SW388R6

1
Crosstabulated Table Problems

• Using cross-tabulated tables
• Dependent and independent variables
• Number, proportion, or probability for sample
• Number, proportion, or probability within
  category
• Comparing the likelihood for two groups

2
Comparison of groups independent and dependent
variables

• Very often, our research involves a comparison of
  the characteristics of two or more groups within
  the data set.
• Comparisons utilize two variables from the data
  set one variable defines the groups that will be
  compared the second variable defines the trait
  or characteristic on which we are comparing the
  groups.
• The variable that defines the groups is called
  the independent variable, while the variable
  being compared is called the dependent variable.

3
Relationship between variables

• The basis for the comparison of groups is the
  expectation that different groups will have
  different numbers or proportions of subjects in
  the categories of the dependent variable.
• When one group had a higher proportion of members
  in a category of interest than the other group,
  we interpret this difference as evidence that
  there is a relationship between the independent
  and dependent variables.
• The relationship between variables is usually
  identified as the influence which the independent
  variable has on the dependent variable.

4
Constructing a cross-tabulated table

• A cross-tabulated table is constructed by
  counting the number of cases that appear in each
  of the combinations of categories for the two
  variables.
• By convention, the independent, or grouping,
  variable is used for the columns of the table.
  The dependent, or test, variable is used in the
  rows.
• Percentages for a cross-tabulated table can be
  computed across rows, down columns, or for each
  cell in the table. The availability of multiple
  percentages for tables is a source of confusion.

5
Request the cross-tabulated table
To compute a frequency distribution in SPSS,
select the Descriptive Statistics Frequencies
command from the Analyze menu.
To compute a crosstabulated table in SPSS, select
the Descriptive Statistics Crosstabs command
from the Analyze menu.
6
Specify the column variable
First, highlight the variable for which we want
SPSS to use in the columns of the table.
Second, click on the arrow button to move the
highlighted variable to the list box of variables.
7
Specify the row variable
Second, click on the arrow button to move the
highlighted variable to the list box of variables.
First, highlight the variable which we want SPSS
to use in the rows of the table.
8
Request the dialog to specify cell contents
First, click on the "Cells" button to open a
dialog box that lets us specify what will appear
in the cells of the table.
9
Specify cell contents
First, we click on the "Column" checkbox in the
Percentages panel to include column percents in
the table.
Third, click on the Continue button to close this
dialog.
Second, we click on the "Total" checkbox in the
Percentages panel to include total percents in
the table.
10
Complete the cross-tabulated table request
First, click on the OK button to complete the
request for the cross-tabulated table.
11
The crosstabulation output
The crosstabulation table will be used to answer
a variety of questions.
12
Problem 1 Number, proportion, or probability for
sample

• Based on the dataset GSS2000.SAV, is the
  following statement true, false, or an incorrect
  application of a statistic?
• Among all survey respondents in the sample, the
  percentage who were male and who said they had a
  great deal of confidence in the U.S. Supreme
  Court was 30.1.
• 1. True
• 2. False
• 3. Incorrect application of a statistic

13
Solution 1 Number, proportion, or probability
for sample
The asks about males and subjects who had a great
deal of confidence in the Supreme Court. The
answer will be in the cell defined by those two
categories.
When the question asks for the percentage for all
cases in the sample, we use the of total to
answer the question. The total percentage who
were male and who said they had a great deal of
confidence in the U.S. Supreme Court is 12.6,
not the 30.1 stated in the question. The answer
to the question is false.
14
Problem 2Number, proportion, probability within
category

• Based on the dataset GSS2000.SAV, is the
  following statement true, false, or an incorrect
  application of a statistic?
• Within the group of survey respondents who were
  female, the probability that a survey respondent
  said he or she had hardly any confidence in the
  U.S Supreme Court was 0.176.
• 1. True
• 2. False
• 3. Incorrect application of a statistic

15
Solution 2 Number, proportion, probability
within category
When the question asks for the percentage within
a specific group of respondents, we use the
within the category to answer the question. The
percentage in the female column who said they had
a hardly any confidence in the U.S. Supreme Court
is 17.6, equivalent to a probability of
0.176. The answer to the question is true.
16
Problem 3Comparing the likelihood for two groups

• Based on the dataset GSS2000.SAV, is the
  following statement true, false, or an incorrect
  application of a statistic?
• When survey respondents who were male were
  compared to survey respondents who were female,
  those who were male were more likely to have said
  they had a great deal of confidence in the U.S.
  Supreme Court.
• 1. True
• 2. False
• 3. Incorrect application of a statistic

17
Solution 3 Comparing the likelihood for two
groups - 1
The proportion of males who had a great deal of
confidence in the Supreme Court is 30.1.
The proportion of females who had a great deal of
confidence in the Supreme Court is 31.4.
18
Solution 3 Comparing the likelihood for two
groups - 2
Since the proportion within males (30.1) is less
than the proportion within females (31.4), males
were less likely to have a great deal of
confidence in the Supreme Court. The answer to
the question is false.