SUNY College at Buffalo
Sociology 301: Social Statistics
Study Guide for Exam #3
Dr. ZHANG Jie
**Chapter 7
Testing differences between means
Decision making
Testing hypothesis
Comparing groups
Differences due to pure chance or sampling errors
The null hypothesis:
two samples from the same population
Accept or reject the null hypothesis
The Research hypothesis:
real difference between two samples
Testing the hypothesis with the distribution of differences between means
the z score
the 95% confidence level
the t score
the level of significance
alpha value and P value
the degree of freedom (df)
Run tests with SPSS
**Chapter 8
Analysis or Variance (ANOVA):
to compare two or more means
Why not using t tests for comparing two or more means?
The logic of ANOVA
variation between groups and variation within groups
the null hypothesis
at least two of the groups differ
the level of measurement of the independent variables
Mean squares: used to calculate the F ratio
the MS between
the MS within
The F Ratio = MS between / MS within
Making decisions in ANOVA
calculating the F ratio
calculating the df's
setting an alpha value
finding the critical value in the table
comparing the obtained F with the critical value
rejecting the null hypothesis if F > the critical value
Making decisions in ANOVA with SPSS
conparing the obtained P value to the alpha
rejecting the null hypothesis if P < the alpha
Multiple Analysis of Variance (MANOVA)
multiple comparison of means
three or more variables
Requirements for using ANOVA, MANOVA, or F ratio
comparing two or more independent means
interval level of data for the dependent variable
random sampling
normal distribution
equal variance (homogeneity of variance)