*Spoiler* AAMC FL 1 Question 56 P/S

[Dr.Qball]

Hello all, Just doing the usual review but this question I cannot justify why B is better than C. For me I think B is more wrong because there is no way you can justify that statement with a simple discrimination file analysis and I think B implies a more casual role. Also for C I feel its more in line with what the data is saying because equivalent isnt the same as equal to.

---

Members don't see this ad.

 

[aldol16]

B is better than C because race and gender discrimination do manifest in similar ways because the distribution of categories of discrimination in each case is similar. C is also wrong because the incidence of each is clearly different from the other. In the 15k claims, there were 9k race discrimination and 6k gender discrimination.

Also, "equivalent" is the same as "equal to."

 

[Dr.Qball]

Ahh okay didnt know I had to bring outside knowledge for this one since it said based on the passage. Also I always thought equivalent was more like saying similar as than directly equal to but I get what youre saying thanks.

 

[aldol16]

Ahh okay didnt know I had to bring outside knowledge for this one since it said based on the passage. Also I always thought equivalent was more like saying similar as than directly equal to but I get what youre saying thanks.

What outside knowledge? Do you mean the definition of "incidence"?

 

[Dr.Qball]

Yes. But even still I could of gotten the right answer without knowing incidence.

 

[Mr. otcoD]

See my thinking was this:
Just because they have the same similar percentages doesn't indicate the same number of incidences. Whats the definition of incidence? The number of new cases that pop up each year.
Take for example Exclusion. For Race its 5% of 9013, which is ~450 cases. For Gender, its 4% of 6162, which is ~246 cases.
Answer C is saying that the number of cases are equivalent, which as shown are not. Hence B is a better answer because B is referring more towards the percentages being similar compared to the actual number of new cases

 

[Thoth]

Ahh okay didnt know I had to bring outside knowledge for this one since it said based on the passage. Also I always thought equivalent was more like saying similar as than directly equal to but I get what youre saying thanks.

It does mean equal to. Here is an example:

Drug A is equivalent to drug B in efficacy in reducing risk of death in patients with severe heart failure.

That means that they have equal, or the same, effects.

Now, if you said drug A is similar to drug B, that is much more general because you are not specifying any particular metric. It is more accurate, because the side effects are probably slightly different, as are other nitpicky stuff that isn't relevant. The first one is appropriate to say equivalent to because you are comparing them in a specific population with a specific outcome that can be measured and shown to be the same. The second is not because if it was equivalent, literally 100% of clinical outcomes would be the same, including side effects, etc.

We often use them somewhat interchangeably in real life, but for the MCAT, you need to know it means 'the same as.'


It cannot be C because the incidence is clearly different. The incidence of race-related discrimination is 50% more than gender-related discrimination.

 

[aldol16]

Just because they have the same similar percentages doesn't indicate the same number of incidences. Whats the definition of incidence? The number of new cases that pop up each year.
Take for example Exclusion. For Race its 5% of 9013, which is ~450 cases. For Gender, its 4% of 6162, which is ~246 cases.

You don't even need to do that kind of math. Just look at the "n"s for each case. The "n" for race is clearly different from that of gender and therefore incidence is different.

 

[Dr.Qball]

Thank you all for the responses! Its been really helpful. I honestly didnt even see the "n=" on the chart when looking at the answers so B is not even close to being right making C the best option.

 

[AnotherLawyer]

You don't even need to do that kind of math. Just look at the "n"s for each case. The "n" for race is clearly different from that of gender and therefore incidence is different.

I'm not sure this is right. Knowing the n values is not enough to decide one way or the other without knowing how many women and minorities are in the relevant work force. If n(race) = 10 and there are 100 minorities in the work place and n(gender) = 5 and there are 50 women in the workplace, then the incidence of discrimination would be the same even though the n values are different.

Thus we cannot conclude C is incorrect, but there is no support for it either.

 

[aldol16]

I'm not sure this is right. Knowing the n values is not enough to decide one way or the other without knowing how many women and minorities are in the relevant work force. If n(race) = 10 and there are 100 minorities in the work place and n(gender) = 5 and there are 50 women in the workplace, then the incidence of discrimination would be the same even though the n values are different.

That depends on what you take to be the definition of "incidence." Normally, "incidence" is defined as the number of new cases during a specified period of time. Which is what the "n"is meant to give here. Another definition, which is not used here, is the number of new cases per specified number of people per unit time. So something like 5 cases per 50 workers. That would be a valid challenge and a nuance ignored by the question-makers.

However, I don't think anybody subscribes to your definition of incidence as being frequency of occurrence of some event in an arbitrarily-defined subset of the population in some unit time. This is because that subset cannot be easily defined and when defined, is done arbitrarily. What if you have n(gender) = 10 and 3 of those were reported by men? Let's say there are 50 women and 50 men in the workplace. What is the incidence of discrimination by gender in that case? One could say that the incidence of discrimination by gender in women is 7 and in men is 3, but what are the important sets here? 7 in 50 women and 3 in 50 men? 10 in 100 workers? Usually people just go by the latter because it's a better description of the population of interest, i.e. all workers.

 

[AnotherLawyer]

That depends on what you take to be the definition of "incidence." Normally, "incidence" is defined as the number of new cases during a specified period of time. Which is what the "n"is meant to give here. Another definition, which is not used here, is the number of new cases per specified number of people per unit time. So something like 5 cases per 50 workers. That would be a valid challenge and a nuance ignored by the question-makers.

However, I don't think anybody subscribes to your definition of incidence as being frequency of occurrence of some event in an arbitrarily-defined subset of the population in some unit time. This is because that subset cannot be easily defined and when defined, is done arbitrarily. What if you have n(gender) = 10 and 3 of those were reported by men? Let's say there are 50 women and 50 men in the workplace. What is the incidence of discrimination by gender in that case? One could say that the incidence of discrimination by gender in women is 7 and in men is 3, but what are the important sets here? 7 in 50 women and 3 in 50 men? 10 in 100 workers? Usually people just go by the latter because it's a better description of the population of interest, i.e. all workers.

It seems like you are in complete agreement that the n values do not tell the full story and therefore neither prove nor disprove option C.

 

[aldol16]

It seems like you are in complete agreement that the n values do not tell the full story and therefore neither prove nor disprove option C.

Like I said, it depends on your definition of "incidence." Sociologists don't have a universal definition of "incidence." Some define it as "number of new cases per unit time" and others define it as "number of new cases per unit people per unit time." This sort of ambiguity is why I'm a chemist and not a sociologist - how are you going to explain anything to anybody if you can't even define the thing you're explaining?

 

[Dr.Qball]

And this is why no one scores perfect on the MCAT.