BioNinja

Measures of dispersion (or spread) describe how scattered or clustered a dataset is around a central point

The degree of spread reflects the precision of the data – the less variability, the greater the precision
The data spread can be represented graphically via error bars (the larger the error bar, the lower the precision)

The measure of dispersion employed is dependent on the type of data used:

Range – Used when there are less than five replicates within a data set
Quartiles – Used when the data set is skewed and the central point is measured by the median
Standard deviation – Used when a data set is normally distributed and the central point is shown by the mean
Standard error – Used to measure the variability of many sample means around a population mean

The precision of a data set is influenced by random errors (unpredictable variations in the measurement process)

Random errors create inconsistent displacements of data values and cannot be avoided (they are irregular and indiscriminate)
Random errors can be limited by keeping all extraneous conditions constant (i.e. controlled variables)

Outliers are any results that deviate by a significant margin from all other values (and should be removed from data processing)

For normally distributed data, a value is considered an outlier if it is more than three standard deviations outside of the mean
For skewed data, a value is considered an outlier if it is outside the median by more more than 1.5 times the interquartile range

Normal Distribution