Thank you for trying DataScope! Any questions are welcome on CompuServe ID 100263,574 and fax number +36 62 325 928. This demo version can be distributed freely. --------------- 1. Installation --------------- Simply copy all files into one directory. Then create an icon for DataScope in Program Manager, or use the File/Run command to start it. --------------- 2. Introduction --------------- The short description below introduces DataScope's main fea- tures, thus helping you to evaluate the software. DataScope is a Microsoft Windows 3.1 application which helps you analyse your database in a very efficient manner. The software graphically displays the contents of a database and provides incredible tools to explore the relations between the records. For further analysis, it also allows you to create new databases with specific selections of records and fields. DataScope can be used immediately without formal training. It is an excellent tool for managers, their staff, consultants and scientists. Operational, financial, customer, product, scientific and other databases can now be interrogated to unveil hidden business challenges and opportunities, meaningful relationships leading to new scientific con- clusions, and more. Some of the most important features of DataScope: - The DataScope system attaches itself to any database through Microsoft ODBC, and checks its contents. It classifies the various database fields to three categories: 1. Identifier fields. This category helps in identifying the records. An identifier field is usually a name or other unique ID. 2. Number fields. These fields contain quantities, such as age, population, price, etc. DataScope is also able to handle numeric fields where no data available for all records. 3. Discrete fields. In this category, fields usually have only a few different values that are used not for quantitatve analysis but record classification. Examples are fields which contain type identifiers, yes/no answers, and so forth. - DataScope displays the database in up to 16 different windows. Each windows represents one or two database fields, so you can analyse various alternatives using 16 different fields or field pairs. All windows are related to each other in that they all show information about the currently selected alternative. The mode of display depends on the field category: 1. Identifier fields are displayed as a list. The list con- tains the contents of the identifier field, and, also, the contents of the field(s) in the active, i.e. selected window. The list is automatically sorted based on the field in the active window, so you can examine the data sorted by any of the fields by simply activating another window. Changing the sorting order does not take time at all, because of DataScope's special indexes that it cre- ates in a preprocessing phase. 2. Number fields are represented by their empirical distribution function. The value that this function gen- erates for a selected data element is a number between 0% and 100%. The percentage shows you where the selected data element is situated relative to the others and, thus, allows you to examine one value relative to another very easily. Suppose you have a database containing car price/ performance data. If the value generated by the function based on pricing data amounts to 90%, then the car falls in the top range of all list prices. 3. Discrete fields are displayed as traditional charts, such as the pie chart or the bar chart. Charts show the cate- gory to which the currently selected data element belongs. - Relational diagrams can be used to display information about the relationship between two numeric fields. These diagrams are useful when examining trends and if you want to point out exceptional features. The more a point (representing a specific data element) deviates from the average, the farther it is from the diagonal line of the diagram. A re- lational diagram can also be well used to examine whether a better alternative exists. Many different relational diagrams can be displayed. If you see an exceptional data element in one, you can click on it to determine whether it is exceptional in other windows as well. The 16 windows that can be opened show the selected data element from various angles of view. The identifier window shows its identifier, the numeric windows show its quantitative characteristics, the discrete windows show the categories to which it belongs, while the relational windows show the trends it might follow. - DataScope is not only suitable for analysis of individual records. In any window (i.e. for any feature), you can select a group of records. These selections are independent from each other and are called local selection. A global selection can be created from the union ('or' operation) or the intersection ('and' operation) of the local selections. In this way, you can select records satisfying complex conditions without having to type in any commands or formulas. For example, you can select the high-performance and low-consumption cars in a database with car-related data. The simultaneous update of the windows causes all windows to show globally selected records. When you globally select a group of alternatives, the identifier window helps you to identify them, numeric windows show their distribution relative to all other records, discrete windows show their distribution among the categories and relational windows show the ones with exceptional characteristics. - You can create a new database from the globally selected records. In this way, you can quickly filter out the uninteresting records when preparing for a decision. - Record markers can be assigned to records you wish to study in more detail. These markers appear in all diagrams, showing the position of the record at a glance. - DataScope imports data through the Microsoft Open Database Connectivity (ODBC) standard. This standard makes possible to access all popular database types. Drivers are delivered for text files, dBase, Excel, Paradox, MS-Access, FoxPro and SQL-Server databases. Additional drivers can be obtained from Microsoft and manufacturers of database systems. - With DataScope, the emphasis is on the graphical display of data, which makes quantitative information more understandable for people. The numeric information is also preserved, so DataScope also replaces traditional database query systems. Typed commands in DataScope are not needed. You just need to use your mouse (and your intuition, of course) while you explore. - The instant update of the windows allow you to examine a database from many angles of view. You can select database elements with certain features and check how they relate to others. - The unique features of the distribution function used to display numeric fields allow you to perform a context- dependent analysis. How good is a car that has a performance of 80 HP? Can you tell whether the female life exceptancy in Germany is high or low compared to the other countries? Is one point increase for a chemical company on the Stock Exchange high or low? Generally to answer these questions you need an expert or a significant time investment to evaluate the statistics. This is not necessary any longer. Now, you just need DataScope!