Take-Home/In-Class Final Exam – STAT 601

This exam uses the NC Birth Data from the year 2006. The variable descriptions are contained are found on the website and are available by clicking here. Also I have another device that can be used to help determine which method of analysis to use, it is linked to the website and available by clicking here.This table covers many of the methods of analysis that we used in this course, however it also covers several methods that we did not cover as well. You also have your all your course notes, assignments, etc. as a resource to help with these questions.

For each of the general research questions specify as completely as you can what statistical method(s) you would use in answering them. You should address what descriptive and inferential methods you would use for each. YOU DO NOT NEED TO CONDUCT THE ANALYSIS, JUST SAY WHICH METHOD YOU WOULD USE. If you want to play around with the data set to help you decide which methods to use I have provided a link to the data set on the website: NC Birth (Final 601).JMP.

Here is an example what I expect -

Is there evidence that the mean education level of the mother’s and father’s of infants born in North Carolina differs?

Answer: As these data consist of education levels of the mother and father of the same child they are meaningfully paired. Thus we have dependent samples and could use either a paired t-test or a Wilcoxon signed-rank test to compare the education levels. As we have a large sample size here we could comfortably use the paired t-test and CI for the paired difference.

Even though I am not asking you to conduct the analysis, feel free to include it in your write-up. Also because you have the data you can certainly confirm your choice of analysis by actually doing it in JMP.

QUESTIONS

1. Is the mean/median birth weight of infants born to women who smoked during pregnancy less than that for infants born to women who did not smoke during pregnancy? (3 pts.)

2. Is the proportion of infants born with low birth weights born to women who smoked during pregnancy less that that for the population of infants born to women who did not smokes during pregnancy? If so, what can you say about the risk of low birth weight associated with smoking during pregnancy? (4 pts.)

3. Adjusting for the following three potential confounding factors
(fill in the blank, choose three potential covariates)

______

______

______

what can we say about the difference in mean birth weight of infants born to women who smoked during pregnancy and that for infants who were born to non-smoking mothers? (4 pts.)

4. Is there a relationship between the month of the first prenatal visit and the years of formal education a mother has? If so, what is the nature of that relationship? (2 pts.)

5. Researchers are interested in determining if the computed gestational age and the one appearing on the birth certificate differ (see variables Gest Age and Gest Age (BC)) (2 pts.)

6. Is there an association between hypertension during pregnancy (HYPERPR) and low birth weight status (Low Birth)? Is so what is the nature of the relationship? (2 pts.)

7. Is there evidence that a majority of infants born in North Carolina have 1 minute APGAR scores of 8 or higher? (2 pts.)

8. What is the percentage of infants in North Carolina that are born vaginally? (2 pts.)

9. What is the mean weight gain during pregnancy for women in North Carolina? (2 pts.)

10. Is there an association between minority status of the mother (Mother Minority) and the Kotelchuck Index? (2 pts.)

11. Do the mean birth weights of infants differ across populations of mothers determined by Kotelchuck’s Index of the Adequacy of Prenatal Care? If so, which populations significantly differ in terms of mean birth weight? (3 pts.)

Now come up with your own…

12. Give an example of a question that could be answered using each of the following
statistical methods: (3 pts.each)

a)One-sample t-test

b)Two-sample t-test

c)Fisher’s Exact Test

d)Paired t-test

e)One-way ANOVA

f)Chi-square test of independence

g)Pearson’s product moment correlation (r)

h)Simple linear regression

i)Multiple regression

j)Kruskal-Wallis Test

k)Mann-Whitney Test

l)Wilcoxon Signed-Rank Test

For each state the research question that would be addressed and list the variables involved (using the variable abbreviations is fine).

1