GRE Argument Writing

It is pretty unrealistic to describe a group and then expect that every single member fulfills that characteristic. You can remember this fallacy by thinking about stereotypes. We generally think of stereotypes as harmful because they unfairly limit a certain group to one definable characteristic that is often founded on little to no evidence. In order to avoid the member-group fallacy, the argument should clearly state that a member is a representative of the group as a whole; most of the time, however, it won’t.

The speaker of an argument may assume that a certain course of action is necessary or sufficient to achieve a result. The “necessary” line of reasoning is particularly weak if the speaker does not provide evidence that no other means of achieving the same result is possible. For example, a superintendent of a school argues that adopting a certain marketed reading program is necessary–i.e. the only means–to increase reading skills of students.
The “sufficient” line of reasoning is weak if the speaker fails to provide evidence that the proposed course of action would be sufficient to bring about the desired result by itself. In the above example, the superintendent may not have shown that the reading program by itself is enough to raise reading levels. There are other factors involved in this proposed outcome: preparedness of teachers and attentiveness of students.

The speaker may come to a conclusion about one thing on the basis of another thing. For example, if the manager of a business, say a trading card shop, may find that a big competitor in a different city has increased sales by moving from a downtown location to a suburban one. The argument may seem sound, but we can’t completely analogize these different trading-card shops. First of all, the demographics in their respective cities may respond to different incentives. Maybe that particular city’s downtown district was already on the rise, and the relocation merely reaped the benefits? Without this thorough background info, we can’t make this analogy.

This fallacy, more lovingly known as the post hoc fallacy, may be one of the most common you’ll encounter when examining the pool of arguments, so it’s essential that you master it. There are two basic ways a fallacious cause-and-effect claim can be made. First, the speaker may claim that a correlation suggests causation; just because two phenomena often occur together, it doesn’t mean that one event causes the other. Second, the speaker may claim that a temporal relationship suggests causation; by the same logic, just because one event happens after another, it doesn’t mean that event caused the other to occur.
A speaker may often use correlation to simply causation when a lurking variable is present. Take this argument for example: As ice cream sales increase, the rate of drowning deaths increases, so ice cream causes drowning. This one may take some head-scratching to realize that ice cream is more popular in the summer months, when water activities are also more popular.

You will often find that these arguments cite statistical evidence to bolster their claims. As you may find out, simply citing evidence does not prove a claim since the statistics may be faulty, unrepresentative, or inapplicable. The speaker may often cite a statistic that polled a sample group in order to draw a conclusion about a larger group represented by the sample. This is where problems can arise. For a sample to adequately represent a larger population, it must be of significant size and characteristically representative of the population. For example, a speaker may try to make a broad claim about graduate school’s impracticality by citing statistics from one particular university, e.g. 80 percent of University X undergrads were employed within one year of graduating, while only 50 percent of the graduate students of the same university were employed after one year. The statistics of one university simply cannot account for a sweeping claim about graduate education. To really identify the source of the employment disparity, we’d have to compare the admission standards for undergrads and grad students, examine the economy of the surrounding area, compare the types of jobs sought by undergrads and grads, and show the distribution of majors among grads and undergrads.

Tainted data is the second problem that could arise with data samples. For data to be considered legitimate it has to be collected in an unbiased, fair, and scientific manner, otherwise the quality of the data is compromised. For example, if there is reason to believe that survey responses are dishonest, the results may be unreliable. Further, the results may be unreliable if the method for collecting the data is biased, e.g. if the survey is designed, consciously or unconsciously, to yield certain responses. To spot tainted data, make sure that if a survey should be conducted anonymously–like in the workplace–then it is indicated. Also, watch out for surveys that try to manipulate responses by providing narrow options. For example, a survey asking the question “What is your favorite ice cream flavor?” should have more options than simply “coconut” and “mint;” from those findings, we might fallaciously conclude that 78% of people identify “mint” as their favorite ice cream flavor.