The median is the observation in the middle of a continuous distribution that divides the distribution into equal halves. Exactly 50% of the observations in the distribution should be above and below the median. If there is an even number of observations in a distribution, the two observations in the middle are averaged to obtain the median. Medians play an important role in understanding the nature of skewed distributions and non-parametric statistics like one-sample median test, Mann-Whitney U, Kruskal-Wallis, Wilcoxon, and Friedman's ANOVA . The results of non-parametric statistics using ordinal or non-normal outcomes should be interpreted in the context of medians and interquartile ranges.
1. The data is entered into a column of a database.
2. Click Analyze.
3. Drag the cursor over the Descriptive Statistics drop-down menu.
4. Click on Explore.
5. Click on the outcome or variable to highlight it.
6. Click on the arrow to move the ordinal variable into the Dependent List: box.
7. Under the Display table, click on the marker for Statistics to select it.
8. Click OK.
2. Click Analyze.
3. Drag the cursor over the Descriptive Statistics drop-down menu.
4. Click on Explore.
5. Click on the outcome or variable to highlight it.
6. Click on the arrow to move the ordinal variable into the Dependent List: box.
7. Under the Display table, click on the marker for Statistics to select it.
8. Click OK.
1. In the Descriptives table, look for the Median row. This is the median value that is interpreted.
Click on a button below to continue.