Analyzing Flaw Stimuli from a Logical Perspective

Download a preview of

LR Perfection Free Preview: Chapter 8 Part II

At the same time as we analyze the argument’s structural shifts, we must also keep a close eye out for the presence of conditional and causal logic in the author’s argument.

We have covered conditional and causal logic in depth in “Core Habits: Logic,” so here, we will focus on these relationships’ flawed/invalid variations.

A Flaw argument may contain conditional or causal reasoning, but that doesn’t necessarily mean that they will be erroneous. Roughly 30-40% of all Flaw Questions will contain either conditional or causal reasoning, but only 3/4 of these questions have a conditional or causal flaw.

Just as I try to analyze the argument’s structure as soon as I start to read the stimulus, I am also looking for signs of conditional and causal logic. I do this by specifically looking for indicator words (see the previous chapter). If I do find either conditional or causal logic, the first thing I do is check to see if it is valid.

***

8.2.1 Conditional Flaws

Given the conditional relationship A → B, the only valid extrapolation is that BA. Other than the contrapositive, there is absolutely nothing that we can derive from a conditional relationship. Conditional flaws come in two forms: the author can erroneously reverse the relationship without negating the conditions, or the author can negate both sides of the conditional without reversing the relationship. 

A → B (original conditional relationship)

BA (proper contrapositive, this is correct)

B → A (reversed original relationship without negation, this is wrong)

AB (negated conditions without reversing relationship, this is also wrong) 

The same rule applies when the conditional relationship involves and/or relationships. Only now the “and” is turned into “or,” and vice versa. 

A → B and C (original conditional relationship)

B or CA (proper contrapositive)

A and B → C (original conditional relationship)

CA or B (proper contrapositive) 

A → B or C (original conditional relationship)

B and CA (proper contrapositive)

A or B → C (original conditional relationship)

CA and B (proper contrapositive) 

Whenever you recognize conditional logic in the author’s argument, make sure to check to see if they commit a conditional logic flaw. If a conditional relationship is present, does the author reverse the relationship without negating it in his subsequent reasoning? Does the author negate the terms without reversing the relationship? Both would be wrong. 

Conditional flaws are fairly easy to detect, because once you are comfortable identifying conditional relationships, you can mechanically check for errors without having to think too much about it. 

Another benefit of recognizing conditional flaws is that when you are trying to match up a conditional error with the correct answer choice, the process is also straightforward. Here are some examples of answer choices describing conditional flaws: 

Takes a condition necessary to be a condition sufficient

Takes for granted that an assumption required to establish the argument’s conclusion is sufficient to establish that conclusion

Treats a statement whose truth is required for the conclusion to be true as though it were a statement whose truth ensures that the conclusion is true

Confuses a stated requirement with a sufficient condition

In the hardest conditional flaw questions, the test makers will try to throw you off by dressing up the error committed in the argument as another flaw type. Misdirection, or “Red Herrings,” are a recurrent technique used in the most challenging flaw questions. We will talk about this once we have covered all 23 flaws. 

Take a look at the following question:

PT22 S2 Q25

A recent survey showed that 50% of people polled believe that elected officials should resign if indicted for a crime, whereas 35% believe that elected officials should resign only if they are convicted of a crime. Therefore, more people believe that elected officials should resign if indicted than believe that they should resign if convicted.

The reasoning above is flawed because it

A. Draws a conclusion about the population in general based only on a sample of that population

B. Confuses a sufficient condition with a required condition

C. Is based on an ambiguity of one of its terms

D. Draws a conclusion about a specific belief based on responses to queries based on responses to queries about two different specific beliefs

E. Contains premises that cannot all be true  

The stimulus starts off talking about a survey. As soon as I see this, I make a note to double-check if this is a sampling bias flaw question. The stimulus also contains conditional logic, so it could also be a conditional flaw hiding in the stimulus somewhere. Let’s dissect the argument: 

Premise: A recent survey showed that 50% of people polled believe that elected officials should resign if indicted for a crime

Poll: 50% believe Indicted → Resign

Premise: whereas 35% believe that elected officials should resign only if they are convicted of a crime

Poll: 35% believe Resign → Convicted 

Conclusion: more people believe that elected officials should resign if indicted than believe that they should resign if convicted

Number of people who believe Indicted → Resign > Number of people who believe Convicted → Resign

Notice the subtle shift from the premises to the conclusion? The author has swapped the ordering between “resign” and “convicted” going from the premise to the conclusion. In the second premise, 35% of the people believe that elected officials should resign only if they are convicted of a crime. “Resign” is the sufficient, while “convicted” is the necessary. 

In the conclusion, however, the author is talking about whether officials should resign if convicted. Here, “convicted” is the sufficient condition, and “resign” is the necessary condition. 

This argument contains an obvious conditional flaw. But it also contains a survey, so does it have a sampling problem too? We don’t know: we don’t have enough information to decide whether the sample size is adequately representative, whether there is sampling bias, and whether the methodology has flaws. In other words, it could be a problematic sample, or it could be a perfectly fine sample. 

As we mentioned previously, a flawed argument often contains multiple or multiple potential flaws. Don’t rest on your laurels after you’ve found just one. In addition, if there is one certain flaw and one possible flaw in an argument, as seen here, we always go with the most certain one. (The conditional flaw in this case.) 

Let’s take a look at the answer choices: 

A. Draws a conclusion about the population in general based only on a sample of that population

What the test makers are describing here is a sampling bias fallacy. The previous question where the author concludes anyone fishing for trout based on how the best fishermen felt about the best selling bait would be such a flaw. Here, even though a survey and sampling are involved, we simply do not have enough information to know whether such a flaw is committed. On the real test, I would keep this answer and move on. 

B. Confuses a sufficient condition with a required condition

This is the flaw we are looking for, the conditional logic flaw. 

C. Is based on an ambiguity of one of its terms

The flaw this answer is talking about is called Equivocation, where one word has two meanings and the meaning of the word shifts through the argument. 

D. Draws a conclusion about a specific belief based on responses to queries about two different specific beliefs

This answer is tricky because it’s half wrong half right. The author drew a conclusion about two specific beliefs (more people believe Indicted → Resign than Convicted → Resign) based upon two specific beliefs, one of which is the same (Indicted → Resign), and one of which is different. (Resign → Convicted)

E.. Contains premises that cannot all be true

This is the Self Contradiction flaw, it does not appear here. 

The correct answer is B. 

8.2.2 Causal Flaws

We discussed different types of logic used in the LSAT in detail in the previous chapter and explored the nature of causal logic in depth. Here is a refresher: 

Causal logic is not exclusive. Just because A can cause B doesn’t mean C, D, or E cannot also cause B. Take the following example: 

Cardio helps one lose weight. Peter lost so much weight, he must have done some insane cardio. 

What’s wrong with this argument? It’s possible that Peter did other things which caused him to lose weight, for example, maybe he dieted, or lifted weights, or got surgery, or a combination of the above. 

Whenever the argument presents us with some form of causal logic in a Flaw question stimulus, always ask yourself, could there be alternative causes or contributing causes

See 6-3-16 for an example of this type of causal flaw.

Similarly, one cause can have multiple effects. One trick the test makers use in flaw questions is to confuse side effects with the intended effect. For instance, drinking wine can have the side effect of giving you a headache, but the reason you drank wine wasn’t to get a headache. Just because A causes B, doesn’t mean A is intended to cause B. 

So whenever the author presents causal logic in the argument, ask ourselves the following questions:

Is the cause provided reasonable? Could there be alternative causes that the author has ignored?

Even if the cause provided is reasonable, could it be only one of multiple contributing causes all leading to the effect in question?

Can the cause provided by the author lead to multiple effects

Even if the cause-effect relationship the author provides exists, is the effect in question an intended or side effect

A common way causal flaws appear in Logic Reasoning questions is in the famous Correlation-Causation format. The author presents a correlation, A ~ B, and concludes that the relationship is causal, or A ⇒ B. 

A correlation could be due to causation, other reasons, or even a statistical fluke. Don’t automatically assume cause and effect when presented with a correlation. Whenever the argument presents a correlation and concludes that there is a causal relationship, we ask ourselves three questions: 

If the author sees A ~ B, and concludes A ⇒ B, is it possible that

C/D/E ⇒ B? (Could it be that alternative factors caused B?)

C ⇒ A and B? (Could it be that there is a third common cause?)

B ⇒ A? (Could the author have reversed the causal relationship?)

Sometimes the author will make the argument slightly more complicated by negating the correlation in the premise and thereby negating causation in the conclusion. For example, the author will argue that because there is no correlation between coffee consumption and work performance, drinking coffee does not cause you to work more effectively. This variation of the Correlation – Causation argument type is more common in Strengthen/Weaken Questions but will occasionally pop up in Flaw Questions. Just be aware that even without correlation, causation can exist as well. 

Take a look at 58-1-11 and 62-4-19 for this type of flawed argumentation.

Let’s look at a more complicated variation of the causation flaw:

PT86 S1 Q19

Researcher: in an experiment, 500 families were given a medical self help book, and 500 similar families were not. Over the next year, the average number of visits to doctors dropped by 20 percent for the families who had been given the book but remained unchanged for the other families. Since improved family health leads to fewer visits to doctors, the experiment indicates that having a medical self help book in the home improves family health.

The reasoning in the researcher’s argument is questionable in that

A. It is possible that the families in the experiment who were not given a medical self help book acquired medical self help books on their own

B. The families in the experiment could have gained access to medical self help information outside of books

C. A state of affairs could causally contribute to two or more different effects

D. Two different states of affairs could each causally contribute to the same effect even though neither causally contributes to the other

E. Certain states of affairs that lead families to visit the doctor less frequently could also make them more likely to have a medical self help book in the home

This question is hard because it’s not the usual Correlation – Causation variation we have encountered so many times before. The test makers came up with a slightly more complicated alternative.

As we read the first half of the stimulus, we discover a correlation between having the medical help book and fewer visits to the doctor’s office. Those who have the book didn’t go to the doctor’s as much, while those who didn’t have the book had the same number of visits as before. So there is a correlation, but is there causation? I suppose having the medical self-help book can cause you to go to the doctor’s office less. You can probably find some treatments or diagnose your disease on your own with the help of the book. Let’s see what the author’s reasoning is. 

Instead of arguing that because Medical Self Help Book ~ Less Visits to the Doctor, Medical Self Help Book ⇒ Decreased Visits to the Doctor, the author comes up with a more complicated, alternative causal possibility. The author says that since improved family health leads to fewer visits to the doctors, the experiment indicates that having a medical self help book in the home improves family health. 

If we diagram out the author’s reasoning? It will look like this:

Medical Self Help Book ~ Less Visits to the Doctor’s Office (correlation found in the experiment)

Improved Family Health ⇒ Less Visits to the Doctor’s Office 

Medical Self Help Book ⇒ Improved Family Health

In abstract notation form, it will look like this:

A ~ B,

Because C ⇒ B, 

A ⇒ C (in the author’s view, A ⇒ C ⇒ B)

The author’s causal chain is certainly a possibility, but could it also be possible A ~ B can simply be explained by A ⇒ B.

Take a look at the following analogy if you are still a little confused:

Law students have stronger reading skills than students who didn’t go to law school 

Since reading Shakespeare improves one’s reading ability,

Going to law school causes students to read Shakespeare. 

This example is a direct parallel to the author’s reasoning above. The author’s explanation, while a possibility, ignores a glaring alternative: it’s also very possible that having the medical book was the direct cause for fewer doctor visits, just as a legal education can be the direct cause for students having stronger reading ability. 

So the flaw in this question, while a little non-conventional, is still causal in nature. The author has overlooked an alternative cause. 

Let’s now take a look at the answer choices:

A. It is possible that the families in the experiment who were not given a medical self-help book acquired medical self-help books on their own

Even if the families who were not given the book got the book on their own, where does that lead us? Their visits to the doctor’s office were not affected. So that can only mean one thing: the correlation between having the medical self help book and less visits to the doctor’s office isn’t as clear as the stimulus makes it out to be. This answer is sowing doubt in our minds about the phenomenon, rather than the author’s explanation. It is not attacking the author’s premise – conclusion core itself.

B. The families in the experiment could have gained access to medical self help information outside of books

If the families gained access to medical information outside the books, how do we explain the correlation between those given the books and decreased visits to the doctor? Did only those who were given the books do extra research? We can’t be sure. If everybody did outside research, how is it that only those given the books visited doctors less?

C. A state of affairs could causally contribute to two or more different effects

This answer is saying that a cause can have many different effects. For instance, eating sugary food can make you happy but also make you gain weight. In the stimulus, the author’s mistake was to assume that one effect can only have one cause. The author thinks less visits to the doctor is caused by improved health, which is in turn caused by having the medical book. The more likely alternative is that people with the books simply consulted the books when they got sick and saved a trip to the doctor’s.

D. Two different states of affairs could each causally contribute to the same effect even though neither causally contributes to the other

This is the correct answer, but in super abstract format. So we have two causes that can contribute to the same effect. Cause 1 (medical book) and Cause 2 (improved family health) each independently contribute to the same effect (less visits to the doctor), and there is no causal relationship between cause 1 and cause 2.

The correct answer, as we will see over and over again, will often be worded more vaguely and in more abstract terms than the wrong answers. So don’t eliminate an answer simply because we are unclear what it means. Try to extract what keywords we can from it and match it to the stimulus to have a better idea of what it’s trying to say.

E. Certain states of affairs that lead families to visit the doctor less frequently could also make them more likely to have a medical self help book in the home.

This answer choice is suggesting that instead of A ⇒ B, there is a common cause for both A and B. But in this question, we know that people who received the medical books did so because of their participation in the experiment, could participation in the experiment also have led to them going to the doctor less? This sounds rather far-fetched to me.

The correct answer is D.

Causation flaws are fairly common in Flaw Questions. Whenever we sense causal logic at play in an author’s argument, ask ourselves whether there are alternative causes or contributing causes that the author has willfully ignored. Similarly, whenever there is a correlation involved in a stimulus, and the author provides a causal explanation for it, try to think of alternative ways to explain the correlation. 

***

Now that we have a complete list of the structural and logical flaws to frequently appear in Flaw Questions, we must strive not only to memorize them, to commit them to heart, but also to be able to detect such flaws even when the stimulus arguments get vague, abstract, or are enshrouded in complicated language. Personally, I found the best way to quickly come to a strong grasp of these fallacies is to do some additional research (google these fallacies), and try to come up with your own examples. 

Being able to quickly figure out what flaw each of the answer choices are describing is also a skill that will greatly aid in our Flaw Question successes. The advanced test takers will be able to take a look at the answer choices, quickly realize the flaws many of the answer choices are describing, and be able to match up/eliminate these choices accordingly. 

Lastly, before we turn to the next segment of the chapter, the academically oriented student can further explore the topic of formal and informal fallacies in Douglas Walton’s Informal Logic: A Pragmatic Approach.