and statistically measured and calibrated and adjusted is one of the responsible things our profession can do to ensure quality engineers. No, it isn't perfect, and many bad engineers slip through and perform bad work, but it's the best thing we can do. So, the exam gets scrutinized, and the results get evaluated to ensure that the test was sufficient, that the security was intact, and that all of the potential issues have been worked out. You either accept this fact or not, but moving forward, let's now talk about what all these steps are.
According to NCEES, each exam is automatically graded by machine (the scantron) and then a sample of those are taken and manually graded. I don't know the number of manually graded score cards, but we can assume that the math has been done to provide a statistically significant sample size by NCEES. Now, there are some 30,000 examinees each administration (per NCEES video http://ncees.org/exams/scoring/), which equates to 60,000 exam score sheets accounting for both morning and afternoon. If they sample 10%, that's 6,000 sheets that need to be manually graded. That right there will take some time. So, let's assign a random value to that, let's say it take s a week to mail in all of the exams to NCEES after the exam (Friday, they won't move till Monday, get to NCEES between Tuesday and Friday of the following week). Then lets say it takes one person 5 minutes to manually grade one score card. That equates to 750 hours of hand scoring. If one person was doing all of them, it would be 18.75 weeks. So, say they put 10 people on this task, then theoretically they'd have them all done in about two weeks. That's now three weeks after the exam. Then they have to compare the manually graded results to the computer scored results, let's give that a week (including running the statistical analysis). So now we're up to 4 weeks. At this point they've probably noticed some outlying data points and need to evaluate why this is the case. Let's give this a week. Up to five now or about 35 calendar days. Now it's time to start notifying the boards of the results. Boards say, got it acknowledged, go ahead and release results (and of course each board takes a different amount of time to do this), so we start seeing results pour in and lo and behold, this is happening around the 40 day mark, which is consistent with the results release times we've been seeing for the past few years.
Now, the state specific surveying exam results on the other hand, we shoulda had that shit yesterday!
Bar Exam results take longer. From what I can tell, the Bar exam is generally administered in July each year (any lawyers who want to correct me feel free) and results aren't usually released until November. That's worse than we have it, for sure.
Medical Board results take about three months. This is also currently longer than PE exam results. A few years ago, PE results took about 3 months, but now it's down to about 6 weeks. Medical professionals wait longer.
So, okay, so others have it worse than we do, sucks to be them, we're not lawyers or doctors, why can't we get our results faster? The reason is simple, because the results really matter. If you were taking a cosmetology test, it'd be pretty ridiculous to wait for two months for the results. Your practice of cosmetology will not result in dead people, most likely, if you do something wrong. However, as engineers, what we design and calculate has real meaning and real implications in the real world with regard to health, safety, and welfare. This may not be true for all engineering professions, but it certainly is for the majority of them. While we can easily fall back on the old adage of "well, my work gets reviewed by so many people anyway, it's never going to get built wrong" we know that is simply a cop out. When those other people retire and move on, if you haven't been thoroughly tested, you won't be able to review the younger folks. The entire profession will decline. So, a testing procedure that is taken very seriously and statistically measured and calibrated and adjusted is one of the responsible things our profession can do to ensure quality engineers. No, it isn't perfect, and many bad engineers slip through and perform bad work, but it's the best thing we can do. So, the exam gets scrutinized, and the results get evaluated to ensure that the test was sufficient, that the security was intact, and that all of the potential issues have been worked out. You either accept this fact or not, but moving forward, let's now talk about what all these steps are.
According to NCEES, each exam is automatically graded by machine (the scantron) and then a sample of those are taken and manually graded. I don't know the number of manually graded score cards, but we can assume that the math has been done to provide a statistically significant sample size by NCEES. Now, there are some 30,000 examinees each administration (per NCEES video http://ncees.org/exams/scoring/), which equates to 60,000 exam score sheets accounting for both morning and afternoon. If they sample 10%, that's 6,000 sheets that need to be manually graded. That right there will take some time. So, let's assign a random value to that, let's say it take s a week to mail in all of the exams to NCEES after the exam (Friday, they won't move till Monday, get to NCEES between Tuesday and Friday of the following week). Then lets say it takes one person 5 minutes to manually grade one score card. That equates to 750 hours of hand scoring. If one person was doing all of them, it would be 18.75 weeks. So, say they put 10 people on this task, then theoretically they'd have them all done in about two weeks. That's now three weeks after the exam. Then they have to compare the manually graded results to the computer scored results, let's give that a week (including running the statistical analysis). So now we're up to 4 weeks. At this point they've probably noticed some outlying data points and need to evaluate why this is the case. Let's give this a week. Up to five now or about 35 calendar days. Now it's time to start notifying the boards of the results. Boards say, got it acknowledged, go ahead and release results (and of course each board takes a different amount of time to do this), so we start seeing results pour in and lo and behold, this is happening around the 40 day mark, which is consistent with the results release times we've been seeing for the past few years.
Now, the state specific surveying exam results on the other hand, we shoulda had that shit yesterday!
