Avoid Statistical Flaws In Critical Reasoning

by on November 4th, 2012

So it is upon us. The much-anticipated final Jobs Report before the 2012 Presidential Election has been released, and its results will fuel debates all weekend and can have a significant impact on your GMAT verbal score if you pay attention to the arguments that surround it.

Here’s what happened – the American economy added 171,000 new jobs, beating economists’ predictions by a healthy margin  but the unemployment rate ticked up a tenth of a point from last month’s figure. And in full GMAT Critical Reasoning mode, pundits and political representatives immediately began using those statistics to draw unsupported conclusions. Check out these Critical Reasoning style Weaken questions you could make from the arguments:

1) The unemployment rate rose by a tenth of a point. Clearly this is evidence that the economy is stagnant at best, and likely regressing.

What would weaken this argument? How about:

(A) The rise in unemployment derived from a re-entry to the labor force by many who have regained confidence in the economy.

2) The American economy added 50% more jobs than economists had expected. Consequently we can conclude that the economy is rapidly improving.

What would weaken this argument? How about:

(A) Many of the newly-created jobs were of the part-time or seasonal variety, and thereby only contribute a fraction to the overall economy’s health.

Those arguments and counterattacks will be made throughout the weekend and well into Election Day afternoon on Tuesday. And understanding the flaws in those arguments can help to ensure that the job you create for yourself after business school is higher-paying, as this blueprint for Critical Reasoning questions is one you’ll see many times on your GMAT exam. The blueprint:

*Statistic that has a precise usage

*Conclusion that goes outside that scope or that misses the statistic’s nature by a small margin

*Correct Strengthen answer that links the stat more closely to the conclusion


*Correct Weaken answer that demonstrates that the statistical premise doesn’t quite match the scope of the conclusion.

Here are some examples from GMAT questions:

Statistic: More than 90% of residential fires are extinguished by someone who resides in that home.
Conclusion: Therefore, the proposed law requiring sprinklers in new homes will not significantly decrease property damage due to fire.

Flaw: The number of fires isn’t the same as property damage based on fires! What if the majority of property damage comes from that ~10% of fires that aren’t put out by a family member? The sprinklers would help!

Statistic: Under a costly intensive supervision program for paroled criminals, the percentage of parolees arrested is no different from the percentage of parolees who were released under the routine supervision program.
Conclusion: Therefore, the intensive supervision program is no more effective than the routine supervision program at preventing released criminals from committing additional crimes.

Flaw: “Arrested” isn’t the same as “committing additional crimes”. If the intensive program catches every parolee who commits a crime, but the routine program only catches half, then these stats would show that the intensive program is more effective.

Intensive: 10 people, 2 crimes, 2 arrests (20% arrested/20% committed crimes)

Routine: 10 people, 4 crimes, 2 arrests (20% arrested is the same, but the “committed crimes” rate was twice as high)

In each of these questions, the answer choices are wordy and  convoluted – if you don’t recognize the logical flaw before you get to the answer choices, you will likely struggle. But if you know what you are looking for you can be quite efficient when attacking the answers.

So take a lesson from the most recent jobs report and all the arguments that will be made this weekend and into next week. The conclusions that will be drawn will almost all go outside the scope of what the statistics can actually prove, and if you recognize those subtleties and propose your own weaken answers, you will be that much more likely to add a nice statistical measure to your own job search, a high GMAT verbal score.


  • Hi, I'm a current Veritas student. I am interested in reading more about common logic fallacies and was wondering if you have any suggestions on the reading list. 

  • Commendable article!

Ask a Question or Leave a Reply

The author Veritas Prep gets email notifications for all questions or replies to this post.

Some HTML allowed. Keep your comments above the belt or risk having them deleted. Signup for a Gravatar to have your pictures show up by your comment.