Below are a number of research questions that can be answered by the techniques presented in Chapters 6 through 10. For each question, select the most appropriate test, compute the necessary statistics, and state your conclusions
To complete some problems, you must first calculate the sample statistics (e.g., means or proportions) that are used in the test of hypothesis. Use alpha = 0.05 throughout. The questions are presented in random order. There is at least one problem for each chapter, but we have not included a research situation for every statistical procedure covered in the chapters.
In selecting tests and procedures, you need to consider the question, the number of samples or categories being compared, and the level of measurement of the variables. These problems are based on social survey data, shown below, collected from a random sample of 25 Canadian adults, and you may have to violate some assumptions about sample size in order to complete this exercise. Abbreviated versions of the survey questions along with the meanings of the codes are also presented.
a. Is there a statistically significant difference in average hours of TV watching
by income level? By sex?
b. Is there a statistically significant relationship between age and happiness?
c. Estimate the average number of hours spent watching TV for the entire population.
d. If Canadians currently average 2.3 children, is this sample representative of the population?
e. Are the educational levels of political conservatives and political liberals significantly different?
f. Does average hours of TV watching vary by level of happiness?
g. Based on the sample data, estimate the proportion of female Canadians in the population.