1b. What happened to the value of t where n = 31 when the alpha-level changed? Why?
2a. Using the SPSS output below, construct three separate 95% confidence intervals for respondent’s age when they had their first child, their highest years of education, and the number of hours they report watching television. Show your work.
Descriptive Statistics
|
|
|
|
|
|
|
N
|
Minimum
|
Maximum
|
Mean
|
Std. Deviation
|
R's age when 1st child born
|
1666
|
12
|
51
|
24.30
|
5.738
|
Highest year of school completed
|
2345
|
0
|
20
|
13.73
|
2.974
|
Hours per day watching TV
|
1555
|
0
|
24
|
2.94
|
2.837
|
2b. Describe your results in a paragraph.
In 2018, Gallup published the results of a poll on legalization of marijuana and found that 66% of respondents supported legalization, an all-time high. They note that “[r]esults are based on telephone interviews conducted October 1-10, 2018, with a random sample of 1,019 adults, ages 18+, living in all 50 U.S. states and the District of Columbia.”
3a. What is the 95% confidence interval for this result?
3b. What would happen if we wanted to construct a confidence interval for only that part of the sample under the age of 40? To answer this, you do not need to calculate anything.
SPSS Problems
For all SPSS problems, you should include relevant SPSS output (not the entire log) to support your answer. Click here for information on accessing the GSS 2018 data file and using SPSS.
We’ll explore support for capital punishment along with other opinions among GSS 2018 respondents. Get frequency distributions for two variables – SPANKING and CAPPUN (go to Analyze --> Descriptive Statistics --> Frequencies). Be sure to check out the codebook for each variable to see how the questions are worded. NOTE: I don’t want YOUR answers to these questions, just a summary of the GSS2018 data.
4a. Include your SPSS output for these frequency tables.
4b. Describe your results for each variable in one well-formulated paragraph. You should not relate them to each other – in other words, separate univariate statistics aren’t able to tell us whether those who support capital punishment are more or less likely to support spanking. We’ll get into multivariate statistics later.
Produce descriptives for the variable PAEDUC (respondents’ fathers’ level of education), MAEDUC (respondents’ mother’s level of education) and EDUC (respondents’ own level of education) by going to Analyze --> Descriptive Statistics --> Descriptives and choosing the variables noted above.
5a. Include your output.
5b. Compare respondents’ educational levels relative to their parents by describing their mean levels of education along with the standard deviations for each. Since we’re not yet performing tests of statistical significance, your comparison should simply describe any differences but not use wording such as “significant differences.”