IBM SPSS 21 manuel d'utilisation

i IBM SPSS Statistics 21 Brief Guide[...]

Note : Before using this informat ion and the produ c t it supports, read the general information under Notices on p. 156. This edition applies to IBM® SPSS ® S tatistic s 21 and to all subsequent releases and modi fi cations until ot herwise indic ated in new editions. Adobe produ ct screensho t(s) reprinted with permi ssion from A dobe Systems[...]

Preface The IBM SPSS Statistics 21 Brief Guide provides a set of tutorials designed to acquaint you with the various components of IBM® SPSS® Statistics. This guide is intended for use with all operating system versions of the software, including: W indo ws, Macintosh, and Linux. Y ou c a nw o r kt h r o u g ht h et u t o r i a l si ns e q u e n [...]

Additionally , SPSS Statistics Ba se offers a broad range of algor ithms for compar ing means and predictive techniques such as t-test, analysis of var iance , linear regression and ordinal regression. Advanced Statistics focuses on techniques often used in sophisticated experimental and biomedical research. It includes procedures for general linea[...]

Regression provides te chniques for analyzing data that do no t fi t traditional linear statistical models. It inclu des procedu res for probit analysis, logistic regression, weight estimatio n, two-stage least-squares regression, and general nonlinear regression. Amos™ ( a nalysis of mo ment s tructures) uses structural equation modeling to con[...]

 SPSS Stati stics command synt ax is not availabl e to the u ser . This means that it is not possible to repeat an analysis by saving a series of commands in a syntax or “job” fi le, as can be done in the full version of IBM® SPSS® Statistics.  Scripting and automation are not available to the user . This means that you cannot create s[...]

Contents 1 Introduction 1 S a m p l e F i l e s .................... ............... ............ ............... ... 1 O p e n i n g a D a t a F i l e ......... ............ ............... ............... ......... 1 R u n n i n g a n A n a l y s i s .... ............... ............ ............... ............ 3 V i e w i n g R e s u l t s ....[...]

C o p y i n g a n d P a s t i n g I n f o r m a t i o n b e t w e e n D a t a s e t s ..... ............... ............... 51 R e n a m i n g D a t a s e t s ... ............... ............... ............... ............ 51 S u p p r e s s i n g M u l t i p l e D a t a s e t s ......... ............... ............ ............... 51 5 Examining[...]

H i d i n g R o w s a n d C o l u m n s........................ ............... ............ 89 C h a n g i n g D a t a D i s p l a y F o r m a t s .. ............... ............... ............... 89 T a b l e L o o k s ............... ............... ............... ............ ......... 91 U s i n g P r e d e f i n e d F o r m a t s ......... [...]

S p l i t - F i l e P r o c e s s i n g ........ ............... ............... ............... ..... 139 S o r t i n g C a s e s f o r S p l i t - F i l e P r o c e s s i n g ................. ............... ........ 140 T u r n i n g S p l i t - F i l e P r o c e s s i n g O n a n d O f f .......... ............ ............... .. 141 S e l e c[...]

Chapter 1 Introduction This guide provides a set of tutorials designed to enable you to perform useful analyses on your data. Y ou can work through the tutorials in sequence or turn to the topics for which you need additional information. This chapter will introduce you to the basic features and demonstrate a typical session. W e will retrieve a pr[...]

2 Chapt er 1 By default, IBM® SPSS® Statistics data fi le s (. sav extension) are displayed. This exam ple uses the fi le demo.sav . Figure 1 -2 demo.sav file in Dat a Editor The data fi le is displayed in the Data Editor . In the Data Editor , if you put the mouse cursor on a variable name (the colum n head ings), a more descriptive variab le[...]

3 Introduction Figure 1 -4 V a l u el a b e l sd i s p l a y e di nt h eD a t aE d i t o r R u n n i n ga nA n a l y s i s If you have any add-on options, the Analyze menu contains a list of reporting and statistical analysis categories. W e will start by creating a simple frequency tab le (table of counts). This example requires the Statisti cs B [...]

4 Chapt er 1 An icon nex t to each variab le provides in formation a bout data type and level of measurement. Numeric String Date Tim e Scale (Continuous) n/a Ordinal Nominal E Click the variable Income category in thousands [inccat] . Figure 1 -6 Va r i a b l e l a b e ls and names in the F requencies dialog box If the variable label and/or name a[...]

5 Introduction Figure 1 -7 Resized dialo g box In the dia log box , yo u choose the variables that you want to anal yze from the source list on the left and d rag and drop them into the V ariable(s) list on th e rig ht. The OK button, which runs the analysis, is disabled until at least one variable is placed in the V ariable(s) list. In many dialog[...]

6 Chapt er 1 E Click Gender [gender] in the source variable list and drag the variable into the target V ariable(s) list. E Click Income category in thousands [inccat] in the source list and drag it to the target list. Figure 1 -9 V ariables selected for analysis E Click OK to run the procedure. Vi ewing R esults Figure 1 -1 0 Vie wer window Result[...]

7 Introduction E Click Income cate go ry in thou sa nd s [inc cat] . Figure 1 -1 1 F requen cy t able of income categories The frequency table for income categories is displayed. This frequency table shows the number and percentage of people in each income category . Creating Charts Although some statistical procedures can cr eate charts, you can a[...]

8 Chapt er 1 Figure 1 -12 Chart Builder dialog box E Scroll down the V ariables list, right-click W ir e less service [wireless] , and then choose Nominal as its measurement level. E Drag the W ireless service [wir eless] variable to the x axis. E Right-click Owns PDA [ownpda] and c hoose Nominal as its measu rement level. E Drag the Owns PDA [ownp[...]

9 Introduction Figure 1 -13 Bar c hart display ed in View er wind ow The bar chart is displayed in the V iewer . The chart shows that people with wireless phone service are far m ore likely to have PDAs than peo ple without wireless service. Y ou can edit charts and tables by double-clic king them in the contents pane of the V iewer window , and yo[...]

Chapter 2 Reading Data Data can be entered direct ly , or it can be import ed from a number of different s ources. Th e processes for reading data stored in IBM® SPSS® Statistics data fi les; spreadsheet applications, such as Microsoft Excel; database applicat ions, such as Microsoft Access; and text fi les are all discussed in this chapter . B[...]

11 Reading Dat a The data are now displayed in the Data Editor . Figure 2-2 Opened dat a file Reading Data from Spreadsheets Rather than typing all of your data directly into the Data Editor , you can read data from applications such as Microsoft Excel. Y ou can also read column headings as variable names. E From the menus choose: File > Open &g[...]

12 Chapt er 2 E Make sure that Re ad variable names from the fi rst row of data is selected. This option reads column headings as variable names. If the column headings do not conform to the IBM® SPSS® Statistics variable-naming rules, they are converted into valid variable names and the original column headings are saved as variable labels. If y[...]

13 Reading Dat a Figure 2-5 Dat a base W izard Welcome dialog box E Sele ct MS Access Database from the list of data sources and click Next . Note : Depending on your installation, you may also s ee a li st of OLEDB data sources on the left side of the wizard (W indows operating systems only), but this example uses the list of ODBC data sourc es di[...]

14 Chapt er 2 Figure 2-6 ODBC Driver Login dialog box E Click Browse to navigate to the Access database fi le that you want to open. E Open demo.mdb . For more information, see the topic Sample Files i n Appendix A on p. 1 47. E Click OK in the login dialog box. In the next step, you can specify the tabl es and variables that you want to import. F[...]

15 Reading Dat a E Drag the entire demo table to the R etrieve Field s In This Order list . E Click Nex t . In the next step, you select which records (cases) to import. Figure 2-8 Limit Retriev ed Cases step If you d o not want to import all cases, you can import a subset of cases (for example, males older than 30), or you can import a random samp[...]

16 Chapt er 2 Field names are used to create variable names. If necessary , the names are converted to valid variable names. The original fi eld names are preserved as variable labels. Y ou can also change the variable names before importing the database. Figure 2-9 Define V ariables step E Click the Recode to Numeric cell in the Gender fi eld. T[...]

17 Reading Dat a The SQL statement created from your selections in the Database W izard appears in the Results step. This statement can be executed now or saved to a fi le for later use. Figure 2-1 0 Resu lt s s tep E Clic k Finish to import the data.[...]

18 Chapt er 2 All of the data in the Access database t hat you se lected to import are now available in the Data Editor . Figure 2-1 1 Data imported from an Access database Reading Data from a T ext File Te x t fi les are another common source of data. Many spreadsheet programs and databases can save their contents in one of many text fi le forma[...]

19 Reading Dat a The T ext Impo rt Wizard gui des you th rough the process of de fi ning how the speci fi ed text fi le should be interpreted. Figure 2-12 T ext Import W izard: Step 1 of 6 E In Step 1, you can choose a prede fi ned format or create a new format in the wizard. Select No to indicat e that a new forma t should be crea ted. E Click[...]

20 Chapt er 2 As stated earlier , this fi le uses tab-delimited formatting . Also, the variable names are d e fi ned on t he t op li ne of this fi le. Figure 2-13 T ext Import W izard: Step 2 of 6 E Select Delimited to indicate that the data use a delimited formatting structure. E Select Ye s to indicate that variable names should be read from t[...]

21 Reading Dat a E Ty p e 2 in the top section of next dialog box to indicate that the fi rst row of data starts on the second line of the text fi le. Figure 2-1 4 T ext Import W izard: Step 3 of 6 E Keep the default values for the remainder of this dialog box, and click Next to continue.[...]

22 Chapt er 2 The Data preview in Step 4 pr ovides you with a quick w ay to ensure that your data are bein g properly read. Figure 2-15 T ext Import W izard: Step 4 of 6 E Select Ta b and d eselect the o ther options. E Click Ne xt to continue .[...]

23 Reading Dat a Because the variable names may have been truncated to fi t formatting requirements, this dialog box gives you the opportunity to edit any undesirable names. Figure 2-16 T ext Import W izard: Step 5 of 6 Data types can be de fi ned here as well. For example, it’ s safe to assume that the income variable is meant to contain a cer[...]

24 Chapt er 2 E Select Dollar from t he Data format drop-down list. Figure 2-17 Change the data t ype E Click N ex t to c ontinue.[...]

25 Reading Dat a Figure 2-18 T ext Import W izard: Step 6 of 6 E Leave the default selections in this dialog b ox, and click Finish to import the d ata.[...]

Chapter 3 Using the Data Editor The Data Editor displays the contents of the active data fi le. The information in the Data Editor consists of variables and cases.  In Data V iew , columns represent variables , and rows represent c ases (observations).  In V ariable V iew , each row is a variable, and each column is an attribute that is asso[...]

27 Using the Dat a Editor E In the second row , type mari tal . E In the third row , type income . New variables are automatically given a Numeric data type. If you don’t enter variable names, unique names ar e a utomatically created. However , these names are n ot descri ptive and are not recommended for large data fi les. E Click the Data Vie [...]

28 Chapt er 3 Currently , the age and marital columns display decimal points, even though their values are intended to be integers. T o hide the decimal points in these variables: E Click the Va r i a b l e V i e w tab at the bott om of the D ata Editor wi ndow . E In the Decimals column of the age row , type 0 to hide the decimal. E In the Decimal[...]

29 Using the Dat a Editor E Click the button on the right side of the Ty p e cell to open the V ariable T ype dialog box. Figure 3-4 B u t t o ns h o w ni nT y p ec e l lf o rs e x E Select String to specify the variable type. E Click OK to save your selection and return to the Data Editor . Figure 3-5 V ariable T ype dialog box[...]

30 Chapt er 3 Defining Data I na d d i t i o nt od e fi ning data t ypes, you can also de fi ne descriptive variable labels and value labels for variable names and data values. These descrip tive labels are used in statistical reports and charts. Adding V ariable Labels Labels are meant to provide descriptio ns of vari ables. These descrip tions [...]

31 Using the Dat a Editor Changing V ariable T ype and Format The Ty p e column displays the cur rent data type for each variable. The most co mmon data types are numeric and string, but many other fo rmats are supported. In the c ur rent data fi le, the income variable is de fi ned as a numeric type. E Click the Ty p e ce ll fo r t he income row[...]

32 Chapt er 3 E Click Add to add thi s lab el to the l ist. Figure 3-8 V alue Labels dialog box E Ty p e 1 in the V alue fi eld, and type Married in the Label fi eld. E Click Add , and then c lick OK to s ave your changes and return to the Data Edi tor . These labels can also be displayed in Data V iew , which can make your data more readable. E [...]

33 Using the Dat a Editor If the V alue Labels menu item is already ac tive (with a check mark next to it), choosing Va l u e Labels again will turn off the display of value labels. Figure 3-9 V a l u el a b e l sd i s p l a y e di nD a t aV i e w Adding V alue Labels for String V ariables String variables may require value labels as well. For exam[...]

34 Chapt er 3 E Click Add to add this label to your data fi le. Figure 3-1 0 V alue Labels dialog box E Ty p e M in the V alue fi eld, and type Male in the Lab el fi eld. E Click Add , and then c lick OK to s ave your changes and return to the Data Edi tor . Because string values are case sensitive , you shou ld be consisten t. A lowercase m is [...]

35 Using the Dat a Editor E Click the button on the right side of the cell, and then choose Fe ma l e from th e drop-down list. Figure 3-1 1 Using variable labels to select values Only de fi ned valu es are listed, whi ch ensures that the entered dat a are in a format that you expect. Handling Missing Data Missing or invalid data are generally too[...]

36 Chapt er 3 Figure 3-12 Missing values d isplay ed a s p eriods The reason a value i s missing may be i mportant to yo ur an alysis. For exampl e, y ou may fi nd it useful to distinguish between those respondents who refused to answer a question and those respondents who didn’t answer a que stion because it was not applicable. Missing V alues [...]

37 Using the Dat a Editor E Select Discrete missing values . E Ty p e 999 in the fi rst text box and leave the other two text boxes empty . E Click OK to save your changes and r eturn to the Data Editor . Now that the missing data value has been added, a label can be applied t o that value. E Click the Va l u e s cell in the age row , and then cli[...]

38 Chapt er 3 E Click OK to save your changes and r eturn to the Data Editor . Now you can add a label for the missing value. E Click the Va l u e s cell in the sex r o w ,a n dt h e nc l i c kt h eb u t t o no nt h er i g h ts i d eo ft h ec e l lt o open the V alue Labels dialog box. E Ty p e NR in th e V alue fi eld. E Ty p e No Response in the[...]

39 Using the Dat a Editor E In V ariable V iew , type agewed in the fi r s tc e l lo ft h e fi rst empty row . Figure 3-16 agew ed variable in V ariable Vie w E In the Label column, type Age Marr ied . E Click the Va l u e s cell in the age row . E From the menus choose: Edit > Cop y E Click the Va l u e s cell in the agewed row . E From the m[...]

40 Chapt er 3 T o apply t he attribute to multiple variables, simply select multiple targ et cells (click and drag down the column). Figure 3-17 Multiple cells selected When you paste the attribute, it is applied to all of the selected cells. New variables are automatically created if you paste the values into empty rows.[...]

41 Using the Dat a Editor T o copy all attributes from one variable to another variable: E Click the row number in the marital row . Figure 3-18 Selected row E From the menus choose: Edit > Cop y E Click the row number of the fi rst empty row . E From the menus choose: Edit > Pas te[...]

42 Chapt er 3 All attribu tes of the marita l variable are applied to t he new variabl e. Figure 3-1 9 All values pasted into a row Defining Variable Propertie s for Categorical Variables For categorical (nominal, ordinal) data, you can use De fi ne V ariable Properties to de fi ne value labels a nd other variable properties. The De fi ne V aria[...]

43 Using the Dat a Editor Figure 3-20 Initial Define V ariable Properties d ialog box In the initial De fi ne V ariable P roperties dialog box, you sel ect the nominal o r ordinal v ariables for which you want to de fi ne value labels and/or other properties. E Drag and drop Owns TV [owntv] throug h Owns fa x machine [ownfax] int o the V ariables[...]

44 Chapt er 3 Figure 3-21 Define V ariable Prope rties main dialog box E In the S canned V ariabl e List, select ownpc . The current level of measurement for the selected variable is scale. Y ou can change the measurement level by selecting a level f rom the drop-down list, or you can let De fi ne V ariable Propert ies suggest a measurement level.[...]

45 Using the Dat a Editor Figure 3-22 Suggest Measurement Lev el dialog box Because the variable doesn’ t have v ery many different values and all of the scanned cases contain integer values, the proper measurement level is probably ordinal or nominal. E Select Ordinal , and then click Continue . The measurement level for the selected variable is[...]

46 Chapt er 3 Figure 3-23 New variable properties defined for ownpc Before w e compl ete the job of modifyi ng the variable properties for ownpc , let’ s apply the same measurement level, value labels, and missing values de fi nitions to the other variables in the list. E In the Copy Properties area, click T o Other V ariables .[...]

47 Using the Dat a Editor Figure 3-24 Apply Labels and Level dialog box E In the Apply Labels and Level dialog box, select all o f the variables in the list, and then click Cop y . If you select any other varia ble in the Scanned V ariable List of the De fi ne V ariable Properties main dialog box now , you’ll see that they are all ordinal variab[...]

48 Chapt er 3 E Click OK to save al l of the vari able properties that you have de fi ned.[...]

Chapter 4 W orking with Multiple Data Sources Starting with version 14.0, multiple data sources can be open at the same time, making it easier to:  Switch back and forth between data sources.  Compare the co ntents of di fferent data sour ces.  Copy and paste data between data sources.  Create multiple subsets o f cases and/or variables[...]

50 Chapt er 4  Only the variables in the active dataset are available for analysis. Figure 4-2 V ariable list cont aining variables in the active dataset  Y ou cannot change the active dataset when any dialog box that accesses the data is open (including all dialog boxes that display variable lists).  At least one Data Editor window must b[...]

51 W orking with M ultiple Data Sources Figure 4-3 Open dat asets display ed on syntax window toolbar Copying a nd Pasting Information between Datasets Y ou can copy bo th data an d variabl e de fi nition attributes from one dataset to another dataset in basically the same way that you copy and paste information within a single data fi le.  Co[...]

52 Chapt er 4 E Click the General tab. Select (check) Open only one dataset at a time .[...]

Chapter 5 Examining Summary Statistics for Individual V ariables This chapter discusses simple summary measures and how the level of measurement of a variable in fl uences the types of statistics that should be used. W e will use the data fi le demo.sav .F o r m o r e information, see the topic Sample Files in Ap pendix A o n p . 147. Level of Me[...]

54 Chapt er 5 E Select Owns PDA [ownpda ] and Owns TV [owntv] and move them into the V ariable(s) list. Figure 5-1 Categorical variables selected for analysis E Click OK t o run the p rocedure. Figure 5-2 F requen cy tab les The frequency tables are displayed in the V ie wer w indow . The frequency tables reveal that only 2 0.4% of t he p eople own[...]

55 Examining Summ ar y Statistics for Individual V ariables Charts for Categorical Data Y ou can graphically display the information in a frequency table with a bar chart or pie chart. E Open the Frequencies dialog box again. (The two variables should still be selected.) Y ou can use t he Dialog Recall button on the toolbar to quickly return to rec[...]

56 Chapt er 5 Figure 5-5 Bar c h ar t In addition to the frequency tables, the same information is now displayed in the form of bar charts, making it easy to see that most people do not own PDAs but almost everyone owns a TV . Summary Measures for Scale V ariables There are many summary measures available for scale variables, including:  Measure[...]

57 Examining Summ ar y Statistics for Individual V ariables E Select Household income i n thousan ds [income] a n dm o v ei ti n t ot h eV a r i a b l e ( s )l i s t . Figure 5-6 Scale variable selected for analysis E Click Statistics . E Select Mean , Median , Std. deviation , Minimum ,a n d Maximum . Figure 5-7 F requen cies S tatistics dial og b[...]

58 Chapt er 5 E Click OK to run the procedure. The Frequencies Statistics table is displayed in the V iewer window . Figure 5-8 F requen cies Statistics table In this example, there is a lar ge dif ference between the mean and the median. The mean is almost 25,000 greater than the media n, in dicating that the values are not normall y distribu ted.[...]

59 Examining Summ ar y Statistics for Individual V ariables Figure 5-1 0 Histo gram The majo rity of ca ses are c lustered at the lowe r en d of the scale, with most falling bel ow 100,000. There are, however , a few cases in the 500,000 range and beyond (too few to even be visible without modifying the histogram). These high values for only a few [...]

Chapter 6 Creating and editing charts Y ou can c reate and e dit a wide varie ty of chart types. In this chapt er , we will create and edit bar charts. Y ou can apply the principles to any chart type. Chart creation basics T o demonstrate the basics of chart creation, we w ill create a b ar chart of mean income for different levels of jo b satisfac[...]

61 Creating and editing charts The Chart Buil der dialog box is an interactive window that allows you to preview h ow a chart will look while you build it. Figure 6-1 Chart Builder dialog box Using the Chart B uilder gallery E Click the Galler y tab if it is not selected. The Gallery includes many different p rede fi n e dc h a r t s ,w h i c ha r[...]

62 Chapt er 6 E Drag the ico n for the si mple bar ch art onto th e “can vas,” which is the large area above the Gallery . The Chart Builder displays a preview of the chart on the canvas. Note that the data used to draw the chart are not your actual data. They are example data. Figure 6-2 Bar c hart on Chart Builder canvas Defining variables an[...]

63 Creating and editing charts E Right-click Job satisfaction in the V ariables list and choose Ordinal . Ordinal is an appropriate measurement level because the cate gories in Job sa tisfaction can be ranked by le vel of satisfaction. Note that the icon changes after you change the measurement level. E Now d rag Job satisfaction from the V ariable[...]

64 Chapt er 6 Figure 6-4 Element Properties window The Element Properties window allows you to change the properties of the various chart elements. These elements include the graphic elements (suc h as the bars in the b ar chart) and the axes on the chart. Select one of the elements in the Edit Prope rties of list to change the properties associate[...]

65 Creating and editing charts Adding text Y ou can also add titles and footnotes to the chart. E Click the Titles/Footnotes tab. E Select Title 1 . Figure 6-5 T i t l e1d i s p l a y e do nc a n v a s The title appears on the canvas with the label T1 . E In the Elem ent Properties window , select Title 1 in the Edit Properties of list. E In the Co[...]

66 Chapt er 6 Creating the chart E Click OK to create the bar chart. Figure 6-6 Bar c h ar t The bar chart reveals that respondents who are more satis fi ed with their jobs tend to have higher household incomes. Chart editing basics Y ou can edit charts in a variety of ways. For th e sample bar chart that you created, you will:  Change colors. [...]

67 Creating and editing charts E Double-click the bar chart to open it in the Chart Editor . Figure 6-7 Bar chart in the Chart Editor Selecting chart elements T o edit a chart element, you fi rst select it. E Click any one of the bars. The rectangles arou nd the bars indicate that they are selected. There are general ru les for selecti ng element [...]

68 Chapt er 6 This opens the Pro perties w indow , showi ng the tabs that apply to the bars you selected. These t abs change d epending on what chart elemen t you select in the Chart Editor . For example, if you had selected a t ext frame instead of bars, different tabs would appear in the Propert ies window . Y ou will use these tabs to do most ch[...]

69 Creating and editing charts Figure 6-9 Fill & Border t ab E Click Apply . The bars in the chart are now light blue. Figure 6-1 0 Edited bar c hart showin g b lue bars[...]

70 Chapt er 6 Formatting numbers in tick labels Notice t hat the numbers on the y axis are scaled in t housands. T o make the chart more attractive and easier to int erpret, we w ill change the numb er format in the tick labels a nd then edit the axis title appropriately . E Select the y axis tick labels by clicking any one of them. E T o reop en t[...]

71 Creating and editing charts Figure 6-1 1 Number Format tab E Click Apply .[...]

72 Chapt er 6 The tick labels re fl ect the new number formatting: There are no decimal places, the numbers are no longer scaled, and each thousandth place is speci fi ed with a character . Figure 6-12 Edited bar c hart showing new nu mber format Editing t ext Now that y ou have ch anged the number format of the tick labels, the axis title is no [...]

73 Creating and editing charts E Delete the following text: in thousands E Press Enter to exit edit mode and update the axis t itle. The axis title now accurately describes the contents of the tick labels. Figure 6-13 Bar c hart showing edited y axis title Displaying data value labels Another common task i s to show the ex act values associated wit[...]

74 Chapt er 6 Figure 6-1 4 Bar c hart showing data value labels Each bar in the chart now displays the exact mean household income. Notice that the units are in thousands, so you c ould use the Number Format tab ag ain to change the scal ing factor . Using templates If you m ake a number of routine chang es to yo ur charts , you can use a chart tem[...]

75 Creating and editing charts Y ou can also enter a description of the template. This description will be visible when you apply the template. Figure 6-15 Sav e Chart T emplate dialog box E Click Continue . E In the Save T empl ate dialog box, specify a location and fi lename for the template. E When you are fi nished, click Sav e . Y ou can app[...]

76 Chapt er 6 E Close the Chart Editor . The updated bar chart is shown in the V iewer . Figure 6-16 Updated bar c h art in Viewer E From the V iewer menus choose: Graphs > Char t Bu ild er ... The Chart Builder dialog box “remembers” t he variables that you entered w hen you created the original chart. However , here you will create a sligh[...]

77 Creating and editing charts Figure 6-17 Options dialog box with te mplate specified The Options dialog box displays the fi le path of the template you selected. E Click OK to close the Options dialog box.[...]

78 Chapt er 6 Figure 6-18 Chart Builder with completed drop zones E Click OK on the Chart Builder dialog box to create the chart and apply the template. The formatting in the new chart matches the formatting in the chart that you previously created and edited. Although the variables on the x axis are different, the charts otherwise resemble each ot[...]

79 Creating and editing charts If you want to apply templates after you’ve created a chart, you can do it in the Chart Editor (from the File menu, choose Apply Char t T emplate ). Figure 6-1 9 Updated bar c h art in Viewer Defining chart options In addition to using templates to format charts, y ou can use the Options to control var ious aspects [...]

80 Chapt er 6 The Options dialog box contains many con fi guration sett ings. Click t he Charts t a bt os e et h e available options. Figure 6-20 Charts tab in Options dialog box The options control how a chart is created. For each new chart, you can specify:  Whether to use the current settings or a template.  The width-to-heig ht ratio (as[...]

81 Creating and editing charts For a simp le chart, the Chart Edit or uses o ne style that y ou specify . For grouped charts, the Chart Editor uses a set of styles that it cycles through for each group (category) in the chart. E Select Simple Char ts . E Select the lig ht green colo r , w hich is thi rd from the right in t he second ro w from the b[...]

82 Chapt er 6 The bars in the new chart are light green. This chart also differs from the last one in other ways. There is n o titl e; the axis labels are in t housands; and there are no data labels. The di fferences occurred because the template wasn’t applied to this chart. Figure 6-22 U p d a t e db a rc h a r ti nV i e w e rw i t h o u tat e [...]

Chapter 7 W orking with Output The results from running a statistical procedure are displayed in the V iewer . The output produced can be statistical tables, charts, graphs, or text, depending on the choices you make when you run the procedure. This section uses the fi les viewertut.spv and demo.sav . For more information, see the topic Sample Fil[...]

84 Chapt er 7 The open book icon changes to a c losed b ook icon, signifying th at th e in formation associated with it is no w h idden. E T o redisplay the hidden output, d ouble-click the closed book icon . Y ou can also h ide all of the ou tput from a particular statistical procedure or all of the output in the V iewer . E Click the box with the[...]

85 W orking with Output E Double-cli ck th e Owns PDA * Gender * Internet Cr osstabulation ta ble. E Right-click Expected Count and choose What’ s This? f rom the pop-up c ontext menu . The de fi nition is displayed in a pop-up window . Figure 7 -3 P op-up definition Pivoting T ables The d efault tables produced may not display information as ne[...]

86 Chapt er 7 Pivoting trays provide a way to move data between columns, rows, and layers. Figure 7 -4 Pivoting tray s E Drag the Statisti cs element from the Row dimension to the Column dimension, below Gender . The table is immediat ely recon fi gured to re fl ect your changes. Figure 7 -5 Moving rows to column s The order of the elements in th[...]

87 W orking with Output E Drag and drop the Owns PDA element before the Internet element in the row dimension to reverse the order of these two rows. Figure 7 -6 Swap ro ws Creating and Displaying Layers Layers can be useful for large tables with nested categories of information. By creating layers, you simplify the look of the table, making it eas[...]

88 Chapt er 7 T o display a differen t layer , select a category from the drop-down list in the table. Figure 7 -8 Choosing a lay er Editing T ables Unless you’ve taken the time to create a custom T ableLook, pivot tables are created with standard formatting. Y ou can change the formatting of any text within a table. Formats that you can change i[...]

89 W orking with Output Note : If you change the values in a table, totals and other statistics are not recalculated. Hiding Rows and Columns Some of the data displayed in a table may not be useful or it may unnecessarily complicate the table. Fortunately , you can hide en tire row s and c olumns without l osing any d ata. E If it’ s not already [...]

90 Chapt er 7 E Ty p e 0 in the Decimals fi eld to hide all decimal points in this column. Figure 7 -1 1 Cell Properties, Format V alu e tab Y ou can also chan ge the data type and forma t in th is dialog box. E Select the type that you want from the Category list, and then select the format for that type in the Format list. E Click OK or Apply to[...]

91 W orking with Output T ableLooks The format of your tables is a critical part of prov iding clear , concise, and meaningful results. If your tabl e is dif fi cult to read, the information contained within that table may not be easily understood. Using Predefined Formats E Double-click the Marital status table. E From the menus choose: Format &g[...]

92 Chapt er 7 Customizing T ableLook Styles Y ou can customize a format to fi t you r speci fi c needs. Almost all aspects o f a table can be customized, from the background color to the border styles. E Double-click the Marital status table. E From the menus choose: Format > T ableLooks... E Select the style that is closest to your desired fo[...]

93 W orking with Output E Then select a new text color . T h eS a m p l ew i n d o ws h o w st h en e ws t y l e . Figure 7 -1 5 Changing table cell formats E Click OK to return to the T ableLooks dialog box. Y ou can save your new style, which allows you to apply it to future tables easily . E Click Sav e As . E Navigate to the desired target dire[...]

94 Chapt er 7 The table now contains the custom formatting that you speci fi ed. Figure 7 -1 6 Custom T ableLook Changing t he Default T able Formats Although y ou can change the format of a table after it has been created, it may be more ef fi cient to change the default T ableLook so that you do not have to change the format every time you crea[...]

95 W orking with Output E Click the Piv ot T ab les tab in the Options dialog box. Figure 7 -1 7 Options dialog box E Select the T ableLook style that you want to use for all new tables. The Sample window on the right shows a preview of each T ableLook. E Click OK to save your settings and close the dialog box. All tables that you create after chan[...]

96 Chapt er 7 Customizing the Initial Display Settings The initial display settings include the alignmen t o f objects in the V iewer , whether objects are shown or hidden by default, and the width of the V iewer window . T o change these settings: E From the menus choose: Edit > O ptions... E Click the Viewer tab. Figure 7 -1 8 Vie wer options [...]

97 W orking with Output Y ou can also hide elements, such as the log and warning messages, that tend to clutter your output. Double-clickin g on an icon automatically changes that object’ s display property . E Double-cli ck th e Wa r n i n g s icon to hi de warni ng m essages in the output . E Click OK to save yo ur chan ges and close the dialog[...]

98 Chapt er 7 Y ou can specify differe nt settings for the outline and contents panes. For example, to show labels in the outline and variable names and data values in the contents: E In the Pivot T able Labeling group, select Names from the V ariables in Labels drop-down list to show variab le name s instead of labels. E Then, select Va l u e s fr[...]

99 W orking with Output Using Results in Other Applications Y our results can be used in many applications . For example, you may want to include a table or chart in a presentation or report. The following examples are speci fi c to Microsoft W ord, but they may work similarly in other word processing applications. Pasting Results as W ord T ables[...]

10 0 Chapt er 7 Figure 7 -23 Pivot t able display ed in Word The table is now displayed in your document. Y ou can apply custom formatting, edit the data, and resize the table to fi t your needs. Pasting Results as T ext Pivot tables can be copied to other applications as plain text. Formatting styles are not retained in this metho d, but you can [...]

101 W orking with Output E Select Unfo rmatted T e xt in the Paste Special dialog box. Figure 7 -24 Paste Special dialog box E Click OK to paste your results into the current document. Figure 7 -25 Pivot t able display ed in Word Each column of the table is separated by tabs. Y o u can change the column widths by adjusting the tab stops i n your wo[...]

10 2 Chapt er 7 Note : Export to PowerPoint is available only on Wi ndows operating systems and is not available with the Student V ersion. In the V iewer ’ s o utline pane, you can select speci fi c items that you want to export or export all items or a ll v isible items. E From the V iewer menus choose: File > Expor t... Instead of exportin[...]

10 3 W orking with Output E Click OK to generate the W ord fi le. When you open the resulting fi le in W ord, you can see how the results are exported. Notes, which are not visible objects, appear in W o rd because you chose to export all objects. Figure 7 -27 Output.doc in W ord[...]

10 4 Chapt er 7 Pivot tables beco me W ord tables, with all of the fo rmatting of t he original pivot table retained, including fonts, c olors, borders, and so on. Figure 7 -28 Pivot tables in Word Charts are included in the W ord do cument as g raphic images. Figure 7 -29 Charts in Word[...]

10 5 W orking with Output T ext output is disp layed in t he same fon t used for the text obj ect in the V iewer . For proper alignment, text output should use a fi xed-pitch (monospaced) font. Figure 7 -30 T ext o utput in Word[...]

10 6 Chapt er 7 If yo u exp ort to a P owerPoin t fi le, each exported item is placed on a separate slide. Pivot tables exported to PowerPoint become W o rd tables, with all of the formatting of the original pivot table, including fonts, colors, borders, and so on. Figure 7 -31 Pivot table s in P owerP oint[...]

10 7 W orking with Output Charts selected for expor t to PowerPoint are em bedded in the PowerPoint fi le. Figure 7 -32 Charts in P ow erP oint I fy o ue x p o r tt oa nE x c e l fi le, results are exported differen tly . Figure 7 -33 Output.xls in Excel[...]

10 8 Chapt er 7 Pivot table rows, columns, and cells become Excel rows, columns, and cells. Each line in the text output is a row in the Excel fi le, with the entire contents of the line contained in a single cell. Figure 7 -34 T ext output in Excel Exporting Results to PDF Y ou can export all or s elected i tems in t he V iewer t o a PDF (portabl[...]

10 9 W orking with Output E In the Export Output dialog box, f rom the Export Format File T ype drop-down list choose P or table Doc ume nt Format . Figure 7 -35 Export Output dialog box  The outline pane of the V iewer document is converted to bookmarks in the PDF fi le for easy navigation.  Page size, orientation, margins, content and disp[...]

11 0 Chapt er 7 be 1200 DPI. Not e : High -resolution documents may yield poor results when printed on lower-resolution printe rs. Figure 7 -36 PDF file with bookmarks[...]

111 W orking with Output Exporting Results to HTML Y ou can also export results to HTML (hypertex t markup language). Whe n saving as HTML, all non-graphic o utput is export ed into a single HTM L fi le. Figure 7 -37 Output.htm in Web browser When you e xport to HTML, charts can be expo rted as well, but not to a single fi le. Each chart will be [...]

Chapter 8 W orking with Syntax Y ou can save and automate many common tasks by using the powerful command language. It also prov ides some functionality not found i n the menus an d dialog boxes. Most commands are accessible from t he menus a nd dialog boxes. Howe ver , some commands and options are available only by using the command language. The[...]

11 3 Wo r k i n g w i t h S y n t ax E In the Charts dialog box, select Bar char ts . E In the Chart V alues group, select P ercentage s . E Click Continue . E Click Pa st e to copy the syntax created as a result of t he dialog box selections to the Syntax Editor . Figure 8-2 F requen cies synt a x E T o run the syntax currently displayed, from the[...]

11 4 Chapt er 8 E Click on the item labelled FREQ for frequencies. Clicking on a n item in the auto-completion c o n t r o lw i l li n s e r ti ta tt h ec u r r e n tc u r s o rp o s i t i o n( t h eo r i g i n a l PERCENT keyword w as manual ly deleted). By default, the auto-completion control will prompt you with a list of available terms as you [...]

11 5 Wo r k i n g w i t h S y n t ax Figure 8-5 FREQUENCI ES syntax help Y ou may have noticed that text disp layed in the syntax window i s color ed. Color coding al lows you to quickly identify unrecognized terms, since only recognized terms are colored. For example, you misspell the FORMAT subcommand as FRMAT . Subcommands are colored green by d[...]

11 6 Chapt er 8 Understanding the Error Pane The error pane displays runtime errors from the m ost current run. It contains the details of each error as well as the line number of the command on which the error occurred. Figure 8-6 Error p ane di splayed in the Syntax E ditor Clicking on the entry for an error positions the cursor on the fi rst li[...]

11 7 Wo r k i n g w i t h S y n t ax Figure 8-7 Breakpoint displayed in the Syntax Editor wi ndow When you run command syntax containing breakpo ints, execution stops prior to each command containing a breakpoint. Figure 8-8 Execu tion stopped at a breakpoint The do wnward poi nting a rrow to the left of the command text shows the progr ess of the [...]

11 8 Chapt er 8 T o resume execution fo llowing a breakpo int: E From the menus in the Synta x Edito r choose: Run > Conti nue[...]

Chapter 9 Modifying Data V alues The data you start with may not always be organiz ed in the most useful manner for your analysis or reporting needs. For example, you may want to:  Create a categorical variable from a scale variable.  Combine several response categories into a single category .  Create a new variable t hat is t he computed[...]

120 Chapt er 9 Figure 9-1 Initial Visual Binning dialog box In the initial V isual Binning dialog box, you select the scale and/or ordinal variables for which you want to create new , binned variables. Binning means taking two or more contiguous values and grouping them into the same category . Since V isual Binning relies on actual values in the d[...]

121 Modifying Dat a V alues Figure 9-2 Main Visual Binning dialog b ox E In the main V isual Binning dialog box, select Household income in thousand s [incom e] in the Scanned V aria ble List. Ah i s t o g r am displays the distributi on of the selected variable (which in this case is highly skewed). E Enter in ccat2 for the new binned variable nam[...]

122 Chapt er 9 Figure 9-3 Visual Binning: Make Cutpoints dialog box E Select Equal Width Inter vals . E Enter 25 for the fi r s tc u t p o i n tl o c a t i o n , 3 for the number of cutpoints, and 25 for the width. The number of binned categories is one greater than the number of cutpoints. So in this example, the new binned variable will have fou[...]

123 Modifying Dat a V alues Figure 9-4 Main Visual Binning dialog box with define d cut po ints The values now displayed in the grid represent the de fi ned cutpoints, w hich are the upper endpo ints of each category . V ertical lines in the histogram also indicate the locations of the cutpoints. By defaul t, these cutpoint values are included in [...]

124 Chapt er 9 Figure 9-5 Automatically gen erated value labels This automatically generates descriptive value labels for each category . Since the actual values assigned to the new binned variable are simply s equential integers starting with 1, the value labels can be very useful. Y ou can also manuall y enter or change cutpoints an d labels in t[...]

125 Modifying Dat a V alues The new variable is displayed in the Data Editor . Since the variable is added to the end of the fi le, it is displayed in the far right column in Data V iew and in the last row in V ariable V iew . Figure 9-6 New variable displayed in Dat a Editor Computing New V ariables Using a wide variety of mathematic al functions[...]

126 Chapt er 9 E Select Y ears with curr ent employer [employ] and click the arrow b utton to copy it to the expression. Figure 9-7 Compute V ariable dialog box Note : Be careful to select the correct employment va riable. There is also a recoded categorical version of the variable, which is not what you want. The numeric expression should be age?[...]

127 Modifying Dat a V alues The new variable is displayed in the Data Editor . Since the variable is added to the end of the fi le, it is displayed in the far right column in Data V iew and in the last row in V ariable V iew . Figure 9-8 New variable displayed in Dat a Editor Using Functions in Expressions Yo u c a n a l s o u s e p r e d e fi ne[...]

128 Chapt er 9  Cross-case functions  String functions Figure 9-9 Compute V ariable dialog box displaying function grouping Functions are organiz ed into logically distinct groups, such as a group for arithmetic operations and another for compu ting statistical met rics. For convenience, a nu mber of commonly used system va riables, such as $[...]

129 Modifying Dat a V alues The function is inserted into the expression. If you highlight part of the expression and then insert the function, the highlighted portion of the expression is used as the fi rst argument i n the functio n. Editing a Function in an Expression The functi on is no t complete until yo u enter t he arguments, represen ted [...]

130 Chapt er 9 Figure 9-1 0 If Cases dialog box E Select Include if case satisfies condition . E Enter the conditional expression. Most conditional expressions contain at least one relational operator , as in: age>=21 or income*3<1 00 In the fi rst example, only cases with a value of 21 or greater for Age [age] are selected. In the second ex[...]

131 Modifying Dat a V alues W orking with Dates and T imes A num ber of tasks commonly perform ed with dates and times c an be easily accomp lished u sing the Date and T ime Wiz ard. Using this wizard, you can:  Create a date/time variable from a string variable containing a date or time.  Construct a date/time variable by mergin g variables [...]

132 Chapt er 9 If you’re new to dates and times in IBM® SPSS® Statistics, you can select Learn how dates and times are represented and click Next . This leads to a screen that provides a brief overview of date/time variables and a link, through the Help button, to more detailed information. Calculating the Length of T ime between T wo Dates One[...]

133 Modifying Dat a V alues Figure 9-13 Calculating the length of time bet ween t wo dates: Step 2 E Select D ate of next r elease for Date1. E Select D ate of last upgrade for Date2. E Select Y ears for the Unit and T r uncate to Integer for the Result Trea tment. (These are the default selections. ) E Click Nex t .[...]

134 Chapt er 9 Figure 9-1 4 Calculating the length of time bet ween t wo dates: Step 3 E Enter Y earsLastUp for the name of the resul t vari able. Result variab les cannot have the same name as an existing variable. E Enter Y ears since last upgrade as the label for the result variable. V ariable labels for result variables are optional. E Leave th[...]

135 Modifying Dat a V alues Ad d i n gaD u r a t i o nt oaD a t e Y ou can add or subtract durations, such as 10 days or 12 months, to a date. Con tinuing with the example of the software company from the previous section, consider determining the date on which each customer’ s initial t ech support contract ends. The data fi le upgrade.sav cont[...]

136 Chapt er 9 Figure 9-17 Adding a d uratio n to a date: Step 3 E Enter SupEndDate fo r the name of the result v ariable. Result variables cannot have the same name as an existing variable. E Enter End date for support as the label for the result variable. V ariable labels for result variables are optional. E Click Finish to cr eate the new variab[...]

[...]

Chapter 10 Sorting and Selecting Data Data fi les are not always organized in the ideal form for your speci fi cn e e d s . T op r e p a r ed a t af o r analysis, you can select from a wide range of fi le transformations, including the ability to:  Sort data. Y ou can sort cases based on the value of one or more variables.  Select subsets [...]

139 Sorting and Selecting Dat a in years [age] . For string variables, uppercase letters pr ecede their lowercase counterparts in sort order (for examp le, the string value Ye s comes before yes in the sort o rder). Split-File Processing T o split your data fi le into separate groups for analysis: E From the menus choose: Data > Split File ... [...]

140 Chapt er 1 0 If you select Compare groups , res ults from all split- fi le gr oups will be included in the same table(s), as shown in the following table of summary statistics that is generated by the Frequencies procedure. Figure 1 0-3 Split-file output with single pivot table If you select Organiz e output b y groups and run the Frequencies [...]

14 1 Sorting and Selecting Dat a By default, Split File automatically sorts the data fi le based on t he values o f the groupi ng variables. If the fi le is already sort ed in the proper order , you can save pr ocessing time if you select File is alread y sor ted . T urning Split-File Processing On and Off After you invoke spl it- fi le processi[...]

142 Chapt er 1 0 Figure 1 0-6 Select Cases d ialo g box Selecting Cases Based on Conditional Expressions T o select cases based on a conditional expression: E Select If condition is satisfied and click If in the Select Cases dialog box. This ope ns the Select Cases If dialog box.[...]

14 3 Sorting and Selecting Dat a Figure 1 0-7 Select Cases If d ialo g box The conditional expression can use existing variable names, constants, arithmetic operators, logical operators, relational operators, and functions. Y ou can type and edit the expression in the text bo x just l ike text in an ou tput windo w . Y ou c an also use the cal cula[...]

144 Chapt er 1 0 Figure 1 0-8 Select C ases Random Sample dialog box Y ou can select one of the fo llowing altern atives for sample size:  Approximately . A user-speci fi ed percentage. This option g enerates a random sample of approximately the speci fi ed percentage of cases.  Exactly . A user-speci fi ed number of cases. Y ou must also [...]

14 5 Sorting and Selecting Dat a Figure 10 -1 0 Select Cases Range dialog box (time series) T o generate date variables for time series data: E From the menus choose: Data > Define Dates ... T reatment of Unselected Cases Y ou can choose one of the following alternati ves fo r the treatment of unselected cases:  Filter out unselected cases. U[...]

146 Chapt er 1 0 Figure 10 -1 1 Case selection st atus[...]

Appendix A Sample Files The sample fi les i nstalled with the produ ct can be found in t he Sample s subdir ectory o f the installation directory . There is a separate fo lder within the Samples subdirectory for each of the following languages: En glish, French, Germa n, I talian, Japanese, Kor ean, Polish, Russian, Simpli fi ed Chinese, Spanish,[...]

148 Appendix A  behavior .sav . In a classic example , 52 students were asked to rate the combinations of 15 situations and 15 behaviors on a 10-point sca le ranging from 0=“extremely appropriate” to 9=“extremely inappropriate.” A veraged over individuals, the val ues are taken as dissimilarities.  behavior_ini.sav. This data fi le c[...]

14 9 Sample F iles  catalog.sav . This data fi le contains hypotheti cal monthly sales fi gures for three products sold by a catalog company . Data for fi ve possible predictor variables are also included.  catalog_seasfac.sav . This data fi le is the same as catalog.sav except for the addition of a set of seasonal factors calculated from[...]

150 Appendix A  demo.sav . This is a hypothetical data fi le that concerns a purchas ed customer database, for the purpose of mailing monthly offers. Whether or not the customer responded to the offer is recorded, along w ith variou s demographic information.  demo_cs_1.sav . This is a h ypothetical data fi le that concern s the fi rst ste[...]

151 Sample F iles  health_funding.sav . This is a hyp othetical data fi le that contains data on health care funding (amount per 100 populatio n), disease rates (rate per 10,000 populati on), an d visits to health care providers (rate per 10,000 population ). Each case represents a different city .  hivassay .sav. Thi s i s a hypothetical da[...]

152 Appendix A  ozone.sav . The data include 330 observations on six meteorological variables for predicting ozone concentration from the remaining varia bles. Previous researchers , , among others found nonlinearities among these variables, which hinder standard regression approaches.  pain_medication.sav . This hypoth etical data fi le con[...]

153 Sample F iles  recidivism_cs_sample.sav . This is a hypothetical data fi le that concerns a government law enforcement agency’ s efforts to understand recidivism rates in their area of jurisdiction. Each case corr esponds to a previous offender , released fro m their fi rst arrest durin g the month of June, 2 003, an d records their d em[...]

154 Appendix A  stroke_survival. Thi s hypotheti cal data fi le concerns survival times for patients exi ting a rehabilitati on prog ram post-ischemic stroke face a number of challenges. Post-stroke, the occurrence of myocardial infarction, ischemic stroke, or hemorrhagic stroke is noted and the time of the event recorded. The sample is left-tr[...]

155 Sample F iles  ulcer_recurrence.sav. This fi le contains partial inf ormation f rom a study designed t o compare the ef fi cacy of tw o therapies for preventing the recur rence of ulcer s. It prov ides a good exampl e of in terval-ce nsored data a nd has been pr esented and ana lyzed elsewhere .  ulcer_recurrence_recoded.sav. Thi s fi [...]

Appendix B Notices This in formation was developed for products and services offered worldwi de. IBM may not offer the products, se rvices, or features discussed in this document in other countries. Consult your local IBM representative for information o n the products an d service s currently available in your area. Any reference to an IBM p roduc[...]

157 Notic es Such information may be available, subject to a ppropriate terms and conditions, including in some cases, payment of a fee. The licensed program described in this documen t and all licensed material available for it are provided by IBM u nder terms of t he IBM Customer Agr eement, IBM Int ernational Program License Agreement or any equ[...]

158 Appendix B Microsoft p roduct screenshot(s) reprinted with permission from Microsoft Corporation.[...]

Index Access (Micro soft), 12 bar charts, 55 cases selecting, 141 sorting, 138, 141 categor ical data, 53 summary measures, 53 charts bar, 55, 60 chart options, 79 creating charts, 60 editing charts, 66 histograms, 58 templates, 7 4 computing new variables, 125 conditional expressions, 129 continuous data , 53 copying variable attributes, 3 8 count[...]

160 Index PowerPoint (Microsoft) exporting resu lts to, 10 1 qualitative data, 53 quantitative data, 53 ratio data, 53 recoding valu es, 119 renaming datasets, 51 sample fi les location, 147 scale d ata, 53 scale variables summary measures, 56 selecting cases, 141 sorting cases, 138 split- fi le processing, 139 spreadsheet fi les reading, 1 1 re[...]