Dear Stefaan:
Good day!
To calculate the standard deviation, you would calculate the variance and then square root it. The first statistical argument of having a variance is that we need to calculate the standard deviation.
Variance is the statistical measure of calculating the dispersion of the data. However since variance is in the squared units, it doesn’t make much sense when using the measure practically. Hence we square root it to get a measure in the same units as of the data. For example variance of a sample of lengths measured in meters will be in squared-meters; however the standard deviation will be in meters which make more sense and helps in calculating the ranges.
Variance is calculated by computing squared deviations (from the mean) for every observation and adding them up together. If the deviations were not squared, the result would have been zero. However, since this process results in measurements in squared-units, we take a square root of it to bring it back to the original units.
For a detailed debate, you need to see text on statistics. If you don't want to go deep in statistics, then please remember that calculating variances is a statistical obligation.
Regards,
Khurram