论文部分内容阅读
随着收视率数据在各电视台广告经营和媒介管理中的地位日益提高,对收视率数据本身的争议和质疑也日益增多。抽样调查得到的收视率是否可信?收视率有没有客观性?几个不同来源的数据有差异的时候,谁更可信?(哪个更准?)样本量越大就越好吗? 这些问题的解释涉及到社会学研究方法,以及统计学常见原理的一些具体应用等方面。下面有针对性地加以说明。一、为什么要使用抽样调查的数据?抽样调查的数据准吗? 和普查相比,抽样调查在人力、时间和经费方面拥有绝对优势自不待言,如果数据有较强的时效要求,如收视率数据,更是只能依靠抽样调查来得到。这里想特别说明的是,一个良好的抽样调查,虽然引入了抽样误差,但是由于样本量小、样本目标明确,可以很有针对性地提高执行质量,减少非抽样误差,反而能够减少甚至避免执行误差。例如,对抽样中的高
With the increasing popularity of ratings data in various television stations’ advertising and media management, there is increasing controversy and doubts about the ratings data itself. Is there any credibility in the sample survey? Is there any objectivity in the ratings? Who is more credible when there are differences in data from several different sources? (Which is more accurate?) Is the better the sample size? The interpretation involves sociological research methods, as well as some specific applications of statistical common principles. The following targeted to illustrate. First, why should we use the sample survey data? Sampling survey data? Compared with the census, the sample survey has absolute advantages in terms of manpower, time and funds, it goes without saying that if the data has strong aging requirements, such as the ratings Data, it can only rely on sample surveys to get. What should be specially mentioned here is that a good sample survey, though introducing sampling error, can reduce or even eliminate the implementation of sampling due to the small sample size and the clear objective of the sample, which can improve the execution quality and reduce the non-sampling error in a targeted manner error. For example, the height of the sample