Stata 14 health stats work – finding percentages in raw data
Im doing a pooled crossectional study of the obesity and overweight rates on productivity within the uk.
Controlling for gender, age, ethnicity, education level and NS SEC level.
I am using the active people survey from 2012-2013, 2013-2014 and 2014-2015. And if it lets me will attach the stata files for all these years if not it is availbale on the essex data archive https://discover.ukdataservice.ac.uk/.
My issue is that my productivity data is in nuts3 regions whilst the active people survey data is in local authority data.
First thing i need doing is someone to get the percentages for all my control variables out of the raw data at the nuts3 level. I have attached a document converting the local authority labels from within the data to their nuts3 groups.
I need the percentages for each category of the variables below within each of the english nuts3 regions:
– Pos. = 48 Variable = d1 Variable label = D1 – RESPONDENT GENDER
This variable is numeric, the SPSS measurement level is NOMINAL
SPSS user missing values = -1.0 thru None
Value label information for d1
Value = 1.0 Label = Male
Value = 2.0 Label = Female
– Pos. = 55 Variable = D3_bands_3age Variable label = D3 – Age Bands (3) – 16-34, 35-54, 55+
This variable is numeric, the SPSS measurement level is NOMINAL
SPSS user missing values = -1.0 thru None
Value label information for D3_bands_3age
Value = 1.0 Label = 16-34
Value = 2.0 Label = 35-54
Value = 3.0 Label = 55+
– Pos. = 73 Variable = D4_bands_6 Variable label = ETHNIC GROUP (6)
This variable is numeric, the SPSS measurement level is NOMINAL
SPSS user missing values = -1.0 thru None
Value label information for D4_bands_6
Value = 1.0 Label = White
Value = 2.0 Label = Mixed
Value = 3.0 Label = Asian
Value = 4.0 Label = Black
Value = 5.0 Label = Other
Value = 6.0 Label = Chinese
– Pos. = 80 Variable = d6 Variable label = D6 – HIGHEST QUALIFICATION OBTAINED
This variable is numeric, the SPSS measurement level is NOMINAL
Value label information for d6
Value = 1.0 Label = HIGHER EDUCATION & DEGREE OR DEGREE EQUIVALENT
Value = 2.0 Label = OTHER HIGHER EDUCATION BELOW DEGREE LEVEL
Value = 3.0 Label = A LEVELS & EQUIVALENTS
Value = 4.0 Label = TRADE APPRENTICESHIPS
Value = 5.0 Label = GCSE/O LEVEL GRADE A*-C (5 OR MORE)
Value = 6.0 Label = GCSE/O LEVEL GRADE (LESS THAN 5 A*-C)
Value = 7.0 Label = OTHER QUALIFICATIONS
Value = 8.0 Label = NO QUALIFICATIONS
Value = 9.0 Label = REFUSED
Value = 10.0 Label = DON’T KNOW
Value = 11.0 Label = RESPONDENT QUITS INTERVIEW
– Pos. = 156 Variable = NS_SEC_bands_3 Variable label = NS SEC (3 bands)
This variable is numeric, the SPSS measurement level is NOMINAL
SPSS user missing values = -1.0 thru None
Value label information for NS_SEC_bands_3
Value = 1.0 Label = NS SEC1 to 4
Value = 2.0 Label = NS SEC 5 to 8
Value = 3.0 Label = NS SEC 9
– and the percentage weight (ie percentage of respondents) of each local authority region within each of the nuts3 regions
The Second thing is to use the local authority weight within each of the nuts3 regions to get a weighted average of the obesity and overweight rates (which is attached in precentages for each local authority region) for each english nuts3 region.
Thirdly i would like you to do a pooled crossectional regression of the nuts3 regions obesity and overweight percentages and all the control variables precentages on the unsmoothed GVA (excluding rental income) per hour worked (£) nuts3 productivity data (also attached) for the years 2013/2012-2013(for the obese,overweight and control vars), 2014/2013-2014 and 2015/2014-2015.
i would also like the whole process screenshotted and pasted on a word document throughout so i know its legitimate and im working with correct data.