Additional Tests of Hypothesis
In the previous chapter, we presented methods for pictorially representing a set of data values and determining whether the data follows a known statistical distribution. We stated that testing the data fell under a general category of tests of hypothesis, of which the chi-square goodness of fit test was one example. In this chapter, we discuss a few more tests of hypothesis that occasionally are of benefit to the data miner. These tests involve comparing the parameters (such as mean and variance) of two or more samples, or comparing one sample’ s parameters to a known population measure. The procedure for conducting each test is much the same as that for the goodness of fit situation, so we review that for you and add a few more interesting details that will provide good party conversation, should you find yourself sipping cocktails with your local statistician.
KeywordsInternet Usage Column Total Statistic Text Monthly Sale Population Standard Deviation
Unable to display preview. Download preview PDF.