CHRR
NLS
Database Investigator
NLS
CD-ROM
Users Guide
December 1999
File size: 540.25 Mb
Total files: 194
This User's Guide illustrates how to begin using the Windows version of the data base access software NLS Database Investigator. This retrieval system accompanies one or more of the National Longitudinal Survey's (NLS) databases on CD-ROM. With it you can extract data and prepare that data for statistical analysis. You can create, modify, and save your own personalized, extracted files for further research. This package is easy to use. If you already have some knowledge of the NLS and a search strategy in mind, you will be able to complete that search and extract a file within minutes of using the guide. Just follow this guide and substitute your own strategy for the examples given.
To get the most out of this User's Guide, you should already be familiar with the NLS and know how your interests can be served by NLS data. You need some idea of the samples covered and the types of information collected by the survey instrument. If you have questions about the NLS data, please consult the NLS Handbook (available in printed edition only). Or you may consult the on-line documents available for the cohort selected. These documents are available on the CD via: Documents in the Contents window on the initial screen display of NLS Database Investigator shown in Figure 5. For the sample cohort they are the NLSY79 User's Guide and the NLSY79 Codebook Supplement. For more detailed information, see Chapter 3.
In addition to knowledge of the NLS and its data, you need to have some knowledge of the MicrosoftÒWindowsÔ operating system. This guide explains in detail how to install and use the Windows version of the NLS data extraction software.
Before you install or start the NLS software make sure you have the following minimum hardware:
1. A PC running Windows 95, Windows 98, or Windows NT.
2. For maximum performance, 32 megabytes of memory are recommended.
For data installation on your PC the space needed on your disk drive may vary. Depending on the cohort you select and the type of installation you choose (complete, optimal, minimal) during installation, the NLS Investigator Installer will display the space needed and calculate the disk space available. (Refer to the installation window shown in Figure 1.) If you choose a data installation option that is too large for your disk drive, it will alert you to that fact and not let you begin the data installation. At that time you may choose a different data installation option, or switch disk drives to one with more capacity. Copying the entire CD-ROM to a disk drive on your PC will greatly enhances performance because the data transfer rate on such a device it is typically faster than that of a CD reader.
The capacity of your PC's hard disk is the factor limiting the size of the extract you may design. The amount of disk space required for a given extract is automatically calculated for you during your selection session so you can make sure you have adequate space available before executing the extract. See File Size Information, page 13.
The following conventions are used throughout this guide.
<Enter> Anything enclosed in angle braces represents a key on the keyboard. Find the corresponding key and press it.
<Alt> + <F1> The plus (+) sign between two keys means hold down the first key and press the second key. Then release both keys.
bold Anything that you are asked to type exactly as it appears in the manual will appear in bold type. For example, if you are asked to type NLSDBA, it will appear in bold. The software is not sensitive to upper or lower case.
italics Anything that is a placeholder for variable information will be written in italics. For example, if you are asked to provide the name of a file it will appear as filename, a letter for a disk drive as d:.
Compact discs should be handled with care. Do not touch the surface. Hold the disc by the outside edge and gently place it in the CD reader. It should be stored in its plastic case. If the surface becomes scratched or damaged, the CD may become unusable. Never wipe the surface with cloth; it may be air-cleaned.
When a new year of data is added to the NLS database, a new compact disc may be purchased along with the current software. As the software is refined and updated, new software versions will be issued with the next available data update. Any major errors in the search software will be released on PC diskette as soon as they are discovered. Whenever you receive any updated version of the data or software, you must re-install the new versions.
If you have any customer service questions or technical support problems please contact:
· The Ohio State University
· Center for Human Resource Research
· 921 Chatham Lane, Suite 200
· Columbus, OH 43221-2418
· 614-442-7366 "User Services"
E-mail: USERSVC@postoffice.chrr.ohio-state.edu
To install the NLS database software on your PC, you must install the software and the data in two steps. In order to install the data you must decide among one of the following three options.
1. Minimal Installation - approximately 80 MB. Installs the software and the required documentation on your PC. The detailed documentation (Codebook) and data files remain on the CD-ROM, so you will need to have the CD-ROM in your drive to access the data and detailed documentation. Most browsing performance will be the same as with the above options, but viewing the codebook and extracting performance will be determined by the speed of your CD_ROM drive.
2. Optimal Installation - approximately 200 MB. Installs all documentation files onto your PC. The data files remain on the CD-ROM, so you will need to have the CD-ROM in your drive to access the data. Browsing performance will be the same as with the "Complete" option, but extracting performance will be determined by the speed of your CD-ROM drive.
3. Complete Installation - approximately 499 MB. Install the software, data, and documentation on your PC. This requires the most space and yields the best performance for browsing and extracting.
To install the NLS Database Investigator you must start the installer program (installer.exe). To start the installer, insert the CD in your CD-ROM drive. Click Start/Run and type drive:\installer.exe, where 'drive' is the location of your CD-ROM. Or you may use Windows Explorer, open your CD-ROM directory and double click on installer.exe.
When the Installer window opens (Figure 1) it will be blank. You must install the software and the data in two separate steps. First, in the '1. Install Software' portion of the window, click on Install NLS Investigator to begin installation. Follow the screen prompts.
Next, in the '2. Install DATA' portion of the window:
1.
Choose the Cohort
from the drop-down menu 
;
2. Select one Data Installation Option
3. Then, type (or Select to browse) the path of the Data Directory.
4. Click the Install Data button to copy the data files.
Depending on the cohort you select and the type of installation you choose (complete, optimal, minimal) during installation, the NLS Investigator Installer will calculate the disk space available. If you choose a data installation option that is too large for your disk drive, it will alert you to that fact and not let you begin the data installation. At that time you may choose a different data installation option, or switch disk drives to one with more capacity.
Help contains detailed instructions that will appear in the large box in the window.
To run the NLS Database Investigator from Windows Explorer, open the drive and directory where you installed the software (default C:\Program Files\CHRR\) and double-click on dbgator.exe. Or, click Start/Programs/CHRR Software/DB'gator.
When you start the NLS Database Investigator software, dbgator.exe, a screen similar to Figure 2 will appear. The first time it appears it will be blank and you must select the source (Cohort) of the data set you wish to use.
To select the data set, use the drop-down menu next to the Cohort field.
When the data set has been selected the Contents window will display the Documentation and Indexes that are available with the data set. From the Contents window, you can control what appears in the other windows by selecting items in the tree.
You may resize the any of the windows to suit your needs. If you loose track of one of the windows (Contents, Variables, Text/Cdbk), go to the Window menu and choose Default. The screen will refresh its format similar to that in Figure 2. The contents of the individual windows will not be lost.
Before specifying the data set, it may be necessary to say a few words about the layout of the NLS Database Investigator screen. The screen is a typical Windows environment with the Command Menu bar just below the Title bar. The register of commands available for each menu components: File, Window, Extract, Data, Option, and Help can be found by clicking on the menu component A complete list of commands under each of these components can be found in the Appendices, Main Menu Bar.
Just below the Menu bar is the 'Navigation' bar.
|
Figure 3 The Navigation Bar |
It contains the Cohort field and three buttons, Contents, Variables, Text/Cdbk (Codebook). The data set to be used must be specified in the Cohort field. The Contents, Variables, and Text/ Cdbk buttons switch between the active windows displayed in the lower portions of the screen. These windows are used to manipulate the data for extraction (Contents), display the results of manipulation and extraction (Variables), and obtain help in the form of documentation (Text/Cdbk).
In a Windows environment switching active windows can also be accomplished simply by clicking in the window. However, if the active window 'covers' one of the inactive windows or the inactive window is minimized, the above buttons can be used to make the window active and bring it to the forefront. If you loose track of one of the windows (Contents, Variables, Text/Cdbk) and its button does not bring it to the forefront, go to the Window menu and choose Default. The screen will refresh its format similar to that in Figure 2 but the contents of the each window will not be lost. You may now resize.
The windows in the CHRR NLS Database Investigator may be resized according to your preference. You must decide for yourself what Windows layout you prefer. Once your preferences are established they will remain that way rather then reverting to the default. It may be easiest to begin with the Contents window maximized in order to select an index for use, locate index terms, and select the terms for the search strategy. Or you may wish to review the on-line, hypertext documentation first in the Text/Cdbk window.
The Contents window is the principal window of the data retrieval program. It allows the user to choose what documentation is displayed in the Text/Cdbk window and which index is used for data retrieval and display in the Variables window.
The Variables window displays the variables found when a search is run or it displays the variables in an extract specification file when that file is opened. "Name, Description, Record Type, Year, Reference Number, and Question Name" information document the variables displayed.
The Text/Cdbk window shown in the figure above provides links to this manual and to other documentation.
In a Windows environment switching active windows can be accomplished simply by clicking in the window. However, if the active window 'covers' one of the inactive windows or the inactive window is minimized, the buttons below can be used to make the inactive window active and bring it to the forefront.
Each window can be minimized by clicking on the X in the size bar:
The double-square button will reduce
the size of the window to less than full size.
The large square will make the active
window fill up the full screen and hide the other windows.
The X button will minimize a particular window. Caution:
However, if the 'X' button in the CHRR NLS
Database Investigator title bar is clicked, this will shut down the
application and any unsaved data will be lost.
Select a Cohort to investigate and begin developing a search strategy within the Contents window. Or maximize the Text/Cdbk window and view on-line, hypertext documentation.
In order to begin to develop a search strategy, a data
set, or cohort, must be specified in the Cohort field. Depending on which of
the three options were selected in "Setting Up the NLS Software,"
you may have to insert the CD into you CD-ROM drive. The figure below shows the default display for the Contents,
Variables, and Text/Cdbk windows. You may resize the windows according to your preferences or click on the Window option in the menu bar and choose a pre-specified command from its drop-down
menu .
To specify a cohort, use the drop-down menu in the Cohort field to see the data sets available. Click on the data set you want to
investigate.
Once a data set is specified a list of the indexes and documents that are associated with the data set will appear in the Contents window. The two other windows, Variables and Text/Cdbk, are used to receive data depending on choices made in the Contents window. The windows will be discussed in the following order Contents, Variables, and Text/Cdbk.
Figure 4 also shows the Text/Cdbk window displaying the NLS Investigator User's Guide documentation.
The Contents window shown in Figure 4 and maximized in Figure 5 is the principal window for data retrieval and is used to navigate among documentation, indexes, and variable displays for a specified cohort. It controls the selection of the components that are available for the selected cohort. Depending on the cohort selected different documentation and indexes may be available. The contents window is also used to develop a search strategy of selected variables. To work with the Contents window, it may be best to maximize it as shown in Figure 5.
When the Contents window is maximized its three frames from top to bottom can be viewed better. The frames from top to bottom are:
1. The top frame or Components of the data set that include available documentation and indexes to the variables that are available in the data set. To select a specific index, click on + Indexes expand the tree and display the contents of the available indexes; then double-click on the desired index name;
2. The middle frame or Index Terms available in the selected Index, and a 'Search for' engine to speed up variable selection; and
3. The bottom frame or Items (i.e., variables) selected from various indexes that are being assembled into a Boolean search strategy.
The Components of the data set (top frame) allow the user to switch access to the principal indexes of the data set and to select the documents to appear in the Text/Cdbk window. The available indexes may vary from one data set to another. For example, the 'Area of Interest' index has replaced the previously used Record Type index. To view the contents of an index, double-click on the index name and its contents will appear in the Index Terms frame (middle).
The available indexes in the currently selected data set are named:
Any Word in Context is an index to all words used in the variable title. The number of occurrences (Entries) appears with the word.
Area of Interest (formerly Record Type) is an index to categories of variables that are grouped by areas of interest and assigned to the variable.
SurveyYear is an index to variables used in a Survey year.
Refnum is an index to variables by reference numbers, i.e., a number that is assigned to variables in order to locate them across survey years.
Qname is an index to variables by question name and locates them across years.
To locate a desired term in an index, scroll down to it in the middle frame by using the scroll bar. Or search for a term, by using the Search for field. Begin typing in the name of the term in the Search for field and the list will scroll down to the first occurrence of that combination of letters. To include an index term in a Boolean search, left-click on a term to highlight it and drag it down to the Item frame (bottom) to select it. Variables must be placed in the Item frame to run a Boolean search on the variables. If the Boolean search is successful, the variables associated with the selected terms will be extracted for further examination and manipulation. The variables extracted, as the result of a search, will appear in the Variables window.
To view all occurrences of a variable (Index term), double-click on its name. Each occurrence (equal to the number of Entries) of the variable along with brief documentation will appear in the Variables window.
Figure 5 above shows three variables each selected from a different index. CHILD was selected from the Any Word in Context index, 1998 from the Year index, and 2ND PREGNANCY from the Keyword in Context index.
A variable that appears in the lower Item frame may be manipulated via the Operators-Commands pop-up menu. To view the Operators-Commands pop-up menu, right-click on the desired item name (variable) to highlight it and bring up the pop-up menu. The pop-up menu is comprised of the following three parts:
· Boolean operators at the top;
· Variable position commands, in the middle; and
· Strategic commands at the bottom.
To change one of the operators in the Operator column of the Search Strategy frame, right-click on it and the operators-commands menu will appear. To select a different operation (AND, OR, NOT), move the highlight bar to it and click. It will change in the Operator column.
To change the position of a variable in the Item column, right-click on it and the operators-commands menu will appear. To move the variable in the list, move the highlight bar to the choice (Move up or Move down) and click. The variable will move accordingly in the Item column.
To clear all variables in the Item column, right-click anywhere in the lower frame and the operators-commands pop-up menu will appear. Move the highlight bar to Clear and click and all the variables in the lower frame will disappear.
To delete a variable in the Items list, right-click on the variable to highlight it and bring up the operators-commands dialogue box. Move the highlight bar to Delete and click.
To run or apply the Boolean search strategy, right-click anywhere in the lower frame and the operators-commands pop-up menu will appear. Move the highlight bar to Apply and click. The search will be run and if the Boolean search is successful, the variables associated with the selected terms will be extracted and displayed in the Variables window for further examination and manipulation. The Variables window will appear displaying the variables.
The Variables window displays the variables resulting from a successful search or those saved in previously tagged set when it is opened. Variables appearing in the window may be saved, tagged for saving or untagged, or examined in further detail.
When a variable is tagged a checkmark is placed in the checkbox in the left most column. When a variable is untagged, no checkmark appears.
To place a checkmark in the checkbox, click in the box next to the desired variable. To remove a checkmark, click in the box where the checkmark appears and it will disappear.
To tag all variables at the same time, right-click anywhere in the variables window and the Tagging Pop-up Menu shown in Figure 8 will appear. Highlight the Tag All option or press <F5>.
In the Variables window, multiple variables may be selected for tagging or untagging. All variables may also be tagged or untagged at once. As shown in the Figure 8, four functions may be used to make multiple selections.
To tag or untag all variables in the list, right-click anywhere within the variables list window. Move the pointer to the Tag All or Untag All and click. Checkmarks will appear or disappear, appropriately. An alternate method is to simply press <F5> or <F6> when the Variables List is the active window.
To tag multiple but not all variables, the variables must be selected. Selected variables are those that are highlighted.
· To highlight a contiguous block of variables, click on the first one in the block then hold down the <Shift> key and click on the last one in the block. <Shift> + Click. When the block is highlighted, place the pointer over one of the highlighted variables and right-click. The Tagging Pop-up Menu will appear. Move the pointer to the Tag Selected and click. A checkmark will appear next to those selected. An alternate method is that when all the desired variables are highlighted, <F3> may be pressed to place the checkmarks.
· To highlight non-contiguous variables, hold down the <Alt> key and click on each variable to highlight. <Alt> + Click. When the non-contiguous variables are highlighted, place the pointer over one of the highlighted variables and right-click. The Tagging Pop-up Menu will appear. Move the pointer to the Tag Selected and click. A checkmark will appear next to those selected. An alternate method is that when all the desired variables are highlighted, <F3> may be pressed to place the checkmarks.
To save tagged variables, click on File in the menu bar, move the highlight bar to Save Tagset and click. The standard Windows dialogue box will appear asking you to name the file and specify the location of the file to save. An extension, in this instance for the NLSY79 public releases, ".ythpub" will be attached to the extract specification files (tagsets) used to store variables Each cohort will have a unique identifier. All other extensions stay as they were. To see a list of file name extensions, see Appendices, File Names.
To view a single variable in detail, double-click on it in the Variables window and it will appear in the Text/Cdbk window. Or, highlight the variable in the Variables window by right-clicking on it. The Tagging Pop-up menu will appear as shown in the figure above. Move the highlight bar to Codebook and click to see the codebook description of the variable. An alternate method to bring up the codebook display is to press <F2> when a variable is highlighted.
The codebook display of a variable appears in the Text/Cdbk window as shown in the figure "Codebook Display of a Variable." Notice how closely the codebook documentation resembles the question in the survey. This is because the OSU CHRR CAPI for Questionnaire Design is a self-documenting system in which each question-response pair becomes a meta-datum documenting itself.
Information that identifies the variable is found at the top of the codebook display: Variable number, Variable label, Question number, Year used, Record type (or Area of Interest), and Variable name.
Variables related to the current one are mentioned as 'see' references in the middle of the of the display: e.g., SEE R(26048.).
The frequencies of the responses selected, including Refusals and Don't Knows, and Valid/Invalid Skips complete the codebook description of the variable. (See cohort specific Users' Guide, section "Accessing Variables By Codebook", for a detailed explanation of codebook conventions.)
Detailed information from the codebook can be used to determine if the variable is one to keep in the extract specification file or if it should be discarded.
The Back and Forward buttons control what is displayed in the Text/Cdbk window. Like in you Internet browser, Back will go back to the previous display in the window, if applicable, and Forward will go to the next one, if applicable.
To go back to the Variables window, press the Variables button on the Navigation bar. To go to the Contents window, press the Contents button.
After completing a search and creating an extract specification file, the next step is to setup and run the extract.
To perform an extract, an existing tagset must be open. To open a tagset, go to the File component of the Menu bar and choose Open Tagset Browse and select the desired tagset from the standard Windows dialogue box. When the tagset is open, go to the Extract component of the Menu bar and choose Extract Saved Variables. Before choosing to extract saved variables you may wish to review them (and edit them) by choosing Review Saved Variables.
When the function Extract Saved Variables is activated, the screen shown below in the Figure 10 will appear. From this window you may finalize the data file to be extracted in several ways:
· The sample universe of cases may be qualified by limiting the number of cases in the sample and/or by pressing the Edit Boolean Specification to create a Boolean phrase (left half of window).
· One of the file characteristics for the output file may be specified under the Extract Data File Type portion of the window by clicking on a file type and additional output files may be specified under the Optional Output File portion of the window.
· Finally two options can be carried out on the file: Run Extract or Write Codebook.
The Menu Bar contains the actions to carry out on the sample universe: Run Extract, Stop Extract, Write Codebook, and Exit the extract.
The left half of the window is used to qualify the sample universe of cases by limiting the range of cases and by creating a Boolean Specification.
The right half of the window is used to select the format(s) of the output file.
Information about the extract file such as name, sample size (Number of variables and Number of cases), output record length, and output file size is found in the lower right half of the window.
To make selections in the extract window, press a button, click in a field and type in the limits, or click on a radio button and or check box.
·
A Radio button (
) is a
button that when clicked excludes other choices in the list.
·
A Check box (
) is a
button that when clicked selects the option listed. Multiple choices may be checked.
The left half of the window allows the user to manipulate the size of the sample universe by combining four methods. Each method is discussed below.
The number of cases in the sample universe may be changed in several ways. The number of cases selected for the extract may be altered and a fraction of the cases may be selected for extraction, the first and last cases read may be changed, and a Boolean specification may be run against the variables in a case to include or exclude them.
Select Number of
Cases
To change the number of cases read,
· Place the cursor in the Select 1 of every __ case(s) field; and
· Type the denominator integer in the blank.
Instead of always beginning with the first case in the file another may be specified as the beginning case.
To change the first case read,
· Place the cursor in the "First case to read" field and
· Type in the integer of the first case to read.
To change the last case to read,
· Place the cursor in the "Last case to read" field; and
· Type in the integer of the last case to read.
When any or all of the "selected number of cases", the "first case to read", or the "last case to read" are changed and the cursor is placed in another field, information in the "Number of cases" field will change correspondingly. (Compare the number of cases in Figure 10 and Figure 11 to the right.)
|
Figure 12 Create a Boolean Specification |
When the Edit Boolean Specification button is pressed, the window shown in Figure 12 will appear. The Boolean phrase constructed will further manipulate the sample universe by performing operations on the extracted variables that are listed in the right box of the window.
The Boolean phrase is constructed in the top box by selecting variables from the list and qualifying them with the Boolean and arithmetic operators listed, constants, and variables.
Rules of arithmetic syntax such as balanced parentheses must be followed.
The maximum length of the Boolean string allowed is 255 characters.
If violations of syntax or length occur, an error message will be displayed.
In Figure 13 the phrase is constructed in the top box. You may type the information in the box, you may use the buttons and variables displayed on this screen to select the variables and operators, or you may use a combination of both. When a variable is highlighted in the list box on the right, the Description of the variable will appear across the bottom of the screen (e.g., 2nd PERIOS OF PAID LEAVE ).
To place one of the
extracted variables (e.g., R5266600) in the construction box at the top, double-click
on it in the box on the right. (You may
also select a variable by clicking on the Variable
button. The Variable field will appear
with the Accept and Cancel buttons. Then select it from the list in the
drop-down menu and press the Accept
button.)
To place one of the operators (e.g., >) in the construction box, type it, or click once on its button.
To place a constant (e.g., 0) in the construction box, type it, or click on the Constant button. (When you click on the Constant button, the constant field will appear with the Accept and Cancel buttons. Use the 5 or 6to increase or decrease the number in the field and press the Accept button when the number is correct.)
In the top box, place the cursor at the location in the phrase where the edit will occur and click to activate the text cursor. You may now begin typing to insert, or press Delete to delete the next character, etc/ In standard Windows fashion, double-clicking over a variable or operator will select the whole term by highlighting it and you can delete it or begin typing to replace the selection.
When the phrase is complete, press the Exit button to exit and save the Boolean specification. Or, press the Quit button to exit and not save the Boolean specification.
Output file specifications may be changed. Each specification available for a selected cohort is listed in the right half of the window in Figure 14.
In the Extract Data File Type portion of the window the desired extract
file type is selected. Depending on the
size of the Cohort selected and the form of the compression used, some extract
file types may not be available (e.g., SAS and SPSS). The extract file types
are mutually exclusive. That means that if the radio button for one
extract data file type is toggled ON (), no
others may be selected.
In the Optional
Output Files portion of the window one or more file types may be selected
by placing a check mark in the check box ().
Multiple choices in the list may be checked.
The following file formats are available in the Extract Data File Type portion of the window.
· A formatted ASCII file is a raw file in which each datum is located in a fixed position, i.e., column, in each record.
· A delimited ASCII file is a raw file in which each datum is separated by a comma. This format can be read by many statistical packages, spreadsheets, and programming languages.
· A DBASE 3 file is a database file that can be read by DbaseÒ software.
· A Stata Dictionary is a file that can be used with Stata software.
Depending on the cohorts available, their size, and due to various forms of compression on the CD the following options may be displayed:
· SASÒ control cards describes a file that can be generated to read the binary data file directly from the compact disk. These can be used to run PC-SASÒ to perform any desired SASÒ procedure such as creating a SASÒ system file. (This specification conserves disk space because it eliminates the need to create a raw file on the PC hard disk that will ultimately be used to create a SASÒ system file.)
· SPSSÒ control cards describes a file that can be generated to read the binary data file directly from the compact disk. These can be used to run SPSSÒ to perform any desired SPSSÒ procedure such as creating a SPSSÒ system file.
Note: The SAS and SPSS control statements generated include basic data location, label and missing value information. While every attempt has been made to provide read-to-use statements, the platform and version of your specific installation may require modification of these generic control programs. Please consult your local SPSS or SAS coordinator for site-specific problems.
In the right half of the window, lower portion, information about the extract file is displayed in the fields listed below. The original size of the sample universe and output file size appears when the window appears. As the size of the sample universe is changed, this information is reflected in the fields. Other fields display the length of the record in the file and the output file name.
· Number of variables
· Number of cases
· Output record length
· Output file size
· Output file name
Several actions may be carried out on the extract file. The buttons in the menu bar controls these actions.
· Run Extract begins the extract.
· Stop Extract cancels the running extract.
· Write Codebook produces the codebook for the extract specification file.
· Exit abandons all processing and closes this window.
Carefully read the screen to assure that all selected specifications are correct.
When they are correct, press the Run Extract button. The "Extract In-progress" figure shows the extract operation in-progress. Information about the progress of the extract will appear at the bottom left of the screen.
The Stop Extract button becomes active and the extract may be aborted when it is pressed.
Summary information on the extract run will appear at the end of processing. This information includes the resulting number of cases read, cases kept, and elapsed time. The resulting output files are listed by name.
Press the Exit button to close the window and return to the Extract window.
Normally the Codebook file of the extract file is produced automatically. But when only the codebook is desired, this button can be used without running the extract. When the Write Codebook button is pressed, no window appears on the screen. A codebook file is created and placed in the directory where the software was installed. This file contains a codebook representation of the variables in the extract file. The name of this codebook file is the same file name as extract file with a .CDB extension, e.g., [filename].CDB.
NOTE: This codebook file should not be confused with the complete codebook file created when the Write Documentation function is selected from the File component of the Main Menu screen. The file name of the complete codebook will be specified at the time of creation.
When the Exit button is pressed, the user is returned to the Main Menu screen.
Chapter 5 - Write Documentation or view sample extract files: chld-88-2ndpreg.CDB, chld--88-2ndpreg.DCT, chld-88-2ndpreg.DOC, chld-88-2ndpreg.SAS, chld-88-2ndpreg.SPS.
When the Write Documentation function is selected from the File component of the Menu bar, the Create Documentation Files window appears and allows the user to select from a number of documentation options.
The buttons at the top of the window perform the actions on the files. To specify an Output Location, type it in the box or use the Browse button to browse for the output locations.
The Create Codebook check box will create the complete codebook for the data set selected in the Cohort box on the Main Menu. Caution: the data sets are large and a codebook for the complete data will be a large file.
The tabs in the middle of the window control the type of documentation produced and the format.
The Index Options Tab - creates documentation for one or more indexes.
The List Options Tab - creates documentation for the variable names and or the question names.
The Field Widths Tab - allows the user to specify the format of the output.
Use the Index Options tab to select the index for which documentation will be produced. When an index is checked additional fields appear defining the default output file name, file size, and the default number of lines per page.
The Output Filename and Lines per Page may be edited. Simply type in them to edit or use the arrows to increase or decrease the number of lines per page.
If the Create Codebook box is checked, the same additional information appears for it. The boxes may also be edited. To create the files specified, press the Create Files button.
Or choose another tab for more documentary options.
Use the List Options tab to select print out of the lists of the variable names.
When a list is checked additional fields appear defining the default output file name, file size, and the default number of lines per page.
The boxes may be edited. Simply type in them to edit or use the arrows to increase or decrease the number of lines per page.
To specify an Output Location, type it in the box or use the Browse button to browse for the output locations.
To create the files specified, press the Create Files button.
Use the Field Widths tab to customize the output files. A field may be excluded with a width of '0' or the width of the field maybe be changed by typing in a new integer.
The figure to the right displays the default field widths.
To create the files specified, press the Create Files button.
|
|
Command |
Description |
Shortcut |
|
|
File |
|
|
|
|
|
|
New Tagset |
Open a new, blank extract specification file. |
|
|
|
|
Open Tagset |
Re-open an existing extract specification file. |
|
|
|
|
Append Tagset |
Re-open an existing extract specification file and append variables to it. |
|
|
|
|
Save Tagset |
Save selected variables in an extract specification file. |
|
|
|
|
Save Tagset As |
Save selected variables in a renamed extract specification file. |
|
|
|
|
Write Documentation |
Create printable text documentation files. |
|
|
|
|
Exit |
Close the program. |
|
|
|
Window |
|
|
|
|
|
|
Default |
Reset NLS Database Investigator windows to the default arrangement. |
Alt+0 |
|
|
|
Cascade |
"Stack" the child windows in NLS Database Investigator. |
|
|
|
|
Tile Horizontal |
Horizontally arrange the child windows in NLS Database Investigator. |
|
|
|
|
Tile Vertical |
Vertically arrange the child windows in NLS Database Investigator. |
|
|
|
|
Contents Window |
Switch to the Navigation window. |
Alt+1 |
|
|
|
Variables Window |
Switch to the Variable List window. |
Alt+2 |
|
|
|
Text/HTML Window |
Switch to the Text/HTML window. |
Alt+3 |
|
|
Extract |
|
|
|
|
|
|
Review Tagged Variables |
Review selected variables. |
|
|
|
|
Extract Tagged Variables |
Create an extract with selected variables. |
|
|
|
Data |
|
|
|
|
|
|
View Case |
Select one case and view its data values. |
|
|
|
|
Close Case |
Stop viewing the selected case. |
|
|
|
Options |
|
|
|
|
|
|
Font |
Change the appearance of text. |
|
|
|
|
Add a Cohort |
May be used to manually install a cohort (data set). In future enhancements, to be used to connect to a dataset that resides in an Oracle database. |
|
|
|
|
Delete a Cohort |
Removes a cohort from the "Cohort" drop down list. Will not remove files from the drive. The currently selected cohort cannot be deleted. |
|
|
|
Help |
|
|
|
|
|
|
About |
Display version information. |
|
|
|
|
Contents |
Link to the Text/Cdbk window which will display a table of contents for the HTML manual that comes on the CD-ROM. |
|
|
After selecting one of the options and naming the extract specification file, a three-letter default file extension will automatically be attached to the name entered. Depending on what data set was originally chosen in the Cohort field, the file extension for the extracted specification file will be one listed in the Extract File Types below. The name of the current working file is identified in the NLS Database Investigator Title Bar (blue bar) as [Tagset: file-name]. If it is a new file the file name will be "untitled" until it is given a name when the "Extract Tagged Variables" option is activated.
The extension depends on the data set used.
*.MSR Men, extract specification file used to store variables.
*.BSR Boys, extract specification file used to store variables.
*.WSR Women, extract specification file used to store variables.
*.GSR Girls, extract specification file used to store variables.
*.ythpub NLSY79, extract specification file used to store variables.
*.CSR Child, extract specification file used to store variables.
All files subsequently created in conjunction with the extract file will use the same 8-letter filename, and depending on the file type to be created, one of the following extensions will be attached.
(tagset_file_name).CDB codebook print files of the total universe codebook for the selected variables.
(tagset_file_name).DAT extracted data files containing the selected variables.
(tagset_file_name).DOC index print file listing the selected variables with descriptive statistics.
(tagset_file_name).SAS SAS control cards.
(tagset_file_name).SPS SPSSX control cards.
Below is a complete list of files that can be produced when the Write Documentation" option is chosen from the File component of the Main Menu. Documentation files are given the default extension of .txt. Most of these files are enormous so if you want to produce them be sure you have enough room on your PC. The "anyword.txt" file alone is 1.1 Gb!
anyword.txt Any word in context index.
Dscrpt_test.txt Area of Interest index
Reference.txt Reference number index.
QYear.txt Survey year index.
Question.txt Question name list.
codebook.txt Complete codebook.
varname.txt Variable name list.
The following sample extract files were created as a result of Performing an Extract on the file set up in Chapter 4. The files displayed in this section are:
|
Codebook file for the variables extracted. |
|
|
Stata Dictionary file for the variables extracted. |
|
|
Documentation File to Document the Raw Data File (*.DAT) |
|
|
SAS file. |
|
|
SPSS file. |
It is important to remember that although CD-ROM disks are practically indestructible, careless handling can permanently damage them. A very small scratch can render a disk unusable. Fingerprints and other dirt can cause fatal errors during program execution. Even microscopic dust particles can cause problems.
Anything you do to reduce disk handling is desirable. CD-ROM drives which use disk caddies have become common. A disk caddie for each CD-ROM you own may seem to be an unnecessary added expense, but it can be well worth the cost if it protects even one of your expensive CD-ROM disks. If you have to handle a CD-ROM disk, handle it only by the edges. NEVER touch the data surface of the disk (i.e., the bottom, shiny side without printing). Don't lay CD-ROM disks on anything. Use caddies or jewel cases to store them.
Error messages generated by NLS DATABASE INVESTIGATOR appear in the following format:
Run-time error nnn at xxxx:yyyy
Where,
nnn is the error number, found in the list on the following pages; and
xxxx:yyyy is the error address, that is useless except during program development.
The errors come from Borland Pascal with Objects Version 7.0Ô. More information about all errors can be found in the Programmer's Reference Manual.
|
1 |
Invalid function number. |
{You should never get this} |
|
2 |
File not found. |
|
{Often associated with CD-ROM drives connected via a network. Try the disk on a PC with a locally connected CD-ROM drive. You can compare directories produced with both setups. Usually the network directory will not show all the files.}
|
3 |
Path not found. |
(You should never get this} |
|
4 |
Too many open files. |
|
{The program has too many open files. You probably don't have a FILES=xx entry in your CONFIG.SYS file or the number (xx) is too small. Add FILES=20 or increase the number to 20 and reboot your computer.}
|
5 |
File access denied. |
{You should never get this} |
|
6 |
Invalid file handle. |
{You should never get this} |
|
12 |
Invalid file access code. |
{You should never get this} |
|
15 |
Invalid drive number. |
{You should never get this} |
|
16 |
Cannot remove current directory. |
{You should never get this} |
|
17 |
Cannot rename across drives. |
{You should never get this} |
|
18 |
No more files. |
{You should never get this} |
|
32 |
|
{The most likely cause is a damaged or dirty CD-ROM disk.} |
|
100 |
Disk read error. |
{This input/output error may occur due to faulty hardware or a dirty or damaged CD-ROM disk. Thus a data field gets read incorrectly and results in the program attempting to read beyond the end of file.} |
|
101 |
Disk write error. |
{Your output disk is full} |
|
102 |
File not assigned. |
{You should never get this} |
|
103 |
File not open. |
{You should never get this} |
|
104 |
File not open for input. |
{You should never get this} |
|
105 |
File not open for output. |
{You should never get this} |
|
106 |
Invalid numeric format. |
{You should never get this} |
Critical errors will usually terminate the program. They often are caused by faulty hardware or dirty CD-ROM disks. For more information refer to a DOS programmer's reference manual.
|
150 |
Disk is write protected. |
|
151 |
Unknown unit. |
|
152 |
Drive not ready. |
|
153 |
Unknown command. |
|
154 |
CRC error in data. |
|
155 |
Bad drive request structure length. |
|
156 |
Disk seek error. |
|
157 |
Unknown media type. |
|
158 |
Sector not found. |
|
159 |
Printer out of paper. |
|
160 |
Device write fault. |
|
161 |
Device read fault. |
{The most common hardware problem usually indicates a damaged or dirty CD-ROM disk. Usually this occurs during data extraction. If the error occurs at the same place repeatedly, the CD-ROM disk is probably damaged; otherwise, the CD-ROM drive may be faulty or the CD-ROM disk may be dusty or fingerprinted.}
|
162 |
Hardware failure. |
Fatal errors immediately terminate the program.
|
200 |
Division by zero. |
{You should never get this} |
|
201 |
Range check error. |
|
{This error is known to occur with certain mouse drivers. It MAY have been corrected. If it recurs, try rebooting without the mouse driver installed in CONFIG.SYS or AUTOEXEC.BAT.}
|
202 |
Stack overflow error. |
-or-
|
203 |
Heap overflow error. |
{Either of these errors indicates that there is not enough memory available to run CHRRNAV. Too many device drivers and TSR's (terminate and stay resident programs) loaded in CONFIG.SYS and AUTOEXEC.BAT can cause this error. Poor memory management or not enough RAM can also cause this.}
|
204 |
Invalid pointer operation. |
{You should never get this} |
|
205 |
Floating point overflow. |
{You should never get this} |
|
206 |
Floating point underflow. |
{You should never get this} |
|
207 |
Invalid floating point operation. |
{You should never get this} |
