Introduction to SPSS

[email protected]

2011.4.6
1
 Preface
Statistical Software
• A program, which takes numbers as
input and creates tables (and figures) ?
• A (collection of) program(s) for
exploration, inference and modeling ?
• A tool for administration, manipulation
and analysis of data ?

2011.4.6
2
• A medium of communication with
CPU (graphics card,printer, . . . ) of our
computer

A?Statistical software is a tool to tell

the computer how to analyze your
data.

3
Additional Definition

• Providing the computational tools

for statistics by using tools and
techniques from computer science.
• Doing statistics by using computer
based tools.

2011.4.6
4
SPSS

• SPSS (Statistical Package for the Social

Sciences) is a statistical analysis and data
management software package.
•SPSS can take data from almost any type of file
and use them to generate tabulated reports,
charts, and plots of distributions and trends,
descriptive statistics, and conduct complex
statistical analyses.

5
 SPSS
1. Introduction
SPSS, which once stood for "Statistical Package
for the Social Sciences",
Sciences was created by Norman H.
Nie, C. Hadlai (Tex) Hull and Dale H. Bent in 1968,
three young men developed a software system based
on the idea of using statistics to turn raw data into
information essential to decision-making. The initial
work on SPSS was done at Stanford University.

6
 SPSS Inc. is a leading worldwide provider of predictive
analytics software and solutions. Today SPSS has more than
250,000 customers worldwide, served by more than 1,200
employees in 60 countries.

SPSS has become a leader in predictive analytics

technologies through a combination of commitment to
innovation and dedication to customers. You will find SPSS
customers in virtually every industry, including
telecommunications, banking, finance, insurance, healthcare,
manufacturing, retail, consumer packaged goods, higher
education, government, and market research.

7
 PASW (formerly SPSS)
SPSS is a computer program used for
statistical analysis. Before 2009 it was called SPSS, but in 2009
it was re-branded as PASW (Predictive Analytics SoftWare). are
The company announced July 28, 2009 that it was being
acquired by IBM for US$1.2 billion.

8
 2. How to use it
SPSS performs any task mainly by a menu,
menu but
can also use program,
program too.

Advantage: You don’t need to remember

many
command names or program rules, you can only
click mouse to perform statistical analysis task.
It is very adapted for a beginner.
Disadvantage: inflexible,

9
 Before you analyze the data, you must set the
SPSS dataset in Data Editor. Like this:

10
 Once you have set the SPSS dataset, you may
use certain statistical method in the Analyze menu
to get the results.

11
For example
linear regression analysis using SPSS

The age (year) and systolic blood pressure (kPa)

were measured from 12 women in certain area.
The data listed in table 4-2.
table 4-2 The age (year) and systolic blood pressure (kPa) for the 12 women in certain area
No. 1 2 3 4 5 6 7 8 9 10 11 12
age 56 42 72 36 63 47 55 49 38 42 68 60
systolic
19.6 16.7 21.3 15.7 19.9 17.1 20.0 19.3 15.3 18.7 20.2 20.6
blood pressure

12
 First, you need input the raw data in Data
Editor, set a SPSS dataset(filename: exp13_1.sav).
it shows the Data View windows below.

13
 From the Variable View window, you can know
the information of all variables in SPSS dataset. it
shows two ( x – age, y – blood pressure ).

14
 Next, you need select proper command in
Graphs menu, draw a scatter plot. Like this:

15
It shows the Scatterplot dialog boxes below.

16
It shows the Scatterplot graph below.

17
 Next, you may select linear regression command
in Analyze menu, do a linear regression analysis.

18
It shows the Linear Regression dialog box below.

19
 It shows the first part of output for linear regression
analysis below.

20
 It shows the last part of output for linear regression
analysis below.

According to the results, you can get the regression

equation and the result of hypothesis test for β.

Yˆ 10.813  0.151X t = 6.320, P < 0.001.

21
• More knowledge, see reference books about SPSS.

• You can also find so many sources about

SPSS in internet by Yahoo or Google.

• You can go to the Website of SPSS directly.

www.spss.com

22
 Part I Set up data file

There are two steps to set up data file in SPSS.

The structure of general database likes this. 

1. Set up variables
In Variable View window, you set up every
variable’s properties in SPSS. Include Name, Type,
Width, Decimals, Label, Values, Missing, Columns,
Align, Measure.

23
 See the sample data file in Variable View . 

2. Input raw data

In Data View window, you need input raw data.

case by case in every row, variable by variable in
each column. If the variable has defined the code
number, just input digital number.

See the sample data file in Data View . 

24
 Part II Data analysis
I. Descriptive Statistics
1. Measurement Data
SPSS’s Commands and Options ：
Analyze→Descriptive Statistics→ Descriptives…
Options … :
Mean(Arithmetic Mean):
Mean
Std. Deviation(Standard Deviation):
Deviation
S.E. mean(Standard Error of mean):
mean
Range:

25
 2. Measurement & Enumeration Data
SPSS’s Commands and Options ：
Analyze→Descriptive Statistics→ Frequencies…
 Display frequency tables:
Statistics … :
Median:
Percentile:
Quartiles:
 You can use “Quartiles” to get
InterQuartile
Range .
26
Charts… :
Histograms:

3. Enumeration Data (Contingency table)

Column

Row

SPSS’s Commands and Options ：

Analyze→Descriptive Statistics→ Crosstabs…
Row:
Column:
27
Statistics … :
Chi-square(Chi-square Test):
Test
Cells … :
Observed (Actual Frequency)
Frequency :
Expected (Theoretical Frequency)
Frequency :

28
 II. Inferential Statistics
1. Normality Test
Analyze→ Nonparametric tests → 1-Sample K-S…
Test Distribution:  Normal:
For Output:
Output
Asymp. Sig. (2-tailed)=P-Value,
Value P>0.05 means
data nearly follows normal distribution.
distribution

29
2. One-Sample T Test ：
Analyze→Compare Means→One-Sample T Test…
Test Variable: for one sample measurement data
Test Value: to input the known parameter (μ0 )
For Output:
Output
t = t-Value,
Value Sig. (2-tailed)=P-Value,
Value

P ＜ 0.05 means refuse H0 : μ ＝ μ0 .

30
3. Paired-Samples T Test ：
Analyze→Compare Means→Paired-Samples T Test…
Paired Variables: for paired data
Variable1: before treatment data
Variable2: after treatment data
For Output:
Output
t = t-Value,
Value Sig. (2-tailed)=P-Value,
Value

P ＜ 0.05 means refuse H0 : μd ＝ 0 .

31
 4. Two-Samples T Test ：
Analyze→Compare Means→Independent-Samples T
Test…
Test Variable: for measurement data
Grouping Variable: to mark the data into two
groups
For Output:
Output
t = t-Value,
Value Sig. (2-tailed)=P-Value,
Value

P ＜ 0.05 means refuse H0 : μ1 ＝ μ2.

32
5. Chi-square Test ：
Analyze→Descriptive Statistics→ Crosstabs…
Row(s): for grouping data
Column(s): for categorical testing data
Statistics … :
 Chi-square(Chi-square Test):
Test χ2
For Output:
Output
Value =χ2 -Value,
Value Asymp. Sig. (2-tailed)=P-Value,
Value
P ＜ 0.05 means refuse H0 : π1 ＝ π 2. (seefig2)

33
6. Linear Regression ：
Analyze→Regression→ Linear…
Dependent: （ y ）
Independent(s): （ x ）

7. Linear Correlation ：
Analyze→ Correlate→ Bivariate…
Variables: paired variables
coefficients  Pearson
Correlation coefficients:

34
 •The structure of database

variable
no name sex age Height (cm) Weight (kg) Blood Type
1 Tom male 23 175 71.5 A
2 Jerry female 18 168 58.3 AB
3 Jean female 25 162 61.4 B
4 John male 31 182 74.5 B
case 5 Wendy female 15 173 56.7 O
6 Mark male 28 176 66.9 A
7 Bob male 67 165 75.6 B
8 Bush male 21 191 68.2 AB
9 Ben male 12 174 73.9 O
10 Lily female 75 167 56.1 B

• One row, one case

• One column, one variable
(index)

35
36
37
 1-male
sex Input code number
2-female

Input the meanings that code number indicates

38
 1-male
2-female

1-A
2-B
3-AB
4-O

39
 2011.4.6
40
 2011.4.6
41
 How to eat carrot for best nutrition

Cumulative
Frequency Percent Valid Percent Percent
Valid I don't know 24 2.6 2.6 2.6
eat directly 477 51.2 51.5 54.1
eat after water-boiled 185 19.8 20.0 74.1
eat after oil-fried 240 25.8 25.9 100.0
Total 926 99.4 100.0
Missing 9 6 .6
Total 932 100.0

Table3-1 “How to eat carrot for best nutrition” question's frequency table
Item Frequency Percent Valid Percent Cumulative Percent
eat directly 477 51.2 51.5 51.5
eat after water-boiled 185 19.8 20.0 71.5
eat after oil-fried 240 25.8 25.9 97.4
I don't know 24 2.6 2.6 100.0
Missing 6 0.6 — —
total 932 100.0 100.0 —

42
 sex * status of study Crosstabulation

Count
status of study
bad good Total
sex male 192 299 491
female 137 280 417
Total 329 579 908

Chi-Square Tests

Asymp. Sig. Exact Sig. Exact Sig.

Value df (2-sided) (2-sided) (1-sided)
Pearson Chi-Square 3.813b 1 .051
a
Continuity Correction 3.547 1 .060
Likelihood Ratio 3.824 1 .051
Fisher's Exact Test .053 .030
Linear-by-Linear
3.808 1 .051
Association
N of Valid Cases 908
a. Computed only for a 2x2 table
b. 0 cells (.0%) have expected count less than 5. The minimum expected count is 151.
09.

2011.4.6
43
 a
Coefficients

Unstandardized Standardized
Coefficients Coefficients
Model B Std. Error Beta t Sig.
1 (Constant) 142.386 5.252 27.109 .000
age 1.243 .258 .157 4.822 .000
a. Dependent Variable: Height

For test:
test
t = t-Value,
Value Sig.=P-Value,
Value
P ＜ 0.05 means refuse H0 : β ＝ 0.
For Regression Equation:
Equation

Yˆ 142.386  1.243age

2011.4.6
44
 Correlations

Height Weight
Height Pearson Correlation 1 .721**
Sig. (2-tailed) .000
N 930 930
Weight Pearson Correlation .721** 1
Sig. (2-tailed) .000
N 930 931
**.Correlation is significant at the 0.01 level
(2-tailed).
For test:
test
Sig. (2-tailed) =P-Value,
Value
P ＜ 0.05 means refuse H0 : ρ ＝ 0.
For Pearson Correlation coefficient:
coefficient

r = 0.721
2011.4.6
45
Thank you!

2011.4.6
46