Argue之第一宗罪---调查类问题(Survey)
第一类:调查或统计的程序性问题(Procedure)
(1) 选择性样本(Selective sample)
In the … place, the author provides no evidence to support the claim that … as a whole …. The example cited, while
suggestive of these trends, is insufficient to warrant their truth because there is no reason to believe that the data drawn
from this …is representative of the entire country. For example, if …, it would clearly be unrepresentative. The reason for
this is obvious. …
(2) 样本的数目及代表性问题(Quantity of the sample. Is the sample large enough to be representative?)
In the … place, the conclusion that … is based on too small a sample to be reliable. The only evidence offered in support
of this conclusion is the fact that A …. Unless it can be shown that A is typical of all …, the fact that … is no ground
for concluding that all ….
(3) 调查对象是否能够坦诚
In the … place, we are not informed whether survey responses were anonymous or even confidential. If they were not, then
respondents might have provided responses that they believed …, regardless of whether the responses were truthful. In this
case, the survey results would be unreliable for the purpose of drawing any conclusion about …, let alone about ….
(4) 调查对象是否被暗示或误导了(Loaded questions?)
In the … place, we are not informed whether the survey required that respondents choose … between alternatives. If it did,
then the results might distort … of the respondents, who might very well prefer a … choice not provided for in the survey.
第二类:调查的结果是否被正确表述和使用(Results of the survey):
(1) 统计数据是否被精确表述(Are the statistics misleadingly vague?)
在职者类问题可以从三个方面着手:论者所说好的东西未必是好的,所说不好的未必是不好的; 忽略了其他指标(GDP capital、deficit、welfare…); 忽略了其他原因(National Economic Climate、前任政策得力)。
模版: In the … place, the author are meant to strengthen his allegation by using some statistics. But these statistics are vague
and oversimplified, and thus may distort … overall … picture. For example, …. Moreover, …. Finally, the survey/poll
indicates that …, but fails to indicate ….
(2) 回应者是否具有代表性(Respondents)
In the … place, there is probable problem of biasness in the survey. Perhaps … who are interested in … were more likely
than other … to respond to the questionnaire---possibly because they found the questionnaire more interesting. Or perhaps a
large number of respondents pretended to be interested in … issues because of some wrongful reasons. In either event, the
survey results would be virtually useless in drawing reliable conclusion about ….
(3) 调查是否具有时效性(When was the survey conducted?)
In the … place, the editorial/article/analysis neglects to indicate how recently the survey was actually conducted. When
samples are used to make general claims about a particular group, the samples should be close enough in time to the
generalization they are used to support, so that historical changes will not invalidate the generalization. All we know in
this editorial/article/analysis is that the survey was …. The less recent the survey itself, the less reliable the results
to indicate …. |