A hypothesis testing (hypothesistesting): also known as statistical hypothesis test is used to determine the sample, the sample is the difference between the sample and the overall error caused by a difference in kind or statistical inference method is essentially caused by the difference.
2, a significant test: Hyun-Soo test a hypothesis, testing the most common method, but also a basic form of statistical inference, laundry basic principle, an assumption is made prior to the general characteristics through statistical sampling study reasoning, this hypothesis should be rejected or accepted the judgment.
The so-called reductio ad absurdum, it is first of all, assuming that the null hypothesis is correct, then accordingly get some news, get the message accordingly if known and contradictory messages (generally based on the principle of small probability) then the null hypothesis can explain the lack of evidence established, or reject the null hypothesis (that the statement is not strictly statistically).
Small probability event, once again test almost does not occur ヽ (¯д¯;) Techno
However, the small probability event is not impossible, but the probability of occurrence is very small, we can not completely rule out the possibility
* Mean
(1) open the data file 'HourlyWage.sav', click (analysis) - (Compare Means) - (mean). The 'hourwage' opt-in due to the variable list, it 'yrsscale' opt-in argument list.
(2) Click (option) button.
(3) is selected from the range, the mean standard deviation, median within a group. Click (OK) tick 'Anova table and eta' and 'linear correlation test' click (continue) and click (OK)
(4) in a layer 1, click (Next) the 'position' variables into the layer 22, the argument list, the default value other reservations. Click (OK)