SPSS I

Course Instructor Guide

Outline:

- Introduction
to SPSS
- Data
Entry
- Frequency
Distributions and Descriptive Statistics
- Output
- Edit
a Chart
- One
Sample t-Test
- Independent
Sample T-Test
- Correlation
- The
Process
- Next
week
- Open
class for Q&A

In Detail:

- Introduction
to SPSS
- Web
site

- SPSS

i.
Statistical Package for the Social Sciences

- Designed
to be a relatively comprehensive data analysis package for use in
research and business
- Since
this is a complex program dealing with statistics it is best if you have
a statistics background and understanding of

i.
Basic Terms

1. Case

2. Variables

a. Nominal

b. Ordinal

c. Interval

d. Ratio

3. Values

4. Normal
Distribution

5. Standard
deviation

6. Analysis

a. One-Sample
T-Test

b. One-Way
ANOVA

c. Correlation

7. It
is expected that you have a grasp of these

- Since
it is expected that you know statistics this course is designed to teach
statisticians how to use this SPSS application

i.
There is not enough time to cover both statistics and
the program

- Please
limit questions to how to use SPSS, not to statistics and quantitative
analysis
- Data
Entry
- Open
SPSS

i.
If a window appears asking, What
would you like to do?

1. Select

a. Type
in Data

b. OK

ii.
Toolbars

1. Cover
toolbars

- Data
editor

i.
You are currently in the SPSS Data Editor

ii.
Spreadsheet-type layout

1. Columns
and rows

2. Headings

3. Row
numbers (no value to these numbers)

4. Use
arrow keys to move from cell to cell

iii.
Type in data and name variables (two ways to type in)

1. After
typing in a number hit enter

a. 1-3,
then, 1,1,1, then 87-92

1 |
1 |
87 |

2 |
1 |
53 |

3 |
1 |
92 |

2. Type
a number then hit tab to complete the row

4 |
1 |
70 |

5 |
1 |
78 |

iv.
Coding

1. SPSS
is a quantitative analysis program

a. Works
with numbers

b. We
try to convert text to code

i.
Gender

1. F
= 1

2. M
= 2

v.
Show how to delete

1. Rows

2. Columns

3. Data

- Variable
View

i.
Choose bottom left tab, Variable View

ii.
Name

1. Variable
names can only be between 1-8 characters

2. Any
combination of numbers and letters, but must begin with a letter

3. Name
variables as: student, gender, and score

iii.
Decimals

1. Change
all to 0

iv.
Label

1. More
description of the field

2. score

a. Midterm
Score

v.
Values

1. Code
Information

a. Click
the button to the right of the text box

i.
Value, 1

ii.
Value Label, Female

iii.
Value, 2

iv.
Value Label, Male

vi.
Go back to Data View

- Frequency
Distributions and Descriptive Statistics
- Example

i.
15 men and 15 women in an introductory psychology class
have taken their midterms

ii.
In this example, we wish to construct frequency
distributions and obtain some basic descriptive statistics for the variables ** gender**
and

- Complete
the following

Student |
Gender |
Score |

1 |
1 |
87 |

2 |
1 |
53 |

3 |
1 |
92 |

4 |
1 |
70 |

5 |
1 |
78 |

6 |
1 |
73 |

7 |
1 |
91 |

8 |
1 |
60 |

9 |
1 |
77 |

10 |
1 |
82 |

11 |
1 |
85 |

12 |
1 |
33 |

13 |
1 |
88 |

14 |
1 |
98 |

15 |
1 |
88 |

16 |
2 |
89 |

17 |
2 |
73 |

18 |
2 |
91 |

19 |
2 |
76 |

20 |
2 |
75 |

21 |
2 |
89 |

22 |
2 |
81 |

23 |
2 |
83 |

24 |
2 |
68 |

25 |
2 |
86 |

26 |
2 |
55 |

27 |
2 |
89 |

28 |
2 |
89 |

29 |
2 |
70 |

30 |
2 |
93 |

- Data
Analysis

i.
On the menu toolbar choose Analyze > Descriptive Statistics > Frequencies

1. Move
the variables from the left to the right box

2. After
choosing which variables to use always check what options or preferences are
available before choosing OK

a. Move
gender and score

b. Check the Display Frequency Tables box

c. Choose
Statistics

i.
Check all boxes for dispersion, central
tendency, and distribution

ii.
Click continue

d. Choose
Charts

i.
Choose Histogram,
with

ii.
Choose Continue,
and OK

e. Choose Format

i.
Leave alone

f.
Click Continue,
and OK

- Output
- Show
how there are now two files open

i.
The data file is the .sav
file

ii.
The output file is the .spo
file

- Save
the data file as midterms.sav on the desktop
- Save
the output file as midterms.spo on the desktop

i.
Look on the desktop at the two files

- Go
back to the output file

i.
Discuss the right and left window views

ii.
Briefly discuss

1. Each
table

2. Both
Histograms

- Edit a
Chart
- Right
click on the Score Histogram

i.
Choose SPSS
Chart Object > Open

1. Now in a third window

a. Run
through the following

i.
The chart
options button

ii.
The line style button

iii.
Label style

iv.
The text button

v.
Select the columns (chart series), and use the color button

2. Close
the window

a. The
change takes place in the output file

- Close
the output file
- One
Sample t-Test
- In
the Data Editor Choose File > Open > Data

i.
In the SPSS I folder choose iqexample.sav

- Example

i.
The principal of the

ii.
In this problem we are testing the null hypothesis that
the mean IQ of all school children in the

- Analysis

i.
Choose Analyze
> Compare means > One-Sample T Test

1. Move
IQ to the box labeled Test Variables

ii.
Then, click the box labeled Test Value

1. Enter
105

iii.
Check Options

1. Confidence
Interval 95%

2. Exclude
cases analysis by analysis

iv.
Click OK

v.
In the output file, examine the results

1. Is
a mean difference of 5.73 large enough to be significantly different 105? The results of the t-test show that t=3.900,
with 29 (N-1) degrees of freedom. The
two-tailed p-value for this result is .001.
The result is considered statistically significant if the p-value is
less than the chosen alpha level usually .05 (and .01), so the result is
considered statistically significant and the null hypothesis is rejected

- Close
the output file
- Independent
Sample T-Test
- Open
independentsamplettest.sav
- Example

i.
The SFSU psychology department conducted a study to
determine the effectiveness of an integrated / experimental methods course as
opposed to the traditional method of taking the two courses separately. It was hypothesized that combining the two
would be better. To determine whether
there actually was a difference in student performance as a result of
integrated versus separate training; the final research projects of 20 students
from an integrated course and 20 students from the traditional course were
evaluated. There scores are listed in this
file.

ii.
In this problem we are testing the null hypothesis that
there is no difference in student performance as a result of the integrated
versus traditional courses, that is, that the mean difference between the
conditions in the population from which the sample was drawn is zero.

- Analysis

i.
Choose Analyze
> Compare Means > Independent-Samples T Test

1. Move
the dependent variable (score)
to the right

2. Move
the independent variable (cond) to the right (grouping variable)

a. The ?? means you need to Define Groups

i.
Cond group 1, and group 2

ii.
Type in 1, and then 2

3. Check
Options

4. Click
continue, and then OK

5. Output File

a. Discuss tables

i.
Levene’s Test for
Equality of Variances represents a test of the hypothesis that the
populations from which the groups were sampled have equal variances.

ii.
The most commonly used test is the row labeled Equal variances assumed. Because we are assuming that the two
population variances are equal, a pooled variance estimate is used to combine
the two sample variances to obtain the most accurate estimate of the variance
common to both populations

iii.
The observed t-value for this problem is 2.043, with
degrees of freedom equal to 38. The
two-tailed probability of .048 is less than .05 and, therefore, the test is
considered significant (thought barely) at the .05 level.

iv.
The null hypothesis is rejected at the .05 level of
significance

- Correlation
- Open
correlations.sav
- Example

i.
Next, Mr. Van Damme believes
that regular aerobic exercise is related to greater mental activity, stress
reduction, high self-esteem, and greater overall life satisfaction.

ii.
In this example each subject filled out a series of
questionnaires. The results are as
follows in this file.

iii.
In this problem, we are interested in calculating the
Pearson product-moment correlation between each pair of variables. In addition, for each pair we wish to test
the null hypothesis that the correlation between the variables in the
population from which the sample was drawn equals zero.

- Analysis

i.
Choose Analyze
> Correlate > Bivariate…

1. Move
all variables over to the right, except subject

2. Select

a. Pearson

b. Two-tailed

c. Flag Significant Correlations

3. Choose
Options

a. Check
Means and Standard Deviations

b. Click Continue

4. Click OK

b. Output

i.
Examine tables

1. Discuss
the similar data information

- The
Process

c. Entering
data and coding

d. Changing
variable information

e. Running
an analysis

i.
Choose the analysis

ii.
Choose variables to analyze

iii.
Set options

iv.
Run analysis

- Next week

f.
Recoding into different variables, cross tabulations,
regression, One-Way ANOVA

- Open class
for Q&A