Stat work

Statistics Work

1."Graphs That Enlighten and Graphs That Deceive" which, of course, need not be just statistical in nature. Find a graph of one of the types shown in the chapter (e.g. bar graph) in current news at an online news Web site (e.g. newspaper, TV or radio stations and networks), give the URL of the article, and critique the graph. 

In your critique I am interested in 2 things:

· The proper construction of the graph as detailed in Chapter 2-4.

Other relevant comments about the quality of the graph including sources of data.

2. Why is the median price used to reflect average home prices in most areas, rather than the mean or mode?  Should all people have similar houses or is it fair that some houses are much more expensive than others?  Why or why not?

3. Construct 3 questions that could be answered using the Binomial Distribution and explain why you know the Binomial Distribution is appropriate to use in each case.

 

4. 2. Why is there more than one process for the Normal Distribution? Why are extra steps needed if the mean is not zero and the standard deviation is not one?

5. Explain what a confidence interval is and how it helps you to better understand the parameter of the population you are interested in?

· 6. Make up 4 different claims that can be tested by a test of a hypothesis.

· For each claim, construct the corresponding hypotheses, i.e. H0 and H1.

· Make up an initial conclusion about H0 for each claim.

· Write the final conclusion for each claim using the above

7. The assignment is to fully describe a situation (not copied from the text or the Internet) where you believe a statistical correlation or regression would be of interest and describe the study and how linear correlation or regression would further the aims of the study. For this assignment - no more than a paragraph or two is needed:

· Describe the study idea generally but clearly and specify if you will be using correlation or regression.

· Describe the sampling plan and the exact paired variables to be measured which should be quantitative. (Using qualitative variables or having more than two variables or trying to compare two or more different samples are common errors).

· Describe how you will verify that the requirements for either correlation or regression are met.

· Describe the utility of the result (either correlation or regression equation) - i.e. why a successful outcome of the study would be useful. Regression is usually easier.

Some tips:

1 Correlation and regression have popular definitions and statistical definitions. Make sure you are using the statistical definitions in this assignment.

2 No data or calculations are required.

3 A simple study is preferred. This course does not cover how you would handle large numbers of variables.

Don't describe a study where you compare two (or more) different groups or samples. That is not covered in this course either.