2 Standard Deviations= everyone passes?

[Dr.UserName]

Ok Its been a long time since I've taken statistics so I'm having a hard time figuring something out. I was reading over the course information for an upcoming class and it stated that Pass was scoring greater then 2 standard deviations below the class mean.

So I wikipedia the value of 2std and learn that about 95% of the population will fall within a passing score and since the remaining 5% is split within the extremes about 2.5% fail/get a non-passing score. Correct?

Also, and this is the biggest debate I'm having with myself, is that wikipedia says that this is only true when you have normal distribution. How this is defined I couldn't find. What would be abnormal?

I think my overall question would be does everyone pass (minus 2.5%) or am I missing something? Also, to defend myself about everyone passing there is also a 70% clause as well.

To quote a favorite radio saying, Long time listener here, first time caller would love to hear your responses.

Thanks

---

Members don't see this ad.

 

[bylee]

normal distribution='bell curve'
how many students in this class?
if pass = >2sd below mean, than it sounds like a great class to me 👍

 

[OncoCaP]

Ok Its been a long time since I've taken statistics so I'm having a hard time figuring something out. I was reading over the course information for an upcoming class and it stated that Pass was scoring greater then 2 standard deviations below the class mean.

So I wikipedia the value of 2std and learn that about 95% of the population will fall within a passing score and since the remaining 5% is split within the extremes about 2.5% fail/get a non-passing score. Correct?

Also, and this is the biggest debate I'm having with myself, is that wikipedia says that this is only true when you have normal distribution. How this is defined I couldn't find. What would be abnormal?

I think my overall question would be does everyone pass (minus 2.5%) or am I missing something? Also, to defend myself about everyone passing there is also a 70% clause as well.

To quote a favorite radio saying, Long time listener here, first time caller would love to hear your responses.

Thanks

I don't know anything about what happens at Penn, but test distributions are often not normal distributions. They may be bimodal, trimodal, etc., in that there are clusters of students at various score ranges. This means that everyone could pass if the overall mean is low enough and there are several high scores (roughly: the 5% who don't fall inside the two standard deviations are at the high scoring end) ... or that roughly 5% fail on the low end. If the distribution is non-normal enough, many more could fail than pass because the actual distribution is such that the mean is say, 70%, and there was not much of a spread for most students with a handful at the very low end (having missed a high value question, etc.).