DataLab is a compact statistics package aiming at exploratory data analysis. Please visit the DataLab Web site for more information....


Univariate Descriptors

Command: Math -> Univariate Statistics -> Univariate Descriptors...

In order to calculate the basic univariate descriptors of a dataset, you may issue the command Math/Univariate Statistics/Univariate Descriptors.... After issuing this command, the calculated parameters are displayed in a table. The descriptors can be calculated either of selected variables (columns of the data matrix) or of marked data.

In order to calculate the variable parameters you have to tick off the corresponding variable names in the list of variables at the top of the window. Alternatively, for calculating the univariate parameters of marked data, you may select "Marked Data". This results in the parameters of both blue and red markings, and the union and the intersection of the marked data.

 

The following table lists the calculated univariate descriptors:

  • number of selected data
  • minimum and maximum
  • sum of selected data
  • mean +/- confidence interval
  • standard deviation of the sample (+ the standard deviation of the population in parantheses)
  • variance of the sample (+ the variance of the population in parantheses)
  • standard error of the mean
  • median and quartiles
  • mode (please note that the mode is calculated by a Gaussian density model, while the bars in histograms are created by binning the data; thus there can be differences between the calculated mode and the histograms, depending on the resolution of the histogram)
  • skewness; the skewness is followed by the p-value for the null hypothesis that the distribution is symmetric
  • kurtosis; the kurtosis is followed by the p-value for the null hypothesis that the distribution is mesokurtic

The confidence level of the confidence interval of the mean can be selected in the top region of the window (the drop-down box offers the most common statistical securities between 90% and 99.99%.

In addition to the calculated parameters both the frequency tables and the box plots are displayed on separate tabs.