Sign test is a non parametric test, that is used to test the null hypothesis and whether or not two groups are equally sized. It is based on the direction of the plus and minus sign of the observation, and not on their numerical magnitude. It is also called the binominal sign test, with the null proportion of .5. Sign test is also used in the case of the paired sample, which is called the alternative test of the paired t-test.
Types of sign test:
Sign test in case of large sample:
Binominal distribution formula = with p =1/2
Available in nonparametric tests, the following steps are involved in conducting a sign test in SPSS:
Select the first paired variable and drag it to the right side in variable 1, and select the second paired variable and drag it to the right side in variable 2. Select the “sign test” from the available test. Click on “option” and select the “descriptive statistic” from there. Now, click on the “ok” button. The result window for the sign test will appear.
In the result window, the first table will be of the descriptive statistics for sign test. These will include the number of observations per sample, the mean, the SD, the minimum and the maximum value for sign tests in both samples. The second table shows the frequency table. This will show the number of negative sign, the number of positive sign for the number of ties, and the total number of observations. In SPSS, the following table will appear for the descriptive table and frequency:
The third table will show the test statistics table for sign test. This table shows the value of Z statistic and the probability value. Based on this probability value, we can make our decision about the hypothesis. For example, if the probability value is less than the significance level at 5%, null hypothesis will be rejected. If the probability value is greater than the significance level, then we will accept the null hypothesis. The following table will appear for the test statistics:
*Click here for assistance with conducting the sign test or other quantitative analyses.