How to use SPSS ? Aug, 17, 2011 Hirohide Yokokawa, M.D., Ph.D. Department of General Medicine, Juntendo University School of Medicine.

How to use SPSS ?

Aug, 17, 2011

Hirohide Yokokawa, M.D., Ph.D.Department of General Medicine,

Juntendo University School of Medicine

The United States, Chicago, IL

Since 1968

Merge with IBM at 2008, and generated to IBM SPSS

Statistical Package for Social Science

http://www-01.ibm.com/software/analytics/spss/

1. SPSS has an excellent potential to handle large scale data.

2. It is easy to conduct high technique analysis.

3. No need to enter long codes. (Clicking only)

Strong Points of SPSS

1. Set Excel Data upOne Column is One variables

One Row is One Case

Caution !Limited to Arabic

numerals!Don’t enter any letters!

2. Install Excel Data

F O Data

1

2

1.Select “ Excel”.2.Chose your target file.3.Open the file.

3

Install success !

3. Check Errors and Distribution (Continuous Variables)

1

2 3

12

3

1.Select variables2.Click “Statistics”.3.Select needed measurements.

Total number

Check missing data

Representative value

4. Try to analyze (Basic analysis) 1. Categorical Data

1. Chi square Analysis2. McNemar Test

2. Continuous Data1. Two group comparison

(Parametric/Nonparametric)2. More than three group

comparison3. (Parametric/Nonparametric)

Categorical Variables

Select Crosstabs

1

2

3

1

3

1.Select variables2.Click “Statistics”.3.Click “Chi-square”.

2

2 x 2 Table

Total number

Significance

4-1-2. MacNemar Test

1

4-2-1. Student t Test1

2

3

4

5

Test for Equality of Variance

If p value ≥0.05, accept p value of “Equal Variances”

If p value <0.05, accept p value of “NOT Equal Variances”

4-2-1. Paired t Test

1

2

3

4-2-2. Mann Whitney Test1

2

3

4

4-2-2. ANOVA1

2

3

4-2-2. Kruskal Wallis Test1

2

3

4

\

Click

5. Try to analyze (Advanced analysis)

1. Correlation Analysis

2. Regression Analysis

1. Logistic regression analysis

2. Multiple regression analysis

5-1. Correlation Analysis 1

2

\Select

appropriate Coefficient

\

Enter two variables

\

5-2. Multiple Regression Analysis

1

2

3

1

23

Β value

P-value

5-2. Logistic Regression Analysis 1

2

3

P-value

Odds ratio