HOMEWORK 1 – Due Tuesday, April 1, 2014 at 10:30 AM Table. List of variables Variable Name Description servitc Serum vitamin c level (continuous µmol/L) ageyrs age (years) height height in centimeters (cm) wt weight (kilograms) sex sex (men 1; women 2) race race (white 1; black 2; other 3) smokever smoking status (never smoker 0; ever smoker 1) booze alcohol consumption (number of drinks per week, continuous) 1. Create a new dataset by restricting the NHANES II dataset (nh2fs2014.sas7bdat, available from the course website under the Data 2014 page) to subjects with complete data on the covariates listed in the above Table. Then create the following variables: - BMI (continuous): calculated by [weight (kg)/ (height(m)) 2 ] - BMI_CAT (categorical): recode BMI_CAT as 0 if BMI < 19 (under-weight), 1 if 19 ≤ BMI < 25 (normal weight), and 2 if BMI ≥ 25 (over-weight). - Gender: recode sex as 0=male, 1=female - Alcohol consumption status (categorical variable): create categories for <3 drinks/week vs. ≥ 3 drinks/week. Please provide SAS code as the answer to this question. /* Q1 and Q2 */ data HW1; set "P:\Spring 2014\EPI204\nh2fs2014.sas7bdat" ; where servitc ne . and ageyrs ne . and height ne . and wt ne . and sex ne . and race ne . and smokever ne . and booze ne . ; race = race - 1 ; bmi = wt/((height/ 100 )*(height/ 100 )); if bmi < 19 then bmi_cat = 0 ; else if 19 <= bmi < 25 then bmi_cat = 1 ; else if bmi >= 25 then bmi_cat = 2 ; sex = sex - 1 ; if booze < 3 then booze_cat = 0 ; else if booze >= 3 then booze_cat = 1 ; sexbmi = sex*bmi; sexbmicat = sex*bmi_cat; run ; 2.How many observations were there in the original dataset? How many observations are in your new dataset? 3.Compute descriptive statistics for serum vitamin C level ((Mean, Median, Max, and Minimum). Graph the distribution in the form of a box plot and a histogram. Describe the distribution of serum vitamin C level.1
4. a. Graph and describe the relation of serum vitamin C with BMI as shown by a scatterplot.
- Spring '14
- Regression Analysis