I would like to compare some WEMWBS (Warwick Edinburgh Mental Well-being Scale) data with that reported by the NHS (https://digital.nhs.uk/pubs/hse2016). The NHS data only reports the mean value, sem and q10, q50 and q90 with no sample size so I do not think that I can do a formal statistical comparison. I have tried to show the comparisons graphically for the various age cohorts where the NHS data is the solid line.
Code:
* Example generated by -dataex-. For more info, type help dataex
clear
input double(y sex agegrp stat) str3 statlab double source
 50.2 1 1.1 1 "am"  1
   51 1   1 2 "q50" 1
   60 1   1 3 "q90" 1
   41 1   1 4 "q10" 1
 49.9 1 2.1 1 "am"  1
   51 1   2 2 "q50" 1
   59 1   2 3 "q90" 1
   39 1   2 4 "q10" 1
 50.2 1 3.1 1 "am"  1
   51 1   3 2 "q50" 1
   60 1   3 3 "q90" 1
   39 1   3 4 "q10" 1
47.72 1 1.4 1 "am"  2
   49 1 1.3 2 "q50" 2
   57 1 1.3 3 "q90" 2
   37 1 1.3 4 "q10" 2
 49.5 1 2.4 1 "am"  2
   50 1 2.3 2 "q50" 2
   53 1 2.3 3 "q90" 2
   45 1 2.3 4 "q10" 2
 33.5 1 3.4 1 "am"  2
 33.5 1 3.3 2 "q50" 2
   35 1 3.3 3 "q90" 2
   32 1 3.3 4 "q10" 2
end

twoway (scatter y agegrp if sex ==1, mlabel(statlab) msymbol(none) mlabp(0))    ///
(line y agegrp if stat != 1 & agegrp == 1 & sex == 1)    ///
(line y agegrp if stat != 1 & agegrp == 2 & sex == 1)     ///
(line y agegrp if stat != 1 & agegrp == 3 & sex == 1)    ///
, yti(WEMWB) sch(s1color) legend(off) xti("") ti("Male")

gr_edit  .xaxis1.reset_rule , tickset(major) ruletype(none) 
gr_edit  .xaxis1.add_ticks 1.2 `"16-24"', tickset(major)
gr_edit  .xaxis1.add_ticks 2.2 `"25-34"', tickset(major)
gr_edit  .xaxis1.add_ticks 3.2 `"35-44"', tickset(major)

Array

Although the plot shows the relationship I was wondering if there was a more appropriate plot.

Thank you.

Martyn