Class Notes
(807,235)

Canada
(492,664)

University of Guelph
(25,939)

SOAN 2120
(388)

David Walters
(110)

Lecture

# November 6.docx

Unlock Document

University of Guelph

Sociology and Anthropology

SOAN 2120

David Walters

Fall

Description

November 6 , 2012
In your project you will be required to include three tables…
1) Table 1: descriptive statistics (univariate)
2) Table 2 and 3: bivariate statistics (i.e. cross-tabs, comparing means, or
correlations)
Descriptive Statistics
Gender
Income
Before you run any analyses, weight your data using the weight variable in SPSS.
In SPSS: data -> weight cases -> weight cases by (select the weight variable at the
end of the list “weight” -> Ok
For categorical variables (gender)
Analyze -> descriptive statistics -> frequencies -> sex -> ok
For quantitative variables (income – dvr47gr1)
Analyze -> descriptive statistics -> frequencies -> dvr47gr1 -? Statistics - > click
mean -> ok
For quantitative variables sometimes we need to know the way in which the
variable is coded
- what does a mean of 7 mean?
- Interpret your results of the mean category
- In the SPSS data editor go to the variable view -> find the variable (dvr46gr1)
and click on values. This will give you a list of the correspondence between
the coded values and actual earnings.
- Thus, the mean (category) corresponds with $20,000 - $24, 999 (note: the
analysis includes those without any income – category 0)
(INSERT TABLE 1 HERE: DESCRIPTIVE STATISTICS FOR THE VARIABLES IN
THIS STUDY… it might be on CourseLink in content under week 10?)
Statistical Analysis
- one quantitative variable and one categorical variable
- comparing means (ANOVA)
- in SPSS: analyze --> general linear method univariate dependent
variable: dvr47g1 fixed factor: dvsex options parameter estimates
(definition: extra earning advantage of males over females) continue ok
Translating the results
The average yearly earnings for males = 5.597 + 3.225 (second value on bottom
table, lower left hand side) = 8.8222 ($25,000 - $29,999) The average yearly earnings for females = 5.597 (intercept) +0 = 5.597 ($10,000 -
$14,999)
To figure this out you must go to the variable view and click on values. Make sure
you understand how to do this.
Reporting p-values in tables
* = means .01 > .001
.021 > .01
.01 > .002 > .001
.01 > .0012 >.001
What if? UNDERSTAND THIS SLIDE…
- the p-value in the Sig. column of SPSS is .06. Are your results statistically
significant? And if so, at what alpha level?
o .06 is greater than .05 and therefore not statistically significant
- the p-value in the Sig. column of SPSS is .052. Are your results statistically
significant? And if so, at what alpha level?
o No, 0.52 is greater than .05 and therefore not statistically significant.
We fail to reject the null hypothesis.
- The p-value in the Sig. column of SPSS is .009. Are your results statistically
significant? And if so, at what alpha level?
o .01 (choose this alpha because less then .01 but not less then .001) –
two **
- .000
o significant at .001
INTERPRETING THE RESULTS IN YOUR RESULTS SECTION
Remember: The purpose of the statistical analysis is to identify relationships
(whether they are significant)
The relationship between gender and income is statistically significant (p

More
Less
Related notes for SOAN 2120