Cross-Tabs and Chi-Square
1. Using the GSS12data, create a crosstab examining the relationship between DEGREE (level of education; independent variable) and RUDEWK (how often the respondent is treated rudely at work; dependent variable). Do you observe a relationship between people's education level and what they report on this question (is there a statistically significant relationship between these two variables in the population)? Test this relationship with the appropriate significance test. Analyze > Descriptive Statistics > Crosstabs > Row: (DV) > Column: (IV) > Cells: Percentages > Statistics: Chi-Square.
a. Provide your outputs from SPSS.
|
Case Processing Summary
|
|
|
Cases
|
|
Valid
|
Missing
|
Total
|
|
N
|
Percent
|
N
|
Percent
|
N
|
Percent
|
|
TREATED RUDELY AT WORK * RS HIGHEST DEGREE
|
1137
|
57.6%
|
837
|
42.4%
|
1974
|
100.0%
|
TREATED RUDELY AT WORK * RS HIGHEST DEGREE Crosstabulation
RS HIGHEST DEGREE Total
LT HIGH SCHOOL HIGH SCHOOL JUNIOR COLLEGE BACHELOR GRADUATE
TREATED RUDELY AT WORK Often Count 1 20 2 9 3 35
% within RS HIGHEST DEGREE 0.8% 3.7% 2.1% 3.6% 2.2% 3.1%
Sometimes Count 8 52 13 18 19 110
% within RS HIGHEST DEGREE 6.7% 9.7% 13.5% 7.2% 14.1% 9.7%
Rarely Count 15 117 25 77 40 274
% within RS HIGHEST DEGREE 12.6% 21.8% 26.0% 30.7% 29.6% 24.1%
Never Count 95 347 56 147 73 718
% within RS HIGHEST DEGREE 79.8% 64.7% 58.3% 58.6% 54.1% 63.1%
Total Count 119 536 96 251 135 1137
% within RS HIGHEST DEGREE 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%
|
Chi-Square Tests
|
|
|
Value
|
df
|
Asymptotic Significance (2-sided)
|
|
Pearson Chi-Square
|
32.851a
|
12
|
.001
|
|
Likelihood Ratio
|
34.104
|
12
|
.001
|
|
Linear-by-Linear Association
|
8.806
|
1
|
.003
|
|
N of Valid Cases
|
1137
|
|
|
|
a. 3 cells (15.0%) have expected count less than 5. The minimum expected count is 2.96.
|
b. What % of those with less than a HS education reports "never"?
c. What % of those with a bachelor's degree reports "never"?
d. What % of those with a graduate degree reports "never"?
e. What is the chi-square value?
f. What is the level of significance?
g. How would you formally interpret this result? (Give a Step 5 statement)
MoA
2. Consider the strength and direction of the relationship you examined in #1, by applying one of the measures of association. Examine the relationship and report your findings. Analyze > Descriptive Statistics > Crosstabs > Row: (DV) > Column: (IV) > Cells: Percentages > Statistics: **Choose the correct MOA**.
a. Provide your output from SPSS (do not need to include cross tabulation again only the association outcome).
b. What is the appropriate measure of association (MOA - lambda or gamma)?
c. Why?
d. Describe the relationship between the two variables, i.e. consider the strength and direction of the relationship.
e. Interpret the output in terms of the error that is reduced by knowledge of the independent variable.
Correlations
3. Is there an association between the violent crime rate (CRS35) and per capita state and local spending on police protection (LES65)? Use the STATES10 data to test this relationship.
a. First, run a scatterplot: Graphs > Legacy Dialogs > Scatter/Dot (simple scatter) > Y axis: LES65, X axis: CRS35. Next, generate a correlation matrix: Analyze > Correlate > Bivariate > Variables: LES65, CRS35> Options: Means and Standard Deviations. Provide your SPSS outputs including the Scatterplot, Descriptive Statistics, and Correlations.
b. What is the correlation (value) between these two variables?
c. What is the level of significance?
d. Describe the relationship between local/state spending and violent crime (what does the correlation coefficient tell you - hint: remember strength and direction)?
4. Select two interval-ratio level variables from any of the datasets listed in D2L and examine the relationship between them by generating a scatterplot and correlation matrix (as you did in #3 above).
a. Provide your SPSS outputs.
b. What is the correlation (value) between these two variables?
c. What is the level of significance?
d. Is the outcome of your test statistically significant?
e. Describe the relationship between your 2 variables.
Generate a regression output based on the two variables you identified in #4. Include the coefficients output and insert the values into the regression equation.