Defending Dignity A Manual for National Human Rights Institutions on Monitoring Economic, Social and Cultural Rights In a spreadsheet, you can do this using the formula =STDEV Say you have the following data on unemployment rates as a percentage of the total active population for four different municipalities: 1, 2, 3, 4 • What is the standard deviation of your dataset? • The mean is 2.5, so following steps 1 to 3 Value Difference to mean Squared difference 1 -1.5 2.25 2 -0.5 0.25 3 0.5 0.25 4 1.5 2.25 1. Sum the square of the differences between the values and the mean = 5 2. Divide that sum by the number of values minus one = 5/(4-1) = 5/3 3. Take the square root = 1.291 If the data is normally distributed (all data points are evenly distributed either side of the mean) then 68.27 percent of the data points will fall within one standard deviation from the mean and 95.45 percent of the data points will fall within two standard deviations from the mean. The bigger the standard deviation value, the more variation there is in the dataset. Using the unemployment example, a standard deviation of 1.291 means that 68.27 percent of the data points will fall within this distance from the mean (assuming a normal distribution). This suggests that there may be reasonable geographical variation in unemployment so reporting the mean alone could be misleading. The median absolute deviation is similar to the standard deviation but is used with the median rather than the mean. The following two steps show how to calculate it: 1. Calculate the median and the absolute differences between each value and the median. 2. Calculate the median of the differences. To take the same unemployment example as above but with data on one additional municipality, your data is 1, 2, 3, 4, 5 1. The median is 3 and the absolute differences are 2, 1, 0, 1, 2. 2. Order the differences 0, 1, 1, 2, 2 and the median absolute deviation is 1. 8.3.5. Normalization Once you have an idea of the data you are dealing with, you might begin to make comparisons. For instance, in an investigation on the extent to which the right to health is being fulfilled, you might want to compare government spending on health in your country with that of another country that differs in a number of aspects. Comparing the total value of government spending on health in this case will only tell you so much if, for example, your country is very large and the country you are comparing it to is very small. The bigger country will most likely spend much more on health in total than the smaller country. Does this mean the bigger country is meeting its obligation to fulfil the right to health better than the smaller country? Not necessarily. To answer that question, the two countries have to be compared on an equal basis. This is usually done using an indicator that tells us how big a country is; often the size of its population. 92