Intro to SPSS Using SPSS through WebFAS http://www.yorku.ca/computing/students/labs/webfas/ Try it early (make sure it works from your computer) If you need help contact UIT Client Services Voice: 416-736-5800 Email: helpdesk@yorku.ca 1
SPSS The basics Efficient and simple program for statistical analyses There are 4 main windows Starting Up Start All Programs IBM SPSS Data can be entered directly into SPSS or imported from other programs (e.g., Excel). 2
First window: Data View Like an excel Spreadsheet Enter and edit data here Saved files extension is.sav. Second window: Variable View Define your variables here 3
Third window: Output viewer Displays results of analyses and errors. Extension of the saved file will be spv. Four window: Syntax Editor Usually for more complex analyses. Extension of the saved file will be sps. 4
Entering Data Always start with an ID number. In Variable View, the Name is what will appear as headings for your columns. For the name The first character must be a letter Cannot use the same name twice. Spaces are NOT allowed. Variable View window Label: use it to specify details Type The two basic types of variables that you will use are numeric and string. 5
Variable View window: Type Values: for categorical variables tell you which numbers represent which category e.g. 1=male; 2=female Ready to enter some data 6
Data format Repeated Measures In SPSS each row is a single case or participant Independent Groups 7
Transforming data Calculate totals for a scale Is it a scale that is summed, averaged, or is there a unique formula to achieve the overall score? Unless you are doing an item analysis, all your calculations will be done using an overall score, rather than the individual items on the scale Transforming data Type your variable name into Target Variable Select SUM or AVERAGE from the menu (arrow up) Example: happi = MEAN (hap1, hap2, hap3) 8
Reverse-coding Are there items that need to be reverse-coded? Transform Recode -> Into Different Variables Move the variable you want to recode into the Numeric Variable box Name the output variable (e.g., the same variable name, but with r at the end) Click Change Old and New Values Example: Old Value = 1, New Value = 5, (Click Add ) Make sure to do this for all the values of your scale Continue & OK Reverse-coding 9
Basic Analyses Data Cleaning Before running analyses, you must check that your data has been entered correctly. Scan your data on screen for any errors (ex. a 33 is entered when the scale is only 1-5). To do a more thorough check, SPSS can check the range of all your variables Run Frequencies for all your variables, Check Minimum and Maximum values. 10
Frequencies (for categorical variables) Analyze Descriptive Statistics Frequencies Charts OK -Move variables into the box -If bar charts are required Descriptives Analyze Descriptive statistics Explore Select variables and move into Dependent list box using arrow If you want to split the stats by group (e.g., males and females), move the categorical variable into Factor List Statistics Select Descriptives & Continue Plots Select histogram and normality plots with tests & Continue OK 11
Selecting Cases If you want to run analyses or operations on only a part of your sample. (E.g., looking at males only, looking at a certain age range, etc.) Select cases (e.g., gender: male = 0, female = 1) Data Select Cases If condition is satisfied Example: only male participants Move the gender variable into the box Gender = 0 (depending on the categories you have, you can use the < or > signs too) Get data files at: http://psych4170d.blogspot.com/ 12
Correlation Some say, You can t buy happiness! A researcher wanted to test this theory. Since he knew he could not manipulate income nor subjective wellbeing, he knew he had a correlational design on his hands. Participants completed a demographic questionnaire (with a question about annual income). He also gave his participants a well validated measure of subjective well-being. To analyze his data he ran a correlation. -go to Analyze -select Correlate -select Bivariate -bring both measures into the white variables box -Pearson should be checked off -Click, OK -the correlation s sign will tell you the nature of the relationship. The numbers will tell you the size and the p-value will tell you if it is a significant relationship. 13
Regression Some research on attraction suggests that familiarity breeds liking while other theories maintain that opposites attract. On researcher wanted to test which it was, did similarity predict attraction or do opposites attract? Accordingly she gave her heterosexual female participants a fake profile of a fictitious man and asked how similar they were on a 10-point scale. She also asked how attracted they were to the person depicted in the profile. She then ran a regression to see if similarity predicted attraction. -go to analyze -go to regression -go to linear -in the dependent box put in your variable of attraction -in the independent box put in your measure of similarity. -click, OK -you don t want to look at the model statistics. You want to look at the coefficients box and across the line where it says Similarity. This will give you your t statistic and your p value of interest 14
T-test A researcher wants to prescribe a medicine which has known side effects for causing stress. She wants to know if she gives small doses of the drug that it will not cause stress (as compared to a high dose). Accordingly, she randomly assigns half her participants to a low dosage level and the other half to a high dosage level. -Go to analyze -Compare means -Choose independent samples t-test -Under test variables choose your dependent measure(s) -Under grouping variables enter your independent measure -values: use the one s you created--check variable information -click OK -ideally you will want the p-value to say it is less than the traditional.05 cut-off. 15
ANOVA Suppose that you are running the same study as in question 3 but now you have three groups: low dose, medium does and high dose. Now you need to see if there are differences among the three groups. -Go to Analyze -Go to Compare Means -Go to One-way ANOVA -Under Factor put click over your independent measure -Under dependent list put your dependent measure -Click on Options and choose descriptive statistics and means plot -The ANOVA table will give you your F statistic and p value and the plot will give you an idea of what the means for the different conditions look like NOTE: SPSS conducts a test for homogeneity of variance, using Levene s test. This test should not be significant (you do not want the 2 variances to be different), so you want the reported Sig. value to be greater than.05. Next, the results of the t-test are presented using 2 different assumptions. The top row shows the outcome assuming equal variances, using the pooled variance to compute t. The second row does not assume equal variances. Each row reports the calculated t-value and the degrees of freedom, and the level of significance (the p value). 16
ANOVA -to test which groups are different from one another, you must exclude one group at a time and compare the other two using a t-test -to exclude a condition, go to, Data and then Select cases and then If condition is satisfied and then bring the Dosage into the box and choose the cannot = sign and choose a dosage level and then click, OK -look at your dataset--there the condition you chose to exclude should be striked through in the margin. -now run a t-test as before. 17