Given the two National Health Interview Survey datasets, the data dictionaries, and the scoring

document, achieve the following:

$1.$ Clean the data in R:

a$.$ In the child dataset, identify and print the duplicate records with the same ID value.

Eliminate the duplicate record for each instance so there is only one of each ID$.$

b$.$ In the child dataset, identify and print records with values that are out of range for the

following variables $($consult data dictionary$).$ Determine how best to deal with these

values. Document your decisions in the dataset cover sheet $($see #$3$ below$).$

i$.$ BSCNWPPL$\_$C

ii$.$ BSCNWPLCS$\_$C

iii. BSCCHG$\_$C

iv$.$ BSCHLOPPL$\_$C

v$.$ BSCCRYALT$\_$C

vi$.$ BSCCLMDWN$\_$C

vii. BSCFUSSY$\_$C

viii. BSCSTHE$\_$C

ix$.$ BSCSCHD$\_$C

x$.$ BSCPTSLP$\_$C

xi$.$ BSCSTYSLP$\_$C

xii. BSCPRLKSL$\_$C

c$.$ In the child dataset, convert the above listed variables into factors and define the levels

according to the data dictionary.

d$.$ In the child dataset:

i$.$ Create a new variable AGECAT grouping the age of the child $($AGEP$\_$C$)$ into the

following categories:

$1.0-7\mathrm{.}99$

$2.8-12\mathrm{.}99$

$3.13-18$

ii$.$ Generate frequencies for the new age groups

iii. Generate a list of all records who have either had borderline diabetes or

prediabetes. Save these records to a new dataset called DIABETES.

$2.$ Generate a permanent cross$-$sectional analysis data set.

a$.$ Create a new variable called HHX$\_$NUM that is the HHX variable value without the H$0$ at

the beginning $($contains only the last $5$ numbers of the variable$).$

b$.$ Generate frequencies for sex, age, Hispanic ethnicity, and general health status.