Bunuel
It appears that the number of people employed by a typical American software firm decreased in the 1980s and 1990s. This trend is borne out by two studies, conducted 20 years apart. In a large 1980 sample of randomly chosen American software firms, the median size of the firms’ workforce populations was 65. When those same firms were studied again in 2000, the median size was 57.
Which of the following points to the most serious logical flaw in the reasoning above?
A. The median number of employees in American firms in many industries decreased during the 1980s and 1990s.
B. During the 1980s and 1990s, many software firms increased the extent to which they relied on subcontractors to write code.
C. The data in the studies refer only to companies that existed in 1980.
D. The studies focused on the number of employees, but there are many ways of judging a firm’s size, such as revenues and profits.
E. The median number of employees is not as sound a measure of the number of employees employed in an industry as is the mean number of employees, which accounts for the vast size of the few large firms that dominate most industries.
Key focus areas:1)
Conclusion: General trend was median number of employees decreased in an industry.
2)
Evidence 1: A LARGE sample of companies operational in 1980 showed median size to be 65.
3)
Evidence 2: The SAME SAMPLE was used and it showed median size to be 57.
4)
Key words: LARGE, THOSE SAME FIRMS, MEDIAN, TYPICAL AMERICAN SOFTWARE FIRM.
All of the above helps us connect to the correct answer and become sure of the choice.
Question type: Logical flaw
A.
Many industries - GMAT doesn't like CR answers where another Geography, Location or Industry is used an evidence to support a certain Geography, Location or Industry in question. Having said that, this option agrees with the conclusion by saying the trend was observed in other industries also so there is no flaw. We know there is a flaw somewhere, we need an option which helps find that flaw. So this is incorrect.
B -
MANY - refers to more than one but not most/majority. Knowing that many firms used contractors instead of employees doesn't explain why the trend decreased.
The trend could only have decreased if it came from more than 50% i.e. Majority of the companies not from MANY companies. So this option is incorrect.
C - Only to companies that existed in 1980s =
THOSE FIRMS = We can see that one of the keyword does match this option. If the
LARGE SAMPLE of companies in 1980s consisted of 100 companies and if let's say 70 companies went bankrupt then only 30 companies remained by 2000 and so those firms would refer to those 30 companies. Now the study done in 2000 actually used a
SMALLER SAMPLE = opposite of the study in 1980 = which means a variable changed and this change showed a wrong picture of decreasing trend.
It would be wrong to GENERALISE the results of the two studies when 2nd study had a very small sample size (not representative of the industry). Therefore, this option clearly points out a logical flaw.
D - An Alternative Method - An option suggesting a better method/approach is usually not a correct answer.
The logical flaw in the study is not pointed out here, this option is just saying there was a better way to do it. It is not saying it was the right way to do it. Therefore, this option is incorrect.
E - Not the right method - This option does exactly what we asked for in D. It say Median is not the correct method, mean is. But the reasoning it gives is where this the option goes wrong. It say mean is better suitable for LARGE SAMPLES. But if you paid attention to option C, you would have realised that
LARGE SAMPLE became SMALL SAMPLE over the 20-year gap.
If that is true, then Mean wouldn't have been the correct method by its own reasoning. Therefore, logically this option is incorrect.
Therefore,
Option C is the correct answer. Posted from my mobile device