1.2. Using UNISTAT User’s Guide
UNISTAT Users Guide is supplied as a single PDF document. Context-sensitive help is available at all times, providing instant information on the currently selected procedure. Where relevant, the algorithm used in the procedure is given and examples are provided. The data for these examples are supplied in example files so that results can be easily reproduced by the user. Section titles of procedures which are not included in Light Edition are shown in dark blue font.
1.2.1. Conventions and Notations
All references to computer keys are enclosed within less than (<) and greater than (>) signs. For instance <2> refers to the number key 2, <f2> to the function key 2, <Enter> to the enter (or return) key. If two keys are to be pressed simultaneously, a plus (+) sign will separate the keystrokes. For instance <Shift> + <Tab> means pressing the shift and tab keys simultaneously. Square brackets ([) and (]) are used to enclose button labels in dialogues. For instance, [Variable] refers to a button on the Variable Selection Dialogue (clicking on which selects the highlighted variable for analysis). The underlined character is the hot-key for this button. Optional parameters that can be appended to some functions are also enclosed in square brackets. The hyperlinks are in blue colour.
Menu selections are represented in bold sans serif font. Multi-level menu selections are separated by an arrow. For example, Graph → 2D Plots → X-Y Plots means clicking on Graph, selecting 2D Plots from the pull-down menu and X-Y Plots from the daughter menu.
References to the sections of this guide, and thus references to all graphics and statistics procedures are in sans serif font, such as X-Y Plots. References to prompts, messages, dialogues and all other aspects of the user interface are in narrow sans serif font, such as Syntax Error. UNISTAT’s own spreadsheet functions are represented in bold gothic font such as HCoSec(). References to variable names are in generic serif italic font and the file names are in upper case.
In order to avoid unnecessary repetitions, information common to more than one section is given under a higher level section. For instance, to find out how to perform a Mann-Whitney U Test, one can start from section 18.104.22.168. Mann-Whitney U Test. However, it will help to have a look at section 6.4.1. Unpaired Samples, 6.4. Nonparametric Tests with One or Two Samples and 6.0. Overview as well. Usually, the types of data used in procedures are explained at the beginning of each chapter.
1.2.2. Reproducing the Examples
UNISTAT’s help system and User’s Guide contain a large number of examples many of which have been published in commonly used books. Data sets used in the examples can be found in the UNISTAT example data files.
The data files can be accessed from Tools → Example Files. Each of these files contains data for more than one example, but in general the data for procedures under the same section are grouped within the same file. For instance, the file PARTESTS contains the data for Parametric Tests and REGRESS contains the data for Linear Regression examples.
Examples will usually make a reference to the book or article in which they were first published, followed by the name of the file containing the example data. Where a suitable published example cannot be found, an example will be provided using the generic DEMODATA file. The user will then be told which columns are to be assigned to which tasks. When all the instructions are followed, it should be possible to precisely reproduce the results given in this manual. These may then be compared with the results in the original published source for validation purposes. In most cases, UNISTAT output would include many more statistics than were given in the original source. Also, UNISTAT’s results will generally be more accurate, as the examples in books often display the intermediate results leading to rounding-off errors.
We believe that by including published examples in this User’s Guide we not only provide a valuable tool to learn UNISTAT by solving real-life problems, but also provide the user with the possibility of verifying and validating the accuracy of UNISTAT’s results by comparing them against respected academic sources.