 This draws a histogram of the distribution of data in the selected
block, including the mean value (red line) and values of skew and kurtosis. Optionally, you can display the median and an approximate normal distribution curve, based on the mean and SD.
- Skew is a measure of the symmetry of the distribution. A normal (symmetric) binomial distribution has a skewness of zero. Very approximately, negative skewness indicates a distribution in which the left 'tail' (relative to the mean) is longer than the right tail, while positive skewness -- as in this example -- indicates a longer right 'tail'.
- Kurtosis is a measure of how 'bunched' or 'spread out' the data are; i.e., the 'tailedness' of the distribution. A univariate normal distribution has a kurtosis of 3; here this is normalized to zero. Again very approximately, negative kurtosis indicates a more compact distribution than normal ('tails' less spread out), while positive kurtosis -- as in the example -- indicates data are more spread out than a normal distributon (longer 'tails').
Important: there are many exceptions to these generalities; see a statistical text to more fully understand skew and kurtosis.
You can adjust either the number of bins (to a maximum of 500; the default is 40) or the bin width, using the textfields and associated buttons. Note that using a large number of bins on a small dataset may give odd results.
The 'Print' button to send an image of the window to a printer or a pdf file.
|
