Lab 6. Producing Bivariate Tables and Using Measures of Association and Chi Square Statistics to Understand Them.

Producing Bivariate Tables and Using Measures of Association and Chi Square Statistics to
Understand Them. Due Dec 4, 2024 by midnight in the lab 6 folder of Blackboard
This lab will introduce you to some more of SPSS’s capabilities for bivariate crosstabulation. The
focus here is on computing and interpreting crosstabulation tables, measures of association and chi
square statistics. The data set will be the GSS2008_lab6.SAV, the 2008 GENERAL SOCIAL
SURVEY DATA set we have been using most of the semester. The data set and the codebook
(2008_GSSCODEBOOK.pdf) are in the Lab 6 folder in blackboard. Download them to a lab 6
folder on your computer to complete the assignment, and then move them to the virtual desktop to
work in SPSS. Make sure to create a lab 6 word document to receive your answers. BE SURE TO
ORGANIZE YOUR ANSWERS USING THE NUMBERING SYSTEM OF THE ASSIGNMENT.
USE FULL SENTENCES TO ANSWER ALL QUESTIONS. There will be penalties for
disorganization. ALL QUESTIONS WEIGHTED EQUALLY. Once the assignment is complete,
upload the word document that has your answers to the lab 6 folder in blackboard.
Part A.
In part A of this lab, we will look at attitudes about taxes on the rich and their relationship to race.
One of the variables used is TAXRICH1, a measure of people’s attitudes towards taxes on rich
people. The variable is measured by the question “Generally, how would you describe taxes in
America today. We mean all taxes together, including social security, income tax, sales tax, and all
the rest: First, for those with high incomes, are taxes too high, about right or too low” and is located
on page 464 of the codebook. The second variable is RACE (RACE-white, black or other). Race is
located on page 316 of the codebook.
1) Make a hypothesis for the relationship between TAXRICH1 and RACE. Make TAXRICH1
the dependent variable and RACE the independent variable. Please make a directional
hypothesis.
2) Provide a theoretical rationale for the hypothesis you made in question 1.
3) What are the levels of measurement for the dependent and the independent variables?
4) What measures of association are appropriate for assessing the strength of the relationship
between TAXRICH1 and RACE?
5) Follow these instructions to perform the crosstabulation analysis to examine your hypothesis
a. Click ANALYZE, DESCRIPTIVE STATISTICS, CROSSTABS.
b. Select TAXRICH1 from the left column and place in the ROW box as the dependent
variable. NOTE:THE VARIABLE IS TAXRICH1 – NOT TAXRICH!
c. Select RACE from the left column and place it in the COLUMN box as the independent
variable.
d. Click on CELLS to open the CROSSTAB CELL DISPLAY and make sure
OBSERVED in COUNTS and COLUMN in PERCENTAGES are checked. Click
CONTINUE
2
e. Click STATISTICS and then click LAMDA and PHI AND CRAMER’S V under
NOMINAL and CHI –SQUARE. Then click CONTINUE and then OK to obtain
results.
6) Copy and paste the CROSSTABULATION table, the CHI SQUARE TEST table, the
DIRECTIONAL MEASURES table, and the SYMMETRIC MEASURES table into your lab
document. If the table is too wide for your page, right-mouse click in the middle of the table
and select TABLE PROPERTIES and select CENTER and change the TABLE WIDTH to
about 8 inches.
3
4
7) Discuss the relationship depicted in the crosstabulation table, using selected percentage data to
support your claims.
8) Is the relationship observed in your crosstabulation results a weak, moderate or strong one?
Report the LAMBDA (look in the value cell of the DIRECTIONAL MEASURES table – the
attitudes about taxes for the rich dependent) and CRAMER’S V (look in the value cell of the
SYMMETRIC MEASURES table for Cramer’s V) produced by SPSS to defend your answer.
Remember that a value of Lambda and Cramer’s V between 0-0.3 means weak relationship,
0.3 to 0.5 is moderate and 0.5+ is strong.
9) Is the relationship depicted in the crosstabs statistically significant or not? Look in the chi
square test table for the relevant information. Report the value of the PEARSON CHI
SQUARE, the DEGREES OF FREEDOM(DF) and the probability associated with the chi
square(look in THE ASYMPTOTIC SIGNIFICANCE CELL OF THE PEARSON CHI
SQUARE for this probability). Remember that when this probability is less than or equal to
.05, the relationship observed in the crosstab is statistically significant. You must then reject
the null hypothesis of no relationship in the population between your dependent and
independent variables, and be left with the alternative hypothesis that there is a relationship
between the two variables in the population. However, if the probability associated with the
chi square statistic is larger than 0.05, then the relationship is statistically nonsignificant,
meaning that the relationship is likely due to random chance and you cannot reject the null
hypothesis of no relationship in the population.
10) Is the relationship observed in the crosstabulation table consistent or not consistent with
your hypothesis? If you cannot reject the null hypothesis of no relationship then you have to
conclude that the relationship observed in the crosstabulation table is not consistent with your
hypothesis, unless you made a null hypothesis, which is not usually done. However, if you can
reject the null hypothesis, you may still conclude that the relationship is not consistent with
your hypothesis. You have to compare the relationship observed in the crosstabulation table
with the hypothesis you made to decide whether the former is consistent or not consistent with
the latter. Please explain your answer in detail.
Part B
In this part we will look at the relationship between attitudes towards capital punishment and race.
The two variables to be used are CAPPUN “Do you favor or oppose the death penalty for persons
convicted of murder?” on page 28 of the codebook and RACE (white, black, other). Inspect the two
variables in the codebook to make sure you understand them and then answer the following
questions.
1) Make a hypothesis involving CAPPUN and RACE. Treat CAPPUN as the dependent variable
and RACE as the independent variable. Again please make a directional hypothesis.
5
2) Provide a theoretical rationale for the hypothesis you made in question 1.
3) What are the levels of measurement for the dependent and the independent variables?
4) What measures of association are appropriate for assessing the strength of the relationship
between CAPPUN and RACE?
5) Follow these instructions to perform the crosstabulation analysis
a. Click ANALYZE, DESCRIPTIVE STATISTICS, CROSSTABS.
b. Select CAPPUN from the left column and place in the ROW box as the dependent
variable.
c. Select RACE from the left column and place it in the COLUMN box as the independent
variable.
d. Click on CELLS to open the CROSSTAB CELL DISPLAY and make sure
OBSERVED under COUNTS and COLUMN in PERCENTAGES are checked. Click
CONTINUE
e. Click STATISTICS and then click LAMDA and CRAMER’s V under NOMINAL and
CHI –SQUARE and click CONTINUE. Then click OK to obtain results.
6) Copy and paste the crosstabulation table, the chi square test, the directional measures table and
the symmetric measures table into your lab document. If the table is too wide for your page,
right-mouse click in the middle of the table and select TABLE PROPERTIES and select
CENTER and change the TABLE WIDTH to about 8 inches.
7) Discuss the relationship depicted in the crosstabs table, using selected percentage data to
support your claims.
8) Is the relationship observed in your crosstabulation results a weak, moderate or strong one?
Inspect the LAMBDA and CRAMER’S V produced by SPSS and use them to answer. Be sure
to report numbers to support your statements.
9) Use the chi-square statistic (PEARSON CHI-SQUARE) and its degrees of freedom (DF) to
examine whether the relationship is statistically significant or not. Is the relationship depicted
in the crosstabs statistically significant or not? Look in the chi square test table for the relevant
information. Report the value of the PEARSON CHI SQUARE, the DEGREES OF
FREEDOM(DF) and the probability associated with the chi square(look in THE
ASYMPTOTIC SIGNIFICANCE CELL OF THE PEARSON CHI SQUARE for this
probability). Remember that when this probability is less than or equal to .05, the relationship
observed in the crosstab is statistically significant. You must reject the null hypothesis of no
relationship between your dependent and independent variables. You will therefore be left
with the alternative hypothesis that there is a relationship between the two variables in the
population. However, if the probability associated with the chi square statistic is larger than
0.05, then the relationship is statistically nonsignificant, meaning that the relationship is likely
due to random chance and you cannot reject the null hypothesis of no relationship in the
population.
10) Is the relationship observed in the crosstabulation table consistent or not consistent with
your hypothesis? If you cannot reject the null hypothesis of no relationship then you have to
conclude that the relationship observed in the crosstabulation table is not consistent with your
hypothesis, unless you made a null hypothesis, which is not usually done. However, if you can
6
reject the null hypothesis, you may still conclude that the relationship is not consistent with
your hypothesis. You have to compare the relationship observed in the crosstabulation table
with the hypothesis you made to decide whether the former is consistent or not consistent with
the latter. Explain whatever answer you give.