conpanda.jpg (8732 bytes)                                   anspss.jpg (10244 bytes)      

                                                                                     nonav.jpg (5444 bytes)


What do you want to do?

SPSS Computing Steps

Data Manipulation

Sorting Cases

Data-Sort Cases-
select one or more sorting variables

Selecting Cases

Data-Select Cases-Select Cases:If
specify the criteria for selecting cases
do not indlude an equal sign

Splitting File

Data-Split File
  • select COMPARE GROUPS if the split-file groups are presented together for comparison purposes or ORGANIZE OUTPUT BY GROUPS if all results are displayed separetely for each split -file group

Merging Files - Adding Cases

Data-Merge Files-Add cases
  • select the data file taht you want to merge with teh currently opened file
  • remove any variables you don't want and add any variables from the unpaired list taht you do want

Merging Files-Adding Variables

Data-Merge File-Add Variables
  • select the data file that you want to merge with the currently opened file
  • remove any variables you don't want and add any variables from the excluded list that you do want
  • indicate the key variable that matches the two files (change to have the same name if necessary)

Data Transformations

Compute Variables

  • type the name of a single target variable
  • build an expression by highlighting variables and using the arrow dey or type directly into teh Expression field

Recoding Variables

Transform-Recode-Into Different Variable
  • select variable you want to recode
  • enter a new output name for the new variable and click CHANGE
  • enter old value or old value range
  • enter new value and click ADD

Missing Values

Double click variable name
  • Define variable
  • Click missing values
  • Enter discrete values or range of values

Bivariate Procedures

Independent Samples T-Test use this method to compare means of one continuous variable across independent groups, the samples are independent.  Example:   males and females in the employment file can be divided into independent groups by teh variable gender.  The t-test can determine if the mean current salaries of the males and females are the same. Analyze-Compare Means-Independent T-Test
Test variable = the variable whose means you are comparing (ex. salary)
Grouping variable = the groups you are comparing (ex.gender)
Define groups = indicate the values you are comparing for the grouping variables (1 = male 2 = female)
One Sample T-Test to test whether a mean equals a predetermined value.  Example:  you can test whether the average salary in this company differs from a national average. Analyze-Compare Means-One-Sample T-Test
Test variable = the variable whose means you are testing (ex. salary)
Test value = the value you are comparing to (ex. national average)
Chi-Square determines a relationship between two categorical variables (example to test whether gender and job category are related) Analyze-Descriptive Statistics-Crosstabs
row (ex. gender)
column (ex. job category)
Statistics - chi-square box needs to be checked
ANOVA used to test for the differences in means among two or more groups simultaneously, if p<0.05 conclude that at least one population mean differs from the others.  We cannot tell at this point which mean is different. Analyze-Compare Means-One-way ANOVA
  • Dependent List: one or more dependent variables that must be coninuous (ex. salary)
  • Factor: a single independent factor variable taht must be categorical (ex. job category)

Describing a Population


Analyze-Descriptive Statistics-Frequencies
Descriptives (Mean, Variance, etc) only for continuous variables Analyze-Descriptive Statistics-Descriptives
Bar Chart (Usually for categorical data)
Simple Line
Histogram (shows distribution of values by dividing the range into equally-spaced intervals and plotting the count of cases in each interval as a bar)
Scatterplot (highlights the relationship between two                         variables)
Pie Chart (displays info. in a circular format)

Graphs-choose type you want (ex. simple)
  • Indicate type of graph and use options to improve quality.  For example, for a pie chart select interactive, pie, simple, and then slice by...slice summary is a method to disply information and pie tab allows you to label the pieces

Regression Analysis

Linear Regression ascertain the probable relationship between two continuous variables, predict or estimate the value of one variable corresponding to a given value of another variable

Dependent (ex. current salary)
Independent (ex. education level)
Method:  indicate teh method you want the variable to enter/remove