Project Guidelines

Your final project assignment requires that you:

1. explore sources of data for your project - see links below,

2. identify and develop specific questions you wish to answer,

3. familiarize yourself with how the data you are using were collected and the limitations, if any, concerning the database.

I recommend that you begin by first exploring the databases to see what data are available and to begin thinking about the topics you are interested in and that you can develop questions about. For example, age specific mortality rates by state, region, gender, ethnicity, population density, year, etc. You may find that your research questions become more focused as you explore the databases and see what data are available. Once you begin to develop your questions you will need to begin identifying the appropriate statistics you'll need to apply to the data in order to answer your questions. In addition, you should begin to search for primary references (peer reviewed journal articles) for your topic. For example, I may initially ask whether access to special education services differs between states. The null statistical hypothesis would be that access doesn't differ between states and the alternative would be that it does differ. As I explore the database I will come to realize that this question will result in me trying to compare single values between each state (%receiving services) and so there is no statistic I can use. This often happens and so don't feel discouraged. You just need to revise your questions. I may then need to revise my question - such as, does access to these services differ between large urban centers and rural areas. Now I can do a literature search to see what researchers have found relating to my topic/questions.Next, I could then pool data for urban centers and rural areas across a state, or, if still resulting in a small sample size, I could pool data for these two categories across all states. At this point I would then consider using a t-test to compare mean proportion of children receiving services in urban and rural areas. When applying the t-test, I would have to test for equal variances and normality, state my null and alternative hypotheses, indicate whether it is a one-tail or two-tail test, and be sure to relate the statistical findings back to the question I was asking (summarize the findings as they relate back to your question).

A minimum of four statistical tests (not including tests for normality and equal variance) must be included in your report. For example, you may be interested in age specific mortality rates by state in the United States. Age would be your one category and your sample for each age would be the state specific mortality rates. This would be a one-way or one-factor ANOVA. If you also wished to consider the influence of ethnicity you could then run a second analysis with two categorical variables (AGE and ETHNICITY). This would be a two-way ANOVA.

SAMPLE DATABASES - there are more out there but these are good.
http://data.worldbank.org
http://www.womanstats.org
https://www.data.gov
http://www.childhealthdata.org/home
https://www.healthdata.gov
https://wonder.cdc.gov/controller/datarequest/D77

Structure of Final Report
12pt font
page numbers bottom center of each page except Title Page
1" margins

For references, use the "Name - Year" citation style - http://www.scientificstyleandformat.org/Tools/SSF-Citation-Quick-Guide.html

1. Title Page - Descriptive Project Title

a. "Fun with Statistics" is not a descriptive title. "The Impact of Race and Gender on Health Care Access in the United States" is a very good, descriptive title.

b. Your Name

c. BIO 598 and Due Date

2. Introduction

a. Provide background information on the topic you'll be exploring.

b. Specify the questions you are asking/hypotheses you'll be testing.

c. Please donot describe methodology or results in this section.

d. 1-2 pages should be sufficient - no more than 2pages.

3. Methods

a. Describe the database being used

b. Describe how the data were collected

c. Explain the statistical analyses you will use to answer each question you are asking.

d. 1-2 pages should be sufficient - no more than 2pages.

4. Results

a. Number all tables and figures and provide a brief description following each table or figure number.

b. Present detailed tables with your statistical results.

c. In paragraph form, describe your results being sure to refer to the relevant table and/or graph.

d. no more than 2 pages of written description of findings but you can include whatever number of pages of tables and graphs to convey your findings.

e. Please incorporate tables and graphs into the resuts section and don't make them too small. DO NOT PLACE THEM AT THE END OF THE REPORT.

5. Discussion
a. Relate your analytical findings back to the questions you were asking.
b. IF there were any limitations to the analysis, provide a brief explanation of these limitations and your recommendations for how these limitations can be avoided when data are collected in the future.
c. 1-2 pages should be sufficient - no more than 2pages.

6. Literature Cited

7. Use the "Name - Year" citation style - http://www.scientificstyleandformat.org/Tools/SSF-Citation-Quick-Guide.html