Curves, low averages and exam structures

If anything, the most common reason why people choose UofT is for its reputation, despite the school being notorious for its difficulty. In terms of most graduate/professional programs, however, reputation means actually matters very little- and many who chose the school for the very reason (while having aspirations of entering X/Y graduate/professional program) are often left with a sub-par GPA and little competitive chance into their desired program.

On the flip side, there are many students here who are extremely bright and have near perfect GPAs - they are the ones who understand the system best and can study accordingly to ensure they score within the same quartile every time.

It'd be nice if they shared more, right?


Lets examine two of the most discussed GPA topics about UofT;

Curves and the 65 average.

65 - a borderline C+ or 2.3 GPA score is the ideal average which first year science departments often attempt to maintain. Reasons? Sheer difficulty of the course, standardized/limited grade distribution, limit grade inflation, crushing dreams, etc. The average itself, however, cannot be used completely as a proper representation of the class/school's difficulty. A 65 average can be obtained by a multitude of distributions- it may mean that, for example, in a class of 100, 50 people have 70 and 50 have 60. In other cases, 50 with 50 and 50 with 80 - statistically speaking, no information about the variance of the classes grades are given. In the extreme case, amongst 3 people, 2 scoring 90 and one with 15 make a 65 average- yes, people do score that low and fail- but on the other hand, people also do amazingly well.

To keep the average at 65, professors may curve tests upward when the average is too low. These curves are often in the form of eliminated questions, which give varying amounts per person, or flat percent boosts. Eliminated question curves are a bit funky- but there are two main ways by which professors choose to distribute, one fairer than the other. In one method, people who answered correctly are rewarded, and in another, the question is eliminated, be it right or wrong. The latter method often gets people complaining, and is used less often than flat percent boosts where everyone gains an equal amount, whereas with elimination, some may lose marks.

The question many have is whether or not professors curve a class down when the grade are too high- it happens, but indirectly. After an easy exam with a class average of say 80, the next exam may be completely devastating and leave the class with a 50 average- effectively bringing the average back to 65. This is the most common method. Alternatively, if no up curve occurs in a class which performs too badly, then the next exam may be extremely easy, or final grades will be adjusted entirely to compensate. It all depends on the professor.


Yet, thankfully, as with any system, there is always counter-play.


The Exam Model

From the many exams I've written and the many intellectual students with which I've conversed with, we seem to reach a consensus of how exams are constructed. Here's how it looks:


    

In all honesty, 40% of the questions are going to be straight up easy- by this I mean they're general facts which the professor has highlighted, points that are repeated emphasized, things that are core to learning more in-depth concepts. For example, if about plants, they'd be questions like: "What organelle differentiates plant and animal cells? A. None, B. Cell wall, C. Mitochondria, D. Golgi Apparatus". Just standard knowledge.

30% will be simply medium.. not too difficult but at the same time not as brainless as the previous question type. This typically includes statement based answers with a true/false fact hidden within. Otherwise, memorization of say a specific enzyme in the Citric acid cycle would suffice.

20% will be more theoretical analytic based questions where there will be one immediate answer used to "bait" the writer, but the real answer would be X since Y condition prohibits the most common sense answer of being true. Look out for these.

There is no defining border between the final two categories.

5% will be just plain hard, whether it be memorizing the most minute detail of a slide (such as name of a species of animal shown on a slide as a brief example), or an extremely difficult or ambiguous application question which forces you to consider various details.

The final 5% are what I call curve resetters- simply questions so ridiculous/hard/specific in detail that I expect they put on just to have as questions to eliminate in the chance of a curve. They fall in similar category to the previous category, except their intention is to cap the maximum score at 95 pre-boost. The guy at 95 goes to 100 if a curve is needed.


Thus, the theoretical individual who understands content at an average level can score up to 70. As one travels further up the scale, they know the content at a deeper level. Some can go as far as to get 100 despite the resetters, given they know what to study and how to manipulate their information well enough.

Conclusion?

The entire scale is relative- just be better than the rest.

That's always easier said than done, of course. I'll share my techniques in a later date.