2 descriptive

7/28/2019 2 Descriptive

http://slidepdf.com/reader/full/2-descriptive 1/9

SPSS Workshop – Descriptive Statistics - 1 of 9

Resource: http://www.usd.edu/~aschwein/Stats/Tutorials/SPSS%20Tutorial.pdf

Descriptive Statistics & Frequency Tables

In this tutorial you will learn:1. How to compute basic descriptive statistics

2. How to split files

3. How to read SPSS Output4. How to create simple frequency tables

5. How to create frequency tables with two variables

Descriptive Statistics

• Convert large sets of data to more meaningful, easier to interpret, chunks or values.

• They summarize the data.

• Ex: mean, median, variance, and range.

SPSS contains a function that will compute many of these statistics.

• Analyze→ Descriptive Statistics→ Descriptives

A new window will appear with

two boxes. The box on the left

contains the variables with which

descriptive statistics may be

calculated (i.e., numeric

variables). Move the variables of

interest to the right-hand box with

the arrow button.

By default, SPSS will provide the

following: mean, standard

deviation, sample size, minimum

value, and maximum value. To select additional options or de-select default options,

click the Options button. Any item with a check-mark will

be computed, for each variable selected in the previous

step.

Under display order, you may select the order in which thevariables appear in the output.

When you are finished, click Continue to return to the

variable-selection window and click OK. SPSS will compute

the statistics and open a new window (Output) with the

descriptive statistics.





Reading SPSS Output

The following output is the result of selecting the variables age and score.

Notice there are two frames: the one on the left lists all the analyses available for

viewing, as well as any notes and titles relating to those analyses; the one on the right

depicts the results of the analyses.

The variables are listed on the left followed by the statistics we selected earlier. We can

see that the average age is 20.20 with a standard deviation of 2.484.

How to split groups

If you want to calculate descriptive statistics on sub-groups (such as males and females

separately), you may split your files. To do this: Data→ Split File

Determine how you want to split

your data set.

To separate by gender, select

Organize output by groups and

move the variable gender to the

right-hand window. Then select

OK.





When you run descriptive statistics on any variable, they will be reported for each level

of sex. See below:

Descriptives

gender = 1 MaleDescriptive Statistics(a)

N Minimum Maximum Mean Std. Deviation

age 12 18 22 19.58 1.505

score 12 45 98 80.08 15.986

Valid N (listwise) 12

a gender =male

gender = 2 FemaleDescriptive Statistics(a)


age 8 18 29 21.13 3.399

score 8 75 94 82.75 6.798


a gender =female

You will have to turn off this feature if you want to compute statistics for the whole

group, by selecting Analyze all cases, do not select groups.





Frequency Tables

Frequency tables include lists of values (categories) within each selected variables and

the number of times each category occurs.

To create a table of frequencies (number of occurrences of given categories), selectAnalyze→ Descriptive Statistics→ Frequencies.

Select the variables to be depicted in the frequency table by moving them from the

left- to the right-hand box.

SPSS provides the user additional options, including statistics, charts, and format:

Click: Statistics…SPSS will, by default, print the values of the selected variables and the frequencies of

each. If you prefer additional information, click Statistics:

Options include percentile values. SPSS will

print quartiles (fourths) or the values that

divide the data into X equal groups (cut

points). The number of groups is defined by

the user. SPSS will also print selected

percentiles. Simply, select Percentile(s), then

type in the percentile of interest and click

Add. We have selected thirds.

You may also select descriptive statistics,

like measures of central tendency anddispersion, as well as statistics describing the

distributions.

When finished, select Continue.

Click: Charts…This option creates bar charts, pie charts, or histograms in addition to the frequency

table. This might be useful if there are many categories for each variable or if two or

more variables are to be compared. The charts may contain frequencies or

percentages.

Click: Format…Use this option to determine the order

categories will appear and whether or not

multiple variables should be compared.

This will impact how results are presented.

To cut back on the amount of output,

users may choose not to view tables with





many categories. When finished click Continue to return to the variable-selection

window. Then click OK.

OutputThe new analyses are added to the descriptive statistics. Notice the addition in the left-

hand frame.

The following statistics are for

the variable age and score.

Notice there are a total of 20

cases, and none are missing.

The mean test score is 81.15

and 20.20 for age. The test

score value of 79.00 cuts off the

33rd

percentile (33% of cases fallat or below this value), and so

on.

If we scroll down the page, we will find additional results:

This table lists the values of the variable age and the frequency of occurrence of each.

age

Frequency Percent Valid PercentCumulative

Percent

18 5 25.0 25.0 25.0

19 4 20.0 20.0 45.0

20 3 15.0 15.0 60.0

21 5 25.0 25.0 85.0

22 2 10.0 10.0 95.0

29 1 5.0 5.0 100.0

Valid

Total 20 100.0 100.0





Next, SPSS provides a bar chart

depicting these frequency results, as

selected under Charts.

A similar table and chart were

created for the variable score.

292221201918

age

5

4

3

2

1

0

F r e q u e n c y

age





Frequency Tables with Two Variables

If you want to create a frequency table with two variables (crossed variables), such as

the number of males and females at each age, use the Crosstabs procedure.

Analyze→ Descriptive Statistics→ Crosstabs

This will allow the user to create a table

with one variable representing rows and

another representing columns. Select the

appropriate variables and move them to

the correct box.

SPSS will create tables with more than two

variables. Simply move the additional

variables to the Layer box.

The Statistics option allows for statistics

evaluating the association between

variables. Cells allow the user to define

what values to include in the cells. Format

provides the option to report categories in

ascending or descending order.

The output is as follows:

CrosstabsCase Processing Summary

Cases

Valid Missing Total

N Percent N Percent N Percent

gender * age 20 100.0% 0 .0% 20 100.0%

gender * age Crosstabulation

Count

age

18 19 20 21 22 29 Total

male 4 2 3 1 2 0 12gender

female 1 2 0 4 0 1 8

Total 5 4 3 5 2 1 20

Notice there are 5 eighteen-year-olds – 4 male, 1 female. There are 3 male twenty-year-

olds, etc.





Practice

1. Open the data “Practice Data – BMI”

2. Calculate the mean, standard deviation and range of BMI, height and weight across

all subjects.

3. Split the file by gender and calculate the mean and standard deviation of BMI for

males and females.

4. Remove the split and create a frequency table for gender.

5. Create a frequency table that includes both gender and height.

Output should look like:

2.

Descriptives

Descriptive Statistics


BMI 10 18.24 29.41 24.6491 3.86056

Weight 10 120 205 155.20 29.123

Height 10 60 71 66.50 3.629


3.

Descriptives

Sex = F

Descriptive Statistics(a)

N Mean Std. Deviation

BMI 5 22.6380 3.94790


a Gender =female

Sex = MDescriptive Statistics(a)

N Mean Std. Deviation

BMI 5 26.6603 2.79934Valid N (listwise) 5

a Gender =male





4.

Frequencies

Statistics

Gender

Valid 10NMissing 0

Gender

Frequency Percent Valid PercentCumulative

Percent

male 5 50.0 50.0 50.0

female 5 50.0 50.0 100.0

Valid

Total 10 100.0 100.0

5.Crosstabs

Case Processing Summary

Cases

Valid Missing Total

N Percent N Percent N Percent

Gender * Height 10 100.0% 0 .0% 10 100.0%

Gender * Height Crosstabulation

Count

Height Total

60 62 64 65 67 68 69 70 71

Gender male 1 0 0 0 1 0 1 1 1 5

female 0 1 1 1 0 1 1 0 0 5

Total 1 1 1 1 1 1 2 1 1 10

2 descriptive

Documents