How to Argue (on the LSAT)
All ARGUE family arguments have at least one problem. Just knowing that they have a problem is not enough. To make a useful prediction, you must be able to say what the problem is.
DESCRIBE family questions might have an error, but generally you shouldn't care.
On INFER family questions, the argument is just facts, so it wont have an error. That said, the answers on INFER questions are often proposed conclusions. On those questions, wrong answers may commit one of the errors described below.
Danger
These argumentative tools are useful on the LSAT.
And they're useful as reflection tools to help you internally navigate an argument.
But typically it isn't wise to use these techniques directly on other humans, especially on loved ones or on social media.
Truth vs. Validity
The LSAT doesn't care about truth; it only cares about logical validity. An argument is valid when it follows the rules of logic and the facts are logically connected to the conclusion.
In real life, hopefully you evaluate arguments based on their truth. But the LSAT is meant to be a closed universe that doesn't require any outside knowledge. On the LSAT, it doesn't matter what's true or false in the real world.
This means that you don't try to prove that the argument is wrong, you try to prove that it was badly made.
To challenge an argument's validity, you first accept the facts as given. Then you question whether those facts adequately support the conclusion.
Chicken Little
In the real world, if Chicken Little 
says to you:
Run! The sky is falling.
You're likely to question their fact. You'll say:
It doesn't seem like the sky is falling to me.
On the LSAT, however, you have to accept the truth of Chicken Little's fact.
Okay, the sky is falling.
What you can do is challenge the connection between their fact and their conclusion:
If the sky is falling, how will running help us?
3 types of logical error
- Gaps: the facts are not completely connected to the conclusion.
- Flaws: the argument committed a specific type of common error.
- What ifs: there's an alternative explanation for the facts or the conclusion.
overlap between flaws, gaps, and what ifs
Some flaws, like fresh ideas in the conclusion and conditional errors are just another way of describing a logical gap.
Other flaws, like the causation, sampling and comparison flaws may be what underlies your what ifs.
MECHANICAL ARGUE questions usually have a gap. What ifs usually only help on ORGANIC ARGUE questions.
Example
Argument:
The t-shirt is on sale. So you should buy it.
Do you see a gap?
One gap: On sale is not connected to should buy
Do you see a flaw?
One flaw: Should is a fresh idea in the conclusion.
Any what ifs come to mind?
Some what ifs:
- You don't like t-shirts.
- They're sold out in your size.
- You're a freegan who never buys anything.
- It's a cotton t-shirt and you're a carnivore who only wears leather and wool.
Notice: What are the chances your exact what if appears in the correct answer? There's a vast range of possible objections; that's why you need to stay flexible on ORGANIC questions.
The Common Flaws
Here are the 5 logical flaws worth learning and memorizing:
- Fresh ideas in the conclusion
- Conditional errors
- Association isn't causation
- Bad samples
- Bad comparisons
there are many other flaws
There are dozens of other logical errors, some even have fancy Latin names.
If you notice another flaw commonly repeat on the LSAT and you don't intuitively spot that flaw, then you can memorize it, too.
But if the flaw isn't common, or you're able to see it easily, then why does it need a name?
Fresh ideas in the conclusion
Charlatans love to begin with obvious facts to make themselves seem commonsensical, then they leap to their grand, unsupported conclusions.
Spot fresh ideas errors in the wild
Look for big, shiny, or strong words in the conclusion.
Spot the fresh ideas in each these three bad arguments:
We all want to be rich. So we need to shrink big government.
New ideas: Where did big government come from? Why is shrink the solution?
Only 5 employees voted yes. Therefore a majority opposed the measure.
New ideas: 5 is an absolute number. majority is a relative number, so majority is a new idea.
We all liked the movie trailer. So we must go see the movie right now.
New ideas: Woah, right now and must? Those are much stronger words than liked.
Conditional errors
Conditional errors read the arrow backwards.
The LSAT often describes conditional logic errors as "confusing a necessary for a sufficient condition."
Example 1: 
All frogs are amphibians. Newts are amphibians, therefore newts are frogs.
Fact 1: Frog --> Amphibian
Fact 2: Newt --> Amphibian
Conclusion: Newt --> Frog
Error: Turning one of those arrows backwards to connect Newt --> Amphibian --> Frog
Example 2: 
True statement: All squares are rectangles.
True statement (contrapositive): If a shape is not rectangular, then it cannot be a square.
False statement (illegal reversal): All rectangles are squares.
False statement (illegal negation): This shape is not a square, so it cannot be a rectangle.
Spot conditional errors in the wild
If you see conditional logic words like "if" "all" "unless" and especially "only" in an ARGUE argument, there's a good chance it made a conditional error.
Association isn't causation
Legally and philosophically, it's hard to prove that something caused something else.
Consider how long it took to prove that smoking causes cancer.
You've probably a pedantic person say "correlation isn't causation." What's that mean?
"Correlation" simply means two things are associated. Things can be associated in time (one happens, then the other happens) or space (wherever we see the one thing, we also see that other thing).
Correlation examples
1: In cities where a larger percentage of cyclists wear helmets, more cyclists suffer brain injuries.
2: As student confidence increases, so does speed.
3: Corporate mergers are typically followed by an increase in the price of goods.
Humans brains like the neatness of causation too much. We're wired to rush to conclude causation from association.
Spot Causation in the wild
Look for association words in the facts and causation words in the conclusion.
Association words: tends to, more/less likely; increases/decreases with; before/after.
Causation words: caused by; effected by; impacts; leads to; (or any active verb).
Association is evidence in favor of causation, but it isn't enough to prove causation. Association is vulnerable to 3 criticisms:
1. random chance,
2. a 3rd cause, or
3. reversed causation.
Example: why we can't conclude causation from correlation.
Whenever we see Ash at a party, we also see Bee. Obviously, we can conclude that Ash has asked Bee out.
Three objections:
- We haven't been invited to that many parties. Maybe someone else often spots Ash at parties without Bee, or Bee without Ash.
- Sia independently invites both Ash and Bee.
- Bee asked out Ash.
Bad Samples
Spot Sampling errors in the wild
To spot this error in the wild: look for studies and numbers.
The more the LSAT describes the methodology of a study, the more suspicious you should be. If the LSAT says "Research shows...." that's probably just a fact you should accept.
Bad samples are (1) small, (2) unrepresentative, and/or are studied with (3) suspect methods.
Observations of 10,000 children revealed that people rarely wear deodorant.
Children might not act like adults.
Maybe children sweat less, so they have another reason for not wearing deodorant.
This study relies on an unrepresentative sample.
A telephone survey of 10,000 Americans revealed that Americans rarely wear deodorant.
The sample size is large and it's comparing Americans to Americans, so it's representative.
But maybe people lie on the phone?
This study relies on suspect methods.
??? Observations of 30 Americans revealed that Americans rarely wear deodorant.
`30` people is too small a group to conclude anything about millions of people.
This study relies on a too small sample.
Bad Comparisons
The comparison error takes two forms:
- Same Same: These things are alike in one way, so they must be alike in another way.
- Different Different: These things are different in one way, so they must be different in this other way.
Example
Same Same: Elephants and Mice are both grey mammals. Elephants are large. So mice must also be large.
Different Different: That apple is red and sweet. But this fruit is green and tart, so this fruit must not be an apple.