ORIENTATION TO THE COMPUTER LAB

ORIENTATION TO THE COMPUTER LAB(S)

PHS 102 has PCs and laser printers (need to pay for printing)

PHS 115 has 3 PCs, a printer, and some network capabilities

Other available computer labs on campus (check for hours, updates, other labs); note that SPSS only runs on PCs and Dual platform Macs and the program might not be installed on all machines in other campus labs.

ED174 has 24 Macs

Music 234A has 22 PCs; Music 232 has 30 Macs

PB133 has 40 PCs and 5 Power Macs

SS202 has PCs

***RULES: NO FOOD OR DRINK IN THE LABS UNDER ANY CIRCUMSTANCES***

Discard unwanted printouts and papers in recycling bins.

PHS 102: You may use the lab when I reserve it for class or during the open lab hours. The open lab times are posted outside the door. You must stop your print job and vacate the lab when asked to do so by instructors for other scheduled classes.

There should be a student assistant in the lab during open hours. If no assistant is present, call the HELPLINE 8-5000 if you need assistance or if there is a problem with the equipment. (The techs will not always be familiar with programs like SPSS.) Note that PHS 115 does not have a lab assistant. The DSWE office staff CAN NOT support either lab, so CALL THE HELPLINE. RE: PHS 115 - Notify me if you encounter problems with hardware like viruses and I will notify the tech folks.

ALWAYS SAVE YOUR WORK TO A FLOPPY DISK (AA:\@ drive or to a zip disk) AND TAKE IT WITH YOU. If you save work to the hard drive, it will be erased during regularly scheduled disk cleaning.

VIRUS WARNING: BE CAREFUL. MANY PEOPLE USE THE LAB MACHINES. ALWAYS CHECK FOR VIRUSES BEFORE USING IT IN YOUR HOME MACHINE. JUST BECAUSE THERE IS A VIRUS PROTECTION PROGRAM ON THE DSWE COMPUTER DOES NOT MEAN IT IS SAFE.

ORIENTATION TO SPSS 10.0 for Windows

Revised for SPSS for Windows 10.0; originally developed by Dr. Donna Hardina. In SPSS 10, there are 2 view sheets. One is a Data View sheet in which you enter your data; the other is a variable definition window (Variable View@) sheet. Note that the program opens to the data view@ sheet and you can switch between the sheets by clicking on the appropriate tab at the bottom of the screen. Other changes from prior versions are small, e.g., the statistics menu is now labeled Analyze.

SPSS is a common statistical package for the computer that is used for data analyses in the social sciences. It has a data input window that looks like a spreadsheet; the tab at the bottom says data. SPSS 10 has a second Asheet@ (see tabs at bottom of screen). This give you a more direct interface for defining your variables. The program handles many common statistical and graphic procedures and produces publication-ready output. In order to effectively use the program, a knowledge of basic statistics is necessary so you will know what analyses you want to run (and where to find them in the program) and be able to interpret the output. The windows Aclick and run@ format makes it tempting to run any and all analyses without really thinking through the research/analysis question and considering the nature of the variables (i.e., level of measurement) in relation to the assumptions of the statistics you are using. The computer will try to run anything you ask for whether it is reasonable to do so or not depending on how you define your variables.

Note that as you work, various windows are activated and you can move between windows (Data Editor, Output, or Syntax/Log) by clicking on the related bar at the bottom of the screen. You can also use the WINDOW menu and select the window you want to view or click on the relevant icon in the icon bar.

ABOUT THE LAB MACHINES: Do not save your work to the hard drives of the Lab Machines because these files get purged on a regular basis; the file won=t be there when you come back another day. Use a Floppy or Zip disk for all of your work.

Get familiar with the menu bars at the top if the data spreadsheet/view spreadsheet. Common tasks and analyses are located in logical places.

ENTERING DATA

UNLESS NOTED OTHERWISE, SELECT MEANS A SINGLE CLICK OF THE MOUSE.

To start the program, click TWICE on the SPSS icon on the screen. If there is no icon, click on START, click on PROGRAMS, click on SPSS, click on SPSS PROGRAM. To create and enter a data file, do the following:

STEP #1 - ASSIGN/DEFINE VARIABLE NAMES, VARIABLE LABELS, AND VALUE LABELS

Click on the tab at the bottom of the screen that says AVARIABLE VIEW@. This opens the second screen that looks like a spreadsheet and has Aname,@ Atype,@ Awidth@ etc across the top row.

Assign a variable name to each variable in the data that you have. Variable names should be no more than 8 letters in length (no spaces and begin with a letter). For example, GENDER, income, AGE3, and INC98b are acceptable variable names. You should assign a name that gives you a good idea about the concept that is measured by the variable/which variable it is. I do not advise using "var1, var2, var3, etc." because you can=t remember which is which. Make it a habit to have your first variable be an identifier of some sort, e.g., CODE, ID, OBS or the like, to identify each "unit" (e.g., each subject) that is represented in the data.

Click once on the cell in the first column/first row in the view spreadsheet under "name"; type the name of the variable, e.g, id, gender, age. Note that defaults are automatically entered for all of the other variable characteristics.

Going across the first row, define the TYPE of variable. Default is numeric (i.e., a number). Left-click on the gray box and a screen comes up that allows you to define the variable as dates, string (characters like "A" or "Fresno"), etc. You will usually leave it as "numeric".

Go to the next column, define a WIDTH for the column if other than 8 spaces. It's OK if your variables will be less than 8 spaces, you don't have to change it. Again, click on the gray box to get the window in which you can redefine the width.

Go to the next column, DECIMALS. Default is 2 places (i.e., .44). To change, activate window by clicking on gray box.

Go to next column, LABEL. Here type in the descriptive label for the variable, e.g., AGE "age in years" (I don't know how long the label can be)

If the variable is nominal or ordinal level of measurement, ASSIGN VALUE LABELS. Under VALUE, click on gray box to activate window to DEFINE LABELS. Enter a value that you use with the variable. In the example of gender, you may use 1 to code for male and 2 to code for female. Enter 1, then move to the VALUE LABEL box and enter "male", click on add and you should get 1 = "male" in the bottom box. Enter the value and value label for the remaining categories. ALWAYS click on ADD or the code will not be entered. When you finish, click OK.

The other two variable view columns that you might be likely to use are MISSING and MEASURE. Missing allows you to define a number or range of numbers to represent missing data (you can also leave the data field blank and the program will realize the datum is missing). Measure lets you specify if the level of measurement is scale (i.e., interval, ordinal, or nominal). Use to gray box to activate the screens for this.

STEP #2 ENTERING DATA

On the Data sheet, the variables you just defined should now be listed across the top of the sheet. Now each observation/respondent has data which you enter by rows (each line across the sheet is one subject). Each column is a variable. (You will notice that, on the "variable view" spreadsheet, you entered the details about each variable going across the sheet; each row is a new variable.) Enter the data for each subject under the appropriate variable.

Select the cell where you wish to enter data with your mouse (or cursor over to it). Enter the value. You can hit the enter key and the next cell in the column is highlighted or cursor to the next cell you want to use. You can also select "view" and "value labels" from the icon bar menu and you will be able to see what your codes mean (e.g., enter a 1 which is defined as "male" and "male" appears in the cell.) This feature toggles on and off.

STEP #3 SAVING YOUR DATA:

When you have finished entering your data (or other work) MAKE SURE YOU ARE IN THE DATA WINDOW. Pull down the FILE MENU from the top left-hand side of the screen. Select SAVE AS. Select the SAVE IN field at the top of the window, click on the downward arrow at the right of the field and select the "A" drive in order to save to your 3 1/2 inch floppy disk; the default is to save to the hard drive on the computer -- do not do this because your file will not remain on the hard drive.

Then click once on the field FILE NAME. Give your file a name using a maximum of 8 characters. Click on SAVE.

The file should save to the A:\ drive as "filename.sav". The .SAV is an SPSS data file default extension; this is a special data file format that only SPSS reads. If you click once on the arrow to the right of FILE NAME (bar in the lower left of the window) you will note that you can save the file in other formats, such as Excel and Lotus. This is useful in making a file transportable. However, the definitions of the variables may not go over in the translation.

When saving, be sure you have the data editor window open (on the screen), what you see on the screen is what the program is trying to save for you. There are 3 different windows of information that SPSS can save for you. And each window is saved in a different format; THIS IS IMPORTANT or you can get very frustrated and lose your work.

Note that when you are working from an existing file from your "A" drive, you can simply use the SAVE command. It will replace your existing file with the updated one. If you don't want to overwrite the file, use SAVE AS and give the revised file a new, unique name.

LOADING AN EXISTING DATA FILE:

Select the FILE menu

Select OPEN FILE/DATA (you can also ask to open OUTPUT, SYNTAX files)

The default FILE TYPE (field in lower part of window) is .SAV (an SPSS data file). (Note that you may also pull up .SPO (SPSS output files) and various spreadsheet formats such as .XLS (Excel), .WK# (Lotus files), .DBF (Dbase files), etc., by highlighting and selecting the appropriate file type. BUT YOU HAVE TO ASK FOR THESE)

The default file location (the field at the top of the window) is wherever the SPSS program is. Click on the arrow to the right of the field and select the "A" drive, or the drive where your file is located.

Once in the correct file location/folder, you should get a list of the files of the desired file type. Select the file you want to open.

Click on OK. If the file fails to open (or looks weird), go to the output window to review any comments about the failure. A typical mistake is that the data file didn't get saved.

RECODING DATA

Say you have a True/False questionnaire where some of the statements were worded in reversed fashion. You want an answer of TRUE to have the same meaning for all of the statements with respect to the concept you are trying the measure. Thus, the answers on some of the statements have to be coded in reverse. (You are using 1=True and 0=False).

1. Select TRANSFORM menu

2. Select RECODE

3. Select SAME VARIABLE. For BEGINNERS, SEE NOTE THAT FOLLOWS.

Select the variable that you want to recode. Note that you can select multiple variables that will be recoded in the same fashion.

Select old and new value.

Put the old value in the OLD VALUE cell and the new value in the NEW VALUE cell to the right. (E.g. if False = 0 and True =1 and you want to reverse the coding, enter 0 as the old value and 1 as the new value; then 1 as the old value and 0 as the new value.

Click continue, then OK.

NOTE FOR BEGINNERS: I recommend making a new variable because that way, you don't mess up your original data through some error. If you make a new variable, select NEW VARIABLE instead of SAME VARIABLE.

Select the variable(s) you want to recode by highlighting them and clicking on the arrow to the right of the variable list. Each old variable should be followed by ->?

Highlight each variable listed under Numeric Variable -> Output Variable

Click on Output Variable field and enter a name for the new variable.

Under OUTPUT VARIABLE, click on NAME, enter the new variable name.

Click on LABEL and enter the new variable label. You will have to define the value labels using the DEFINE VARIABLE routine discussed above.

Select old and new value

8. Click continue, then OK.

COMPUTING A NEW VARIABLE

Say you want to make a new interval variable of income in 1999 (INC99) that combines this year's income (INCOME) and adjusts it by a cost of living percentage (COLA).

elect TRANSFORM

Select COMPUTE

Enter INC99 in target variable box (top left); this names the new variable.

Click on box on right top. NUMERIC EXPRESSION and enter the equation that will give you what you want, e.g., INCOME+INCOME*COLA.

Select OK. You may be prompted for a yes to change the variable. Click Yes.

You can also use this routine for creating ordinal categories from interval data:

For example, you have ages (from 18 to 85) and want to create an ordinal variable (1=18-25; 2=26-40; 3=41-55; 4=56-64; and 5=65 and over.)

1. Under COMPUTE NEW VARIABLE, you define a target variable, CATAGE.

2. Then enter 1 under NUMERIC EXPRESSION

3. Click on IF.

4. Click on INCLUDE CASE IF CASE SATISFIES CONDITION

5. Click on the field to the right of arrow (you can type or use the mouse as you wish) but you want a statement such as: AGE le 25 (this says, AGE is less than or equal to 25)

6. Click on continue and OK.

7. Repeat this process for each numeric category you want to create with a new conditional statement such as those listed below. You have to think through carefully how to specify the ranges of the values of the original variable included in each new category.

catage=2 , if age gt 25 and age le 40

catage=3 , if age gt 40 and age le 55

catage=4 , if age gt 56 and age le 64

catage=5 , if age ge 65

COPYING VARIABLE SPECIFICATIONS IN SPSS

Say you have many of the items (questions) that use the same response format (e.g., 1=strongly agree, 2= agree, 3=neutral, 4=disagree, 5=strongly disagree). Instead of writing these values/value labels out for each variable (item), you can apply the same value labels to a number of variables at once.

1. First, set up your variables by giving them variable names and variable labels.

2. Then, define the values you use to code the responses for one of the variables. Click and copy this cell and paste it to ALL of the variables to which you want to apply the same value labels. This is just like copying a section of a spreadsheet.

You can see on the view spreadsheet whether the values have been applied.

INSTRUCTION FOR RUNNING DESCRIPTIVE STATISTICS

1. Pull down ANALYZE menu.

2. Click on "DESCRIPTIVE STATISTICS"

3. Click on "FREQUENCIES" - this can be used for variables of all levels of measurement for some useful information.

4. Highlight the variables in the left hand box that you wish to analyze. Click on the arrow to the right of that box and the variable names should appear in the right-hand box. (Or double click on a variable and it will be put in the analysis box.)

5. If you want additional statistics, click on statistics button at the bottom of the window. If not, go to line 8.

6. Select those statistics that you want. E.g., with interval data you can get a mean and standard deviation, among other stats.

7. Select continue; this takes you back to the frequencies window.

8. You may also want a GRAPH, so click on charts and select the type of chart you want (bar chart or histogram).

9. The FORMAT button allows you to control some the organization of the output, e.g., ascending vs. descending order of presentation etc. Click continue.

10. Once back at the FREQUENCIES window, click OK and you will get a frequency table and any associated statistics/charts that you requested.

DESCRIPTIVE STATISTICS FOR INTERVAL DATA: MEANS, VARIATIONS, ETC.

1. Pull down ANALYZE menu.

2. Click on DESCRIPTIVE STATISTICS

3. Click on DESCRIPTIVES

4. Select the variables you want analyzed; this is only for interval level dat5. Under the OPTIONS button you can select different statistics

6. Select OK and this will produce the table of results.

CROSSTABULATIONS/CHI-SQUARE

1. Select ANALYZE MENU

2. Select DESCRIPTIVE STATISTICS

3. Select CROSS-TABS

4. Select the variable for the row field (should be the dependent or criterion variable if you have one)

5. Click on arrow pointing at row

6. Select the variable for the column field (should be the independent or predictor variable if you have one)

7. Click on arrow pointing at column

8. Select statistics submenu on CROSS-TABS window

9. Select CHI-SQUARE (You will note that there are many more statistics that could be used in this situation; we will not be getting into those. Generally, they represent special situations with a cross-tab table.)

10. Select continue

11. Select cells submenu

12. Select percentages for row/column/total; also you can ask for observed and expected frequencies, etc. as desired. This will make more sense once you learn how chi-square is calculated.

13. Select continue.

14. Select OK. This will generate a cross-tab table and chi-square statistics.

GRAPHING

(You can also get bar charts and histograms, plus other types of graphics from the graph menu, the basic procedures are similar)

SCATTERPLOT:

1. Pull down the GRAPH menu.

2. Highlight "SCATTERPLOT"

3. Select SIMPLE

4. Click on "DEFINE"

5. Choose variables for the X (horizontal or abscissa) and Y (vertical or ordinate) axis by highlighting the variable and clicking the arrow key to the right of the variable list. Your predictor variable (independent variable) should be placed in the box for the X-axis and the criterion variable (dependent variable) should be placed in the box for the Y-axis.

6. You can also define titles and labels for the graph and use options to define which observations are used in the analysis.

7. Click continue and OK.

8. Editing the graphic: Double click on the table/graph and a smaller edit window pops up. The menu bar and icons are used to change colors, lines, enter titles, labels, legends etc. Using this just takes practice and trying things out to make the final graphic pretty.

INSTRUCTIONS FOR RUNNING INFERENTIAL STATISTICS

CORRELATIONS:

1. Pull down the ANALYZE menu

2. Select "CORRELATE" (move cursor to "CORRELATE" and single click on it; this should take you to the next screen)

3. Select "BIVARIATE" from the "CORRELATE" submenu

4. Select the variables you want to use from the variable list. (Remember that only continuous/interval level variables can be used). You need at least 2 variables.

5. Click the arrow key to the right of the variable list.

6. Click OK. This should give you a correlation "matrix"

(Note that there are options that you can select prior to running the correlations that give you other choices in your output/analysis. Feel free to try some of these out. Also you may want PARTIAL CORRELATIONS so know it is here)

T-TEST - a test of difference of mean between 2 groups

1. Pull down the ANALYZE menu

2. Select "COMPARE MEANS" by highlighting with your mouse and clicking once.

3. Select "INDEPENDENT SAMPLE T-TEST"

4. Highlight the variable representing the 2 groups that you wish to compare. This test can only be used to compare 2 groups. Click on the arrow next to "GROUPING VARIABLE".

5. Select "DEFINE GROUPS" by clicking on the box.

6. Insert the numerical values assigned to the two categories you want to compare, e.g., 1 for group #1 and 2 for group #2.

7. Select continue.

8. Select the dependent variable (a continuous/interval level variable) and click arrow at TEST VARIABLES field.

9. Select OK.

ONE-WAY ANOVA - a test of differences of means between 2 or more groups - an analysis of ONE factor

1. Pull down ANALYZE menu.

2. Select "COMPARE MEANS" by highlighting and clicking once with mouse

3. Select ONE-WAY ANOVA

4. Select a dependent variable (must be a continuous/interval level variable)

5. Select a nominal or ordinal independent variable for "FACTOR." This variable must have 2 or more categories. A 2-category variable will replicate your T-Test analysis.

6. Click continue and OK.

(Again there are many secondary analyses that are important to consider.)

REGRESSION

1. Pull down ANALYZE menu.

2. Select "REGRESSION" by moving to the right and clicking once with your mouse.

3. Select "LINEAR".

4. Highlight dependent variable (must be continuous/interval level). Click on the arrow next to the DEPENDENT field.

5. Highlight one or more independent variables (also continuous/interval level). Click on the arrow next to INDEPENDENT field.

6. Click OK.

(Obviously there are many more options and controls possible that I am not discussing here.)

PRINTING YOUR OUTPUT

Pull up the OUTPUT WINDOW/VIEWER WINDOW.

Click on the table or graph you want to print; multiple choices can be selected by depressing the CTRL key while clicking once on each table/graph you want to print. This may or may not work depending on how well the printer talks to computer.

Select the print icon or go through File, Print. Make sure you are doing this in the output window. This can be tricky and you may end up printing out the data sheet when you want the output of the results.

Click on print selection under PRINT RANGE, number of copies (1), and OK.

SINCE YOU ARE PAYING TO PRINT, YOU WANT TO BE SELECTIVE AND CAREFUL. IF YOU DON'T SPECIFICALLY SELECT WHAT YOU WANT PRINTED YOU MAY GET MANY UNWANTED PAGES. ONE STRATEGY IS TO DELETE/CLEAR ALL OF THE OUTPUT, RE-RUN JUST WHAT YOU WANT AND PRINT THAT. It takes very little time.

SAVING YOUR OUTPUT TO DISK

Again, in the output window, ask for "save as" and select drive A in SAVE IN field.

Give the file a name (maximum of 8 characters) in the FILE NAME field.

The program automatically gives an extension of .SPO to the file (noted in the FILE TYPE field). This can only be read in SPSS.

If you want a file that can be read in a word processor (but is not as pretty), go back to the FILE MENU

Select EXPORT AS