8mo ago

The state of things

You're viewing a single thread.

103 comments

An education system that always fails a set number of people, regarless of how well people do, is a bad system, however.
- There will always be 25% in the bottom quartile, regardless of how well any students perform.
  
  I think their point is that you could have people in the bottom quartile who learned what they are expected to, are capable, but are failed anyway because of how they compare to others.
  (Assuming curved tests really work like that, never bothered reading the pretty long grading policies)
  
  Nothing about the post says anything about how many students passed or failed. Just that the lowest 25% are the lowest 25%.
  Yes, A = A.
  
  Oh... you didn't post this as a joke? This is depressing.
  
  I did post it as a joke. Note how I started my comment with "I think their point is", that isn't my view :)
  There's also difference between pretending to miss for a joke that dividing things into quartiles necessarily means a bottom 25% exists (what the meme does). And noting that it's weird failing people based on how they do compared to others and not if they actually actually learn (me suggesting what the user JackGreenEarth was probably trying to get at).
- As Nougat said, this has nothing to do with passing or failing and is just a consequence of measuring performance. If 100 people take a test and the lowest 25 scorers all have a 95 out of 100 points then they are still in the bottom quartile regardless of the fact that every single student passed with flying colors.
  Showing a bell curve with no context means nothing.
- I’ve been graded on a curve, and I’ve done it myself a couple of times. IMO, it’s usually a sign of a bad class (too much material being crammed in) or a bad teacher (didn’t get the concepts across to the majority of the students).
  That said, it’s usually done when it’s needed to prevent a significant portion of the class from failing. I remember a chem exam I took where a 16/100 was a C.
  The basic idea is that grades are normally distributed (ie a bell curve) which allows you to find the average grade range and shift the letter grade (eg a C or C+). There’s some professors who take the idea too far and rather than working off of an actual normal distribution try to fit the procedure to a simply skewed distribution or use it to pull down an 85/100 to a C, but in my experience that’s the exception to the rule, especially in math/science courses.
  Also, iirc this is a parody account.
  
  I agree. My calc I professor would just silently scribble equations on the board, then turn around, gesture wildly, and shout "You see".
  I remember right before the drop date, I had a 34 in the class, and he took time out of class to beg us to study because if too many people failed, he might have consequences.
  The only grade left was the final. I did much worse on it than the rest of the course, but my course grade shot up to the low 70s. Sure enough, I had the like 4th highest grade in the class.
  
  I don't believe that standardized tests are ever ghraded on a curve.
  
  You’re right, but it depends on how you want to think about it. They’re not necessarily graded on a curve, but with standardized tests you usually have both a history and a design target. They’re intended to produce (for example) a normal curve with a specific mean (eg mean IQ = 100) and they’ll adjust the test year over year to keep within those bounds. In other words, the grades don’t change but the test does.
  Curves exist because failing 90% of your class is a really bad look.
  
  Oh I’m not saying standardized tests are perfect by any means. Plenty of flaws.

103 comments