Introduction 
to 
SPSS/Windows 10 at William and Mary

by Bob Dawson

(SPSS/Windows Version 10.0 is available in the William and Mary Labs in a Windows 2000 environment)

Table of Contents

I. About SPSS/Windows
II. The Introduction: A Simple Example

  1. Defining and Entering Your Data 
  2. Summarizing Your Data Cases: 1st Example of SPSS Processing 
  3. Reviewing Your Summarized Data Cases Output 
  4. Correcting/Editing Your Data 
  5. Getting a Frequency Distribution: Second Processing Example 
  6. Reviewing Your Frequency Distribution 
  7. Specifying the Processing Beyond the Defaults 
  8. Getting a Crosstabs Distribution: Third Processing Example 
  9. Saving Your Data 
  10. Printing Your Output 
  11. Saving Your Output 
  12. Printing/Saving/Editing Your Charts 
  13. Exiting SPSS/Windows 
  14. Restarting SPSS/Windows and Retrieving Your Saved DATA Editor/Output Viewer

Appendix A: More about SPSS/Windows Release 10.0

  1. Basic Items of Interest
  2. PC Computer Resource Requirements for SPSS/Windows 10.0
  3. Differences Between Versions of SPSS/Windows: 9.0 vs. 8.0 and 10.0 vs. 9.0
I. About SPSS/Windows (GoBack)

SPSS/Windows is a user friendly statistical package that requires in most cases no programming to get results in contrast to using traditional SPSS programming. Through menu and dialog box selections (mouse clicking) with occasional actual text entry, you can specify and perform an SPSS operation; SPSS/Windows simply takes your selection information and uses it to automatically compile and execute the necessary programming. 

SPSS/Windows runs on most PC compatible systems with sufficient processor power and memory storage to support a windows environment and the needs of SPSS 10. It is available in the labs at William and Mary on any PC work station connected to a server that distributes SPSS/Windows (most PCs if not all). SPSS/Windows improves on its SPSS/PC predecessor several ways. In addition to offering windows environment with facilities to perform via menu/dialog selections, SPSS/Windows provides the full range of features that hitherto were only available at W & M with main frame SPSS. 

SPSS provides a wide distribution of analysis ranging from descriptive statistics to complex multivariate analysis. Output is accompanied with helpful plot/chart/graphics information. SPSS can also do report writing, data file management, and data massaging via a variety of transformation techniques. As mentioned, the windows facilities of SPSS provides most of this through menu/dialog box selections, SPSS/Windows automatically doing the programming and running; a spreadsheet-like DATA Editor window permits data defining, entry/editing, and IO management; an OUTPUT Viewer window provides output review, editing, IO management. Optionally, if the user wishes to "take over the wheel", a SYNTAX Editor window allows direct program development/execution. Graphics can be achieved either via the usual "menu/dialog box selections" or "interactively". A CHART Editor window is available to edit graphics. A Draft Viewer window can alternately display the output, the statistics tables in typewriter format, the graphics remaining in hi-resolution, and the table borders using box characters for clean, non-break lines. And finally there are HELP windows for both general help and help with a given dialog box; in addition, help with any item in a dialog box is available with just a right click of the mouse.


 

II. The Introduction: A Simple Example GoBack

Assume you wish to use SPSS/Windows to process these five cases of data: 

                  AGE     SEX  EDUC
                   25       1     4
                   27       2     5 
                   31       1     4  
                   33       1     6   
                   40       2     7  
Maybe you want to summarize the cases and then later do a frequency distribution of each variable (to make it practical assume the data is actually a large data set). To do this on a William and Mary lab computer, you start on any lab PC workstation and "click" (select) "Start", "Programs", and then "Statistics"; in the emerging "Statistics" menu, you "click" "SPSS 10", then "SPSS 10 for Windows"; the "SPSS for Windows" dialog box will then open; "click" (select) the "Type in Data" tiny circle and hit Enter. SPSS/Windows now starts, showing a medium size SPSS Data Editor window in the foreground as shown in Fig 1 below.

Figure 1 (SPSS Editor after Start Up)
At the top you find this menu bar:  File Edit View Data Transform Analyze Graphs Utilities Window Help  This bar provides you with these command groupings in Windows/SPSS: 

File is the group that allows you to get or save data, print output, save output to a file on your diskette, etc. Transform allows you to recode variables into the same variable, recode variables into a different variable, create new variables through COUNT and COMPUTE, etc. Analyze allows you to run most of the statistical procedures; these range from basic ones such as DESCRIPTIVES, FREQUENCIES, and CROSSTABS (selected from the DESCRIPTIVE menu under ANALYZE) to the more advanced statistical procedures such as those under the REGRESSION menu. Graph allows you to do bar graphs, line graphs etc., non interactively and interactively. These graphs can be saved in a file, printed or both. Utilities has one particularly useful facility. Its VARIABLES item displays the definition of each variable, one at a time; using the up/down arrow key or mouse to highlight a variable, you see its variable label and value labels as well as its missing values and type (whether character/numeric and what format). 

At the bottom near the middle is a status bar displaying "SPSS Processor is ready" as soon as the start up is complete. Just above on the left are the "Data View"/"Variable View" icons for switching between these two modes. "Data View" (the default as shown in Fig 1) lets you view/edit data; "Variable View" (see Fig 2) lets you view/edit variable definitions (the viewing also available via the VARIABLES item under UTILITIES -- see above). At the very bottom of your PC monitor screen (not shown in Fig 1), you should see the usual icons on the taskbar for foregrounding other opened applications (foregrounding means making an application current by bringing it to the front); among these icons is one for the DATA Editor window itself; should the DATA Editor go into the background (as can occur when another application or another SPSS window such as the OUTPUT window is fore grounded), you can use this icon to foreground the DATA Editor.

(1) Defining and Entering Your Data GoBack

You are now ready to enter the above example data into the DATA Editor (starting with the Editor as shown in Fig 1 above). This window contains cells to hold your data, one value per cell; each cell column holds the data for a given variable and is initially labeled "var" (a faint displayed dummy label) at the top; each cell row holds the data for a given case. A variable contains information about a certain item in your data cases (in the social sciences, respondents to a survey or interview). 

Before entering your data, you should define it; you can start by naming each variable. In SPSS a variable name must be eight characters or less, begin with a letter, have no embedded blanks, and not use these special characters: !, ?, ', and *; SPSS variable names are case insensitive, meaning that lowercase letters are equivalent to uppercase; and finally some words are reserved (invalid as variable names -- you are informed as soon as you specify one). For example when analyzing a survey, if your first question asked respondents their age, you might want to give the first variable the name "age." To do so you must first shift the DATA Editor to the "Variable View" mode by clicking the "Variable View" icon in Fig 1; in this mode (see Fig 2.1), the rows in the DATA Editor are treated as variables (e.g., Row 1 for Variable 1) and the columns as variable definition items (e.g., Column 1 for the variable's name); thus the Row 1, Column 1 cell will have the name of Variable 1. You now register Variable 1 and get its default definitions: right click on the far left cell designated "1" (for Row 1), click "Insert Variable" in the pop up menu, and the results are shown in Fig 2.1; here, Row 1 shows the default definitions of your 1st variable ("age"); currently its name is "var00001", its type is "Numeric", its (display) width is 8, etc. -- again see Fig 2.1. You can now edit "var00001" to be the chosen name for your "age" variable (just double click to edit); you can make it something easy to enter such as "Q1" (then "Q2", "Q3", etc. when defining the other variables) or you can give it the more substantive name "age". In a survey, for example, you must have a variable name for each question (including those such as gender filled in by an interviewer). 

To further define your data and make things easier in your analysis, you may want to explain what the variable refers to; for instance, the name "Q1" isn't very obvious, and you will often find that 8 characters are not enough to make the variable meaning clear. Therefore, you will want to go to the "Label" column (partially shown in Fig 2.1, far right) by scrolling until it shows in full view (if not already); then double click the cell in the row of the variable to be labeled and enter the Variable Label (an explanatory name for the variable that can be 256 characters and can have blanks; e.g., Respondent's Age) 

In some cases, you may also want to label individual values of a variable; obviously if the values speak for themselves, like age, you do not need to label them, otherwise it's a good idea. To do so scroll the "Variable View" until the "Values" column appears; then click the cell in the column of the variable to be value labeled (e.g., Row 2 cell to label "sex"); finally click the tiny icon (labeled "...") that should appear in the cell's far right (see Fig 2.2 where the Row 2 cell's icon has been clicked to value label "sex"); a "Value Labels" dialog box should emerge as in Fig 2.2. Now simply type in the category number to which you wish the label attached (e. g. 1) in the Value box, and the label (embedded blanks permitted), in the Value Label box (see Fig 2.2 for these boxes) For example for the "sex" variable, you might assign the Male label to 1's by entering 1 in the Value box, "Male" in the Value Label box and "clicking" Add; similarly you assign Female to 2's by typing 2 in the Value box and "Female" in the Value Label (as shown in Fig 2.3) and once again "clicking" Add. When you have entered all of the value labels for a variable, click "OK". The box will close showing the "Variable View" window with the corresponding cell that was clicked now showing the label of the 1st labeled value (instead of the default "none" value as shows in Fig 2.2)

Defining your data can also include assigning certain values as "missing" by "clicking" on the variable's cell under "Missing" (see "Missing" column in Fig 2.2); a tiny "..." icon should show at the cell's right end (identical to the one in the "Value" column in Fig 2.2); click the icon and the "Define Missing " dialog box pops up (not show here). In a survey, "missing values" may be responses like "Refused to answer" or, in some cases, "Don't Know" that you want the computer to ignore; the responses may require treating as missing when you want to know the percent of Male and Female while ignoring those cases when the interviewer neglected to indicate the respondent's gender. Similarly, in order to exclude, for example, those who had no opinion on a question, you might define that response as missing. In any case, you can specify up to three missing values under Discrete Missing Values. These can always be changed at a later date. When you are done, "click" "OK".

Another variable defining item of interest is "Type", which lets you define a variable's type beyond the default "Numeric" (as shown under "Type" in Fig 2.1). Thus, if the "sex" variable in your example data see above had "m"s and "f"s instead of "1"s amd "2"s for males and females (rather handy as you don't need to value label it), you would define it as "String"; just click it's cell (Row 2 under "Type"), click the tiny "..." icon at the cell's end -- same type "..." type icon shown under "Values" in Fig 2.2-- and select "String" in the emerging "Variable Type" box. Or if your variable is "income", select "Dollar", and so on (of course as usual click "OK" when finished). 

Figure 2.1 (The DATA Editor Variable View after Inserting the First Variable to be Defined)

Figure 2.2 (The "Value Labels" Dialog Box -- note the Values "..." icon in Row 2 used to open it)

Figure 2.3 (The "Value Labels" Box after Adding the Male Label & before Adding the Female Label)

Now that you know how to define your data, you can begin entering it (see above for data values); before doing so, of course you should return the DATA Editor to the "Data View" mode (just click "Data View" -- see Fig 2.1). Letting column 1 in the DATA Editor be the "age" variable (column 1 now showing "age" at the top assuming you defined Variable 1 to be "age"), you can start by keying "25", the value of "age" for the first respondent; this goes into the default current cell (again the upper left cell); after you type "25", it does not appear in the cell until you move to another cell (instead it appears in a holding area following a blinking vertical bar above column 2). Let us do this by pressing the right arrow key once to go to the next cell (column 2, top cell) in the current row ; "25" should now appear in its cell.

Figure 3 (DATA Editor after Entering Data/Defining the Variable Labels)
The top cell of column 2 should now have bold borders, signaling it as current. This column is the "sex" variable, you type "1", the 1st value of "sex". Again you key the right arrow key to move to the next cell; a "1" now appears in the 1st cell of column 2. In column 3 "educ", you type "4", the 1st value of "educ". Now hold down briefly the left arrow key to return to column 1 ; "4" appears in its cell. 

To enter the remaining cases you press down arrow just once to move to the beginning of the next case (row 2, column 1); you are ready to enter the 2nd value of "age" and can enter case 2 in the above manner: key "27" ("age" 2nd value) then right arrow, "2" ("sex" 2nd value) then right arrow, "5" ("educ" 2nd value); and then hit the HOME key on your keyboard and the down arrow key (in sequence) to advance to row 3, column 1. Similarly you enter case 3, case 4, etc. until all the above example data has been entered. Press ENTER to register the last value (7) and your DATA Editor should look like Fig 3 above.

(2) Summarizing Your Data Cases: 1st Example of SPSS Processing GoBack

You are now ready to process your data; to summarize your data cases, "click" (select) the "Analyze" item on the DATA Editor's menu bar (see top of Fig 3 above), then the "Reports" item, and then the "Summarize Cases..." item; Fig 4 below shows the resulting "Summarize Cases" dialog box. Here a sub box to the left lists "age", "educ", and "sex" in alphabetical order. (You can change the order in which variables are listed to reflect the order of the file (i. e. the order in which the questions are listed in the data file) before you enter your data (or access other data) by selecting Edit, then Options, then under Display Order for Variable List, then "click" on File Order. The next data file you open (or data you enter) will list variables in the order shown in the Data Editor.)

Figure 4 ("Summarize Cases" Dialog Box)
To keep things simple select all three variables by first clicking "age", dragging down to "sex" (highlighting all names), and clicking the right arrow button on the right in the dialog box; all three variable names (age, educ, sex) should now appear in the "Variables(s)" sub box, thereby specifying these variables to list; all you do now is press ENTER or "click" "OK" to list your data.

(3) Reviewing Your Summarized Data Cases Output GoBack

SPSS should now summarize your data case, displaying it in the OUTPUT Viewer window (which should now appear in the foreground); Fig 5 below shows it after the output has been positioned to the "Case Summary" table. Here you should see two panes (sub windows); the one on the left, the output outline (like a table of contents); the one on the right, the output contents (the output itself).

Figure 5 (Output Viewer Summarizing the Example Data Cases)
You can position your output by selecting (clicking) the appropriate items in the left pane; the corresponding output should then show in the right pane. Fig 5 above shows the "Case Summarizes" item as selected (highlighted) in the left pane, representing the currently selected outline item; the associate output displays in the right pane as shown. If you select (click) the "Case Processing summary" item on the left, that output shows on the right (output such as the number of valid cases for each variable, the percentage, etc.) 

You can also position your output by scrolling; you can do this by using the arrow and page keys, clicking arrows on the bottom/side scroll bars, and/or mouse dragging the line/page positioner on the scroll bars. 

You can open the Viewer to full screen by clicking the box shaped icon in the upper right corner (note that these scrolling and expanding techniques are equally applicable to other SPSS windows such as the DATA Editor). Finally, you can also move the boundary separating the two Viewer panes: "click" and drag to the left and you expand the output contents, to the right, the output outline.

(4) Correcting/Editing Your Data GoBack

Assume you found one or more data entry errors from reviewing you output. To correct your data you first select the DATA Editor by clicking its icon (labeled SPSS Data) at the bottom of the screen; the DATA Editor now moves to the front and becomes current. To edit a piece of data you first scroll the DATA Editor until the desired cell appears (actually not necessary in this example because the data set is so small); then "click" the cell (or move to it with the arrow keys) to make it active; then press the backspace key to erase the cell's value; then type the correct one to change the cell's value in the holding area above column 2; and finally press ENTER or move to another cell to permanently change the current cell's value.

Figure 6 ("Frequencies" Dialog Box)

(5) Getting a Frequency Distribution: Second Processing Example GoBack

Assuming your example data is now OK in the DATA Editor, you are ready for a frequency distribution. As in summarizing your data cases in section (2) above, you "click" "Analyze" on the DATA Editor menu bar and then "Summarize" from the resulting menu; this time however you "click" "Frequencies..." instead of "Summarize Cases..." from the next submenu. Again a dialog box but now labeled "Frequencies" (Fig 6 above). As with the "Summarize Cases" box you select all three variables to frequency distribute (click "age", drag down to "sex", then "click" the right arrow button to put these names in the "Variable(s)" selection sub box); to calculate and output the frequencies, you hit ENTER or "click" "OK".

(6) Reviewing Your Frequency Distribution GoBack

Having done (5) above, you should see the OUTPUT Viewer come to the front, containing now a frequency distribution of "age", "educ", and "sex"; initially if the Viewer is full size, you should see the "Statistics Table" ("N" values of all three variables) below which the "age" variable "Frequency Table"; scroll the output and you see the "Frequency Table" for "educ" followed by the table for "sex"(in alpha order). In the outline pane on the left, you should see selectable items for the "Statistics Table" and each of the "Freq. Tables" (labeled "age", "educ", and "sex"); "click" any of these and you get the corresponding table in the contents pane (to see the actual output as it appears, we leave it to you to perform this example). If the Viewer is not full size you can enlarge it by clicking the middle (box shaped) icon in the upper-right corner.

(7) Specifying the Processing Beyond the Defaults GoBack

Up to now you have used the default specifications of the dialog boxes for both "Summarize Cases" and "Frequencies", specifying only the variables (see figs 4and 6above). In the Fig 4 "Summarize Cases" dialog box, you could have, to illustrate, just listed the 1st through the 3rd case by clicking the tiny box containing the "100" value for the "Limit cases to first" specification (near bottom left, Fig 4 above), deleted the "100", and entered "3". You could also have specified numbering the cases in the output by clicking the square sub box labeled "Show case numbers" (Fig 4 above, bottom left).

In the "Frequencies" dialog" box (Fig 6 above), you could have clicked the "Statistics" oval like button to pop up a "Statistics" dialog box (not shown here) to specify the statistics beyond the straight frequency distribution. Or you could have selected the "Charts" button to open a "Charts" dialog box (again not shown) to request "bar charts", "pie charts", or "histograms". Or you could have clicked "Format" to display a dialog box (not shown) from which you could have specified displaying the frequency values in descending order (over the ascending default). 

In general if you are new at using SPSS Windows, you are probably best off using the defaults in the beginning (that is, pick the minimum from the dialog box(es)) and playing it from there. Thus in our example problem, if after getting the frequencies in ascending order by using the defaults, you decided you wanted them in descending order, you can easily output descending frequencies by following the procedure of 5) while choosing "Descending Values": "click" "Format..." (in the "Frequencies" box, Fig 6 above), then "Descending Values" in the "Format" sub dialog box (not shown), then "Continue" to return to the "Frequencies" box, and so on. Note: when you re-run your frequencies for the same variables, you don't have to re-select the variables in the "Frequencies" box; SPSS/Windows always keeps your last settings in a given dialog box, retaining them until you reset them with the box's "Reset" button (see Fig 6 above) or exit SPSS/Windows.

(8) Getting a Crosstabs Distribution: Third Processing Example GoBack

Assume that you want to get a two-way (crosstabs) frequency table between "educ" and "sex" (to see how one relates to the other by frequency). To effectively enable a crosstabs analysis both variables are best confined to just a few categories; therefore assume that "educ" now has just 1's, 2's, and 3's (because you recoded low values to 1's, middle values to 2's, and high values to "3" -- all by selecting "Recode" from the Transform menu -- details not covered here). 

To crosstabs "educ" with "sex" you click "Analyze" on the Data Editor menu bar (or Viewer menu if you wish), then "Summarize", and then "Crosstabs"; the Crosstabs dialog box (see Fig 6.1) that emerges is similar to the Frequencies box in Fig. 6 above except that now there are "Row(s)" and "Column(s)" entry boxes instead of just one "Variables" entry box as in Fig. 6; in these you respectively specify the "Row" and "Column" variables of the two-way table(s). 

In our example instead of selecting all three variables and putting them in one list as in the previous processing example, you just select "educ" and click the button pointing to "ROW(s)", select "sex" and click the button pointing to "Column(s)" (or the other order if you prefer "educ" as columns, "sex" as rows). 

Now click "OK"; a two-way table between "educ" and "sex" displays in the OUTPUT Viewer (as in Fig 6.2where the data has been expanded to 20 cases and the column percentages added thru specifications discussed below). In Fig 6.2 the rows are labeled 1, 2, and 3 for "educ"s low, middle, and high classifications, the columns 1 and 2 for "sex"s males and females; each cell has the case count for its row and column (e.g., the "row 2, column 2" cell might have a 5 count, meaning 5 females with middle education).

Figure 6.1 (Crosstabs Dialog Box)
In addition to just a case count in each cell, you might want to know other information such as expected values, percentages, and residuals (differences between expected and actual counts). To request this you click the "Cells" button of the Crosstabs box (Fig 6.1); a Cells dialog box displays. To illustrate, assume you click (and thus check) the tiny box next to "Columns" under the "Percentages" specifications; after you click "Continue" and then "OK", another "educ" vs. "sex" table in the OUTPUT Viewer appears (exactly as in Fig 6.2); in each cell you now have both the respondent count and the percentage of the count to its column total (e.g., in cell 2, 2 a 5 count accompanied by 20% says that 5 middle educated females are 20% of the total females). 

In general you can crosstabs one list of variables with another; e.g., if in a survey you want to cross questions 1, 2, and 3 (assume variable names q1, q2, and q3) with questions 4 and 5 (q4 and q5), you just move "q1", "q2", and "q3" to the "Row(s) box, and "q4" and "q5" to "Column(s)"; click "OK" and all two-way combinations of the 1st three variables and last two variables appear in the Viewer -- "q1, q2, and q3" as rows, "q4 and q5" as columns. 

You can also get crosstabs between two variables for each category of another. Using our "educ" vs. "sex" example, assume you recode "age" into 1's (for young respondents) and 2's (for old); then move "age" into a "Layer 1 of 1" entry box (located beneath the "Column(s)" box -- again see Fig 6.1) by selecting "age" and clicking the button pointing to the "Layer" box; then click OK and you get two "educ" vs. "sex" two-ways -- one for young, one for old (the two tables actually stacked on top so as to look like one). In general you can do this for a list of variables that you put in the "Layer 1 of 1" box in Fig 6.1, getting a two-way table for each category of each variable in the list.

Figure 6.2 (OUTPUT Viewer Showing Crosstabs of Educ vs. Sex)

(9) Saving Your Data GoBack

Assuming that after reviewing your frequencies output, you corrected any remaining data errors and are now ready to save your data in the DATA Editor onto disk. To save it on your diskette in drive "A:" as file "mydata.sav", you start by bringing the "DATA Editor" to the forefront (if not already), then "click" "File" from the menu bar and then "Save as..." from the resulting menu; this opens the "Save As" dialog box (Fig 7 below) for your DATA Editor; next you insert your diskette in your PC "A:" drive (if not done so already); then you hit the

backspace key to clear the contents of the "File Name" entry box; then type "a:mydata.sav" in this box; when you "click" "OK" or press ENTER, your example data is saved as SPSS/Windows system file "mydata.sav" on your diskette in drive "A:". See section (14) for how to retrieve your saved data back into the DATA Editor (as might be necessary in another SPSS/Windows session). Note: if you inserted your diskette in drive "A:" prior to opening the Fig 7 dialog box, the box references "A:" (as in Fig 7 above) instead of the default drive (commonly "G:" accessing the sample SPSS data directory); no problem since it references "A:" anyway after you specify "a:mydata.sav".

(10) Printing Your Output GoBack

To print your output, the contents of the OUTPUT Viewer, you first select the Viewer (if not already); then "click" "File" from its menu bar, and then "Print" from the "File" menu; a "Print" dialog box displays (not shown here). To Print the entire contents of the Viewer's right pane (all your accumulated output), "click" the small circle next to "all" and then the "OK" (or press ENTER).

If you wish to print just a part of your output, for example just your "frequencies", you first "click" the "Frequencies" item (similar to the "Summarize" item in Fig 5 above) in the output outline (left pane) to select just your frequencies; then you proceed as above to open the "Print" box and this time "click" the small circle next to "Selections" in the print dialog box (again not shown); after that it's "OK" or ENTER again. Alternately you can do this by selecting the unwanted output and deleting it; thus, in this example, if you select the "Summarize" as shown in Fig 5 above (left window, assumed to now include the "Frequency" items), press Delete, you erase your "Case Summary" output; then when you print all your output in the Viewer's right pane, just your frequencies print (cause that's all that's left). This approach has the disadvantage of losing output unless you saved it in a file, a procedure discussed in the next section. Note: to select multiple items, "click" each item while holding down the "Ctrl" key.

(11) Saving Your Output GoBack

To save your output in the OUTPUT Viewer, the procedure is the same as saving your data in the DATA Editor (see section (9) above); thus, to save your output in file "a:myspss.spo" you forefront the Viewer, "click" "File", then "Save As...", then backspace to clear the "File name" entry sub box, then type "a:myspss.spo", and finally ENTER/"OK" to make it happen. The dialog box for saving your output is very similar to Fig 7 above except that it's labeled just "Save As", defaults the "File name" to Output1.spo, and the "Save file as type" to ".spo" (Viewer file type). To save just a selection, delete the unwanted output and save (to delete, first "click" the unwanted items in the output outline, while holding down the "Ctrl" key to retain the selections; then press Delete).

When you save your output in the example "myspss.spo" file with the file type set to ".spo", you are saving in SPSS output format and can only access this information in the future by retrieving it into the SPSS/Windows's OUTPUT Viewer (see Section (14)below for how); from here you can browse and/or print as you wish. To include selected SPSS output in the document of another application such as Microsoft Word or Word Perfect, you are best doing a "Copy Object" (from the Viewer's FILE menu) of selected output in the Viewer and doing a standard paste into the document; this seems to work best with Microsoft Word, except that charts/graphs get reduced in size (resized via Word's graphics tools); see section (12) below for how to do a better chart/graph paste (size less changed) via the CHART Editor.

(12) Printing/Saving/Editing Your Charts GoBack

Assume you created bar charts of each example variable in the frequency output (by clicking "Charts" in the Fig 5 box above and then "Bar Charts" in the resulting dialog box). These charts appear in your Viewer right pane with your frequency table output (discussed in (6) above); the output outline bar chart item for your "age" variable, for example, is labeled "Bar chart of age", the frequency table, just "age". 

You can print these charts just as you print other output (as discussed in (10) above); to print just charts you select just the charts in your output and print the selection just as you generally printed an output selection (again as discussed in (11) above). Specifically, you select the charts to print (click chart items while holding down "Ctrl" in the output outline), "click" "print" in the "FILE" menu, and in the pop up dialog print box, "click" "selections" and then "OK"; e.g., to print just the "age" and "educ" bar charts, you can "click" (while holding the "Ctl" key) the "Bar chart of age" and "Bar chart of educ" items in the Viewer's left pane (to select them), and then follow the steps just discussed to print a selection. 

To save any of the charts into the standard SPSS/Windows output file (extension ".spo"), you save the entire output in the Viewer (in the manner of (11) above); to save just chart(s), you delete all else in the Viewer (select, press Delete) and then do the save. 

To save any of the your charts in non SPSS/Windows output format (e.g., in a WEB image ".jpeg" file), you can export the charts in the chosen format; for example, to export all your bar charts into ".jpeg" files, you "click" "Export" from the FILE menu and in the emerging "Export" dialog box, select "Charts Only" from an "Export" pull down menu and "JPEG File" from the "File Type" pull down; "click" OK and your "age", "educ", and "sex" bar charts go into respectively the files "Output0.jpg", "Output1.jpg", and "Output2.jpeg", now in a format compatible for displaying on the WEB.

You can edit a given chart by opening it into the CHART Editor; e.g., to edit your "age" bar chart, you double "click" it and the CHART Editor automatically opens with your "age" chart ready for editing. Here you can change its color, style, titles, labels, add footnotes, rotate it, etc.; you can even change the chart type, converting it to a "pie chart", "line chart" etc. In the Viewer, your "age" chart is correspondingly changed. 

Finally, from the CHART Editor you can readily paste the chart into another application such as Microsoft Word or WordPerfect; you can probably do this with more success than from the Viewer, as the chart size stays essentially the same when you "click" "Copy Chart" from the Editor's FILE menu, do a "special paste" in the other application, and chose "Picture" from the pop up dialog box.

(13) Exiting SPSS/Windows GoBack

To exit your SPSS/Windows session, "click" "FILE", then "Exit"; if you didn't save your DATA Editor since its last alteration, a small dialog box appears for you to "click" "No" if you don't want to save your DATA Editor, "Yes" if you do; if you "click" the latter, the usual "Save As..." dialog box opens and you can save in the manner of Section (9) above. Similarly, if you didn't save your OUTPUT Viewer since its most recent change (editing, more processing, etc.), you also get a chance to save before exiting. After showing box(es) to let you save unsaved windows, SPSS/Windows exits.

(14) Restarting SPSS/Windows and Retrieving Your Saved Data/Output Viewer GoBack

To restart SPSS/Windows you follow the procedure discussed in the opening statements of this part: click "Start", "Programs", "Statistics", "SPSS 10", and "SPSS 10 for Windows"; in the emerging dialog box click "Type in Data" (in the tiny circle); then hit Enter. 

After SPSS/Windows reaches the stage shown in Fig 1, above, you can retrieve the example data that you saved in file "A:mydata.sav" by first clicking "File", and then "Open"; when get an "Open File" dialog box similar to the one you used to save your data (Fig 7 above), you insert your diskette in drive "A:", hit backspace to clear the "File Name" entry sub box, key "A:*.sav" in the sub box, and press ENTER. All files that have the extension "sav" should list in a larger sub box beneath the entry sub box; in this sub box "click" "mydata.sav" twice (or once plus the usual "ENTER/"OK" procedure). Your example data complete with the "age, sex, educc" variable names should appear in your DATA Editor; you can now edit the data, add cases/variables, process it, whatever.

If you wish to retrieve your saved example output, the procedure is similar to getting your data. First you "click" "File", and then "Open"; in the "Open File" dialog box that pops up, you press backspace to clear the "File Name" sub box and this time key "A:*.spo"; then hit ENTER; this should list all files with "spo" extensions (the default extension for your output files); now "click" "myspss.spo" (the file containing your saved example output) and then "click" it again or follow the usual "ENTER/OK" routine. Your example case summary and frequencies output should now emerge in the OUTPUT Viewer. You can edit it, add on to it with more processing, print it, save it again, etc.

 
 
 
 

Appendix A 
More about SPSS/Windows Release 10.0 
GoBack

I. Basic Items of Interest GoBack

Name: Statistical Package for the Social Sciences/ Windows, Release 8.0 
Function: Data analysis/management with an emphasis on statistical applications in a PC Windows Environment 
Sample Oper: Retrieving Windows SPSS Saved Data File "A:mydata.sav"
Click (select) "FILE" from main menu, then "OPEN", then "DATA", then "mydata.sav" in the file list of the OPEN dialog box (clicking "A" drive before hand if "A" not default), and then click "OK" to activate. 
Reference: SPSS Base 8.0 for Windows User's Guide, copyright 1998 by SPSS Inc, Chicago, Il.
Options:
  • Processing by running SPSS stmts developed in a SYNTAX Window instead of making Menu/dialog box selections. 
  • Creating SPSS stmts by pasting the results of menu/dialog box selections in a SYNTAX window. 
  • Interactively creating charts/graphs. 
  • Editing charts/graphs in the Chart Editor window. 
  • Pasting output in another windows applications such as Word Perfect. 
  • Exporting output in HTML format, charts in a variety of formats. 
  • Creating typewriter format output in the Draft Viewer (charts/graphs remain hi-resolution). 
  • Creating scripts in a Script window and using the programming to customize the SPSS output.
  • II. PC Computer Resource Requirements for SPSS/Windows 10.0 GoBack

    The following lists the minimum computer requirements for SPSS/Windows Release 10.0:
    1. Windows 95, 98, Windows NT 4.0, or Windows 2000.
    2. Pentium or Pentium-class processor, 90mhz or faster.
    3. 64MB recommended.
    4. 80 MB of hard drive disk space.
    5. SVGA Monitor (or any graphics displaying monitor with 800 by 800 resoution)
    III. Differences Between Versions of SPSS/Windows: 9.0 vs. 8.0 and 10.0 vs. 9.0
    1. SPSS/Windows Version 9.0 vs. 8.0
      1. Some menu items are renamed/redistributed
      2. The OLAP Cubes procedure provides the layer reports (descriptive statistics by subgroups)
      3. GLM (General Linear Model) is univariate (formerly multivariate, now available as an SPSS extra).
      4. Interactive graphics is enhanced.
      5. Drafter viewer output improved
      6. Statistical procedures added/enhanced
        1. The Reliability Analysis and Multidimensional Scaling (ALCAL) procedures are added.
        2. A matrix language procedure is added.
        3. The new ROC Curve procedure is added (evaluates classification methods where two category variables classify subjects as default/nodefault potential on a bank loan).
        4. The new Nominal Regression procedure is added (category variables related with multiple independent variables, a Regression Model option)
        5. The Cochran-Mantel-Haenzal statistic is added to the Crosstabs procedure
      7. The "Wizard" is added for easy reading of text files, the input format specified by just pointing, clicking, and/or dragging
    2. Version 10.0 vs. 9.0
    1. Large data sets no longer require large amounts of temporary disk space to analyze data (i.e., SPSS no longer creates a temp work file copy of the input data for making passes to do the analysis -- exceptions: the input data is a text or Excel file or the processing requires sorting the data).
    2. A remote computer can be used to do the SPSS analysis without tying up the desk top (requires installation of software and the data on the remote computer).
    3. Multiple SPSS sessions are possible, permitting opening and processing more than one data file at one time.
    4. Excel 5 files are read directly into SPSS with no special Excel ODBC drivers necessary; just select "Excel" as the file type in the "Open File" box.
    5. The "Data Editor" has a Variable View mode, making it easy to view/define data variables (define name, type, label -- an extensive name, value labels, etc.).
    6. The ROC Curve procedure is expanded to permit comparing multiple test variables (formerly only one test variable); ROC evaluates variables -- often probibilities from discriminate analysis, logistic regression, etc. -- for the ability to classify a subject in one of two categories (e.g., whether or not a bank client would default on a loan).
    7. Interactive Graphics is improved for using in other applications and faster printing.
  • Polytomous Logit Universal models (PLUM) lets you use regression techniques to ordinal outcomes (e.g., low, medium, and high). Accessed via the Advanced Models option.
  • Thermatic mapping added -- graphically summarizes data by geographic areas, using the chart types "bar", "pie", "range of value", etc. Accessed via the new Maps option.
  • Category Principle Component Analysis (CATPCA), a new optional scaling procedure, is available in the Categories option.
  • Logistic/Cox Regression (via Regression Models and Cox Regression options) has improved output (high-quality, flexible, pivot table).