Spotting Percentage Traps and Biased Samples

Adrian Calloway

So if I tell you that a local city council increased its funding for park maintenance by five hundred percent, your instinct, your immediate instinct is to think, wow, they spent a fortune.

Nora Ashford

Right, five hundred percent sounds like, I don't know, millions of dollars.

Adrian Calloway

Exactly. But if the park budget last year was two dollars, a five hundred percent jump brings you to twelve dollars. You can buy, uh, maybe a box of trash bags.

Nora Ashford

Twelve bucks total. That is, that is the classic base rate fallacy right there.

Adrian Calloway

It is the number one quantitative trap on the Logical Reasoning section. The test writers love to exploit the gap between relative shifts and absolute numbers. Percentages cannot be substituted for actual numbers without knowing the group size.

Nora Ashford

You know, I remember when I was prepping, I used to see a huge percentage in a stimulus and my brain would automatically translate it into a massive total quantity. Like, a one hundred percent increase on ten dollars is just ten bucks. But a one percent increase on a million dollars is ten thousand dollars.

Adrian Calloway

Precisely. The percentage tells you the rate of change, not the size of the payload. So whenever I am reading a stimulus, I run what I call a quick premise audit. Is the stimulus giving me a proportion, a rate, or a hard count?

Nora Ashford

A Denominator Audit. That is what we always called it. Before you jump to the answer choices on a Flaw or Weaken question, you have to stop and ask, do I actually know the baseline denominator here?

Adrian Calloway

Right. If an argument says the percentage of commuters taking the train doubled, but it never gives you the total number of commuters, you have no idea if that means ten people or ten thousand people. If the denominator is hidden, the argument is built on air.

Nora Ashford

And on a Weaken question, the correct choice is almost always going to expose that missing denominator. It will say something like, well, the total population of commuters dropped from one million down to fifty people. Which completely undermines the claim that train transit is booming!

Adrian Calloway

Exactly. You attack the hidden baseline, and the entire argument collapses clean to the floor.

Nora Ashford

Okay, but what about when the test writers throw actual numbers at us, but the group itself is totally cooked? Like those survey arguments.

Adrian Calloway

Ah, the self selection trap. My absolute favorite. A premise says, uh, ninety percent of members surveyed at a local health club reported being in excellent physical condition. Therefore, the city population as a whole is overwhelmingly healthy.

Nora Ashford

Right! Because people hanging out at a gym on a Tuesday morning are definitely representative of everyone in town.

Adrian Calloway

Precisely. They confuse correlation with causation, and they rely on an unrepresentative sample. The people who choose to join a gym and fill out a survey about health are self selected. They already care about fitness.

Nora Ashford

On a Weaken question under timed conditions, these can be tricky because the wrong answers sound so authoritative. They will toss in fancy statistics about gym equipment or tuition costs to distract you. How do you cut through that noise in under a minute?

Adrian Calloway

I use a mental drill called the Group Size Swap. The moment I see a specialized sample, I swap it with the broader target population in my head. If the conclusion is about all city residents, but the premise is about health club members, I immediately flag the gap between health club members and all residents.

Nora Ashford

Oh, that is good. So if the stimulus talks about a survey of corporate executives, you instantly swap that out and ask, does this apply to entry level workers? Or if it surveys dog owners, does this apply to all pet owners?

Adrian Calloway

Exactly. You isolate the specific sample, compare it to the conclusion's target population, and ask if there is a reason the sample would be biased. If the survey relies on volunteers or people with a specific interest, it is inherently flawed.

Nora Ashford

So to wrap this all up, whenever we see numbers or statistics on Logical Reasoning, step one, run the Denominator Audit to check if we are dealing with percentages or actual counts. And step two, do the Group Size Swap to see if the sample actually represents the target group.

Adrian Calloway

Spot on, Nora. Master those two checks, and quantitative flaws become free points.

Nora Ashford

Alright, go practice your denominator audits, everyone. Talk to you next time.