The Data Dictionary contains the definitions of all fields that may be included in the client’s databases. Setting up the Data Dictionary is usually the first task to perform when beginning work on a new client. The choices you make when setting up your fields will affect every later step in your project, so be sure to consider your options carefully. Use logical field names and ample descriptions. Be sure that you select the appropriate type for each field. Finally, when establishing coded fields, make sure that you pick appropriate labels for each code.

Fields set up in the Data Dictionary need not have exactly the same structure as in the external data source from which they are to be imported. The import process itself can be used to make certain transformations in the data as specified in the record layout. The Data Dictionary describes how the data will be represented once it has been loaded into ProVal.

Field name must start with a letter and may contain letters, digits, and underscore, but no spaces or other special characters. The capitalization will be retained, but later use of the field is not case-sensitive.

Description can be a phrase of any length. All characters are permissible. Include key information about the contents of the field.

Field type may be any of the five available choices. The rest of the dialog box may change as the field type changes, since special parameters are relevant to certain field types. The available field types are

• Numeric

• Character, a special field type that is used when its contents are alphanumeric data that will not be used in numeric calculations. If you need to use character data for computation, define the field as a coded field instead, assigning (when the coded field is defined) a unique numerical value to each character set that can appear in the field. For example, a field containing a plan participant’s name may be of the character type, but a field containing the division the participant works at almost certainly should be a coded field.

• Coded, a special field type that is used when the field may contain only one of a set of permissible choices. These choices are then identified with character labels and codes (see below). Fields used for certain purposes are required to be of the coded type. Examples of these purposes include participant status, inactive payment form in pension modes, and subtotal breaks. Other fields that, typically, are of the coded type identify a characteristic of a record (for example, an active record hired before a certain date) or a “bucket”, such as division or plant location that the record is to be associated with. In OPEB mode, a coded field, typically, is used to indicate whether a record has spouse coverage (and thus represents two lives, rather than one).

• Date

• Social Security number, a special field type that displays with the format of (U.S.) social security numbers. Although ProVal will recognize and understand this format, it does not require that social security number data be provided with leading zeroes or with hyphens. As is the case for character fields, the data in a social security number type of field cannot be used in numeric calculations.

Column title will be displayed centered above the field in tabular output. The default value for the column title is the field name itself.

Formatting style contains a template for the display format for a numeric or social security number field. For numeric fields, this style determines the width, number of decimals and presence of dollar signs and commas. The numeric data itself is retained to full precision, however, regardless of the formatting style. For social security number fields, formatting style determines how the field is displayed in the spreadsheet view of the database and output. Formats may include "9" (to represent display of actual digit), "X" (to display an X in place of the digit) or "-" (to provide a visual delimiter). The ability to conceal digits enables the user to avoid having full social security numbers appear in printed output. (For example, use of the formatting style XXX-XX-9999 will display only the last four digits of the social security number.) Obscuring social security number digits does NOT make the data secure, however, in the sense that the user can still access the full social security number in various ways. In Spreadsheet Edit, field edits of a social security number will reveal all digits without regard to formatting, as will copying data from Spreadsheet Edit and pasting it (into Excel, for example). Export Data ignores the format and displays the true unformatted number (i.e., the entire number is displayed). List Data uses the Data Dictionary's formatting style as the default; but even if the Data Dictionary's format obscures digits, the formatting may be modified, using the Layout Options button, to display full social security numbers. Note that the dashes displayed in the format examples below the text field in which you enter the formatting style are decorative and thus need not be included in the field value on the database record.

Display width determines how many character places are allotted for the display of a character field. Although this parameter affects the output format, the data itself is retained without any limit on field width.

The entry of Labels and codes is required for coded fields. The character labels consist of character phrases of any length – for example, the names of different locations or divisions. Associated with the labels are numeric codes – often just 1, 2 and so on. There is no limit to the number of code and label pairs that may be specified.

The Sort button allows you to set the order in which the codes and labels are displayed. The available choices are alphabetical, sort by code and custom sort.