代写spss统计专业、Statistical Theory代写

日期：2018-04-19 08:37

960:501– Statistical Theory for Research Workers– Spring, 2018

Homework 5, due

1. The data at fev.dat.txt

contain data on lung function. The file has five fields, age (years), forced expiratory

volume (liters), height (inches), sex (female coded as zero, male coded as 1), and

smoking status (nonsmoker coded as zero, smoker coded as 1).

a. Fit a model that expresses mean FEV as a linear function of age in years. Test the null

hypothesis that FEV has no relation to age, vs. the alternative that there is some linear

relation.

b. Plot the relationship betwen age and FEV, and construct residual plots, both vs. age

and vs. index. Comment on whether regression model assumptions appear to hold.

2. The data at fev.dat.txt

contain data on lung function. The file has five fields, age (years), forced expiratory

volume (liters), height (inches), sex (female coded as zero, male coded as 1), and

smoking status (nonsmoker coded as zero, smoker coded as 1).

a. Fit a model that expresses mean FEV as a linear function of age in years. Give a

confidence interval for the quantity by which FEV changes as age increases by one

year.

b. Identify a measure of strength of association between age and FEV, and interpret it.

c. Give a 95% confidence interval for the FEV of an individual aged 30.

3. The data at fev.dat.txt

contain data on lung function. The file has five fields, age (years), forced expiratory

volume (liters), height (inches), sex (female coded as zero, male coded as 1), and

smoking status (nonsmoker coded as zero, smoker coded as 1).

a. Fit a model that expresses mean FEV as a linear function of age in years, and height

in inches. Give a confidence interval for the quantity by which FEV changes as age

increases by one year, and for the quantity by which FEV changes as height increases by

1 inch.

b. Test the null hypothesis that the in the above model, expected FEV depends on

height.

c. Give a measure for the strength of the relationship explaining FEV on height and age;

choose a measure that adjusts for the fact that two, rather than just one, variables is

used.

1

960:501– Statistical Theory for Research Workers– Spring, 2018

4. Key Words: Test of independence and trend in 2x5 table Kane (2001) presents data

on the amount of video game usage allowed to 100 children, and whether these children

have disicpline problems in school. These data are summarized below.

0 Hours 1–3 Hours 4–6 Hours 7–10 Hours > 10 Hours

Discipline Problems 0 2 2 0 2

No Discipline Problems 11 43 19 1 15

a. Test the hypothesis that exposure (that is, amount of time children are permitted to

use video games) is associated with the disease status (that is, whether the children

have discipline problems.) Comment on on any reasons your calculations might be

suspect. Do not use the ordering of the categories. Interpret your results.

b. There is a part of these data that the above quesiton does not use. What is it?

5. Key Words: ANOVA The data set at

represent blood levels of various chemicals in elderly patients. See for details.

These are measured in one of six labs. Of interest are fifth field, indicating lab, and the

sixth field, indicating calcium level. Test the null hypothesis that all labs have the same

mean calcium level, vs. the alternative hypothesis that these levels vary by lab.

6. Key Words: ANOVA The data set at

represent blood levels of various chemicals in elderly patients. Seefor details.

These are measured in one of six labs. Of interest are second field, indicating age,

fifth field, indicating lab, the sixth field, indicating calcium level, and the eigth field,

representing age group. Ignore ordering among the ages.

a. Fit a model that allows for effects due to lab, age group, and their interactions, and test

the null hypothesis that the interactions are all zero.

b. Fit a the analysis of covariance model that allows for effects due to lab, age, and their

interactions, and test the null hypothesis that the interations are all zero.