Data Analysis Assignment B

Part 1. Use the data contained in the Excel dataset entitled "Violence on TV", located in Doc Sharing. This is data from a survey of adults about violence on television. The definitions of the data values are shown below. Read the data directly from Excel into SPSS . You must enter the variable labels and the labels for the data values into SPSS using the Variable View.

Answer the following questions:

1. Provide basic descriptive statistics for each of the data elements.
2. Treating the "Opinion About Violence" variable as a scalar (scaled variable), answer the following questions:
   1. Do males perceive TV to have more, or less, violence than do females? Using the t-test method, is this difference statistically significant?
   2. Do married people perceive TV to have more, or less, violence than do unmarried people? Using the t-test method, is this difference statistically significant?
   3. Do people with children living at home perceive TV to have more, or less, violence than do people without children living at home? Using the t-test method, is this difference statistically significant?

1. Using the graphing ability of SPSS , produce a bar-chart graph displaying how the average value for Opinion About Violence varies by the values of each of these three categorical variables, and provide an interpretation of your findings:
   1. Income
   2. Education
   3. Amount of TV watched per week
2. Treating Opinion About Violence as a categorical variable, produce a cross-tabulation of respondent age vs. opinion about violence, including a chi-square calculation. What do the table (and its significance as measured by the chi-square statistic) suggest about the association between age and perception of violence?

Survey About Violence on TV: data definitions

Part 2: Analyze the Healthy Lifestyles data again, this time using the version called Healthy Lifestyles Expanded 6-11. Use the regression option, not GLM . Your job is to develop a model that 'explains' the death rate in each state, using the other variables as predictor variables.

The added variables are:

dethrate = The number of deaths per year per 100 population.

medianage = the median age of the population

airparticulates: 1 = the state fails the U.S. standard for particulates in the air; 0 = does not fail

airozone: 1 = the state fails the U.S. standard for ozone in the air; 0 = does not fail

incpercapita = average annual household income per person

Even though the two air-related variables are categorical, you may treat them as scalar for this purpose, since they are binary – i.e. you may use them in the regression analysis. Create the 'best' model you can, in the sense that the explanatory power is high without excessive predictor variables. Therefore, in your final model, all the independent variables should be significant, and no significant variable should be omitted. You may use stepwise regression in some way to help arrive at the final model, or not, as you prefer.

Show some of your steps to illustrate how you arrived at your final model, and explain what your logic was for each step.

Explain the final model in plain English, and comment on what implications it has for public policy.

Price: $28.64

Solution: The downloadable solution consists of 18 pages, 1064 words and 19 charts.
Deliverable: Word Document