A user needs variance estimates for the mean and median values of dwellings for a number of cities in Canada. Statistics Canada publishes these medians and means in tables from the NHS, but I cannot find their standard errors.
Here is an example for windsor:
http://www12.statcan.gc.ca/nhs-enm/2011/dp-pd/prof/details/page.cfm?Lang=E&Geo1=CSD&Code1=3537039&Data=Count&SearchText=Windsor&SearchType=Begins&SearchPR=01&A1=All&B1=All&GeoLevel=PR&GeoCode=3537039&TABID=1
Are these available anywhere, or would it be a special request?
Answer
Here is a new link that should provide you access to what we have available for NHS 2011 , I have tried it but please let me know if you experience any issues with it: <http://www12.statcan.gc.ca/nhs-enm/2011/dp-pd/dt-td/Ap-eng.cfm?LANG=E&APATH=3&DETAIL=0&DIM=0&FL=A&FREE=0&GC=0&GID=0&GK=0&GRP=0&PID=106639&PRID=0&PTYPE=105277&S=0&SHOWALL=0&SUB=0&Temporal=2013&THEME=98&VID=0&VNAMEE=&VNAMEF=>
From what the subject matter has provided regarding your questions the variance estimates for the mean and median values of dwellings for a number of cities would not be available at that level of geography. Only at the national, provincial and territorial geography levels. These would also not be available through special request due to confidentiality.
Below are some questions and answers regarding the 2011 NHS that were publicly released when the survey became available in case it is of any help to you or your researcher:
Q. Why is there no standard error (SE) of average income reported in the National Household Survey (NHS)?
A. The voluntary 2011 NHS replaced the mandatory census long form. As with most voluntary surveys, there are more risks associated with non-response in the NHS. In order to limit the effects of non-response, the systems and methodology normally used for the census long form were changed to introduce new methods at the collection, sampling and estimation stages.
The NHS is now a two-phase sample design, compared to the single-phase sample design in the census. This makes the estimation methodology more complex for the 2011 NHS. A two-phase sample design involves selecting a first-phase sample with data mostly collected via Internet and mail returns. The second phase consists in selecting and following up on a sub-sample of the non‑respondents from the first phase via more effective modes, i.e., telephone and in-person interviews.
Standard errors (SE) released in previous censuses were derived assuming a simple random sample without replacement. The same assumption cannot be made for the 2011 NHS, because of its more complex design. In the fall, Statistics Canada will be releasing coefficients of variation which are another measure of sampling variability, related to the SE, for key variables.
Q. Will Statistics Canada produce standard errors for NHS data at a later date?
A. Statistics Canada will produce coefficients of variation (CVs) at a later date. Standard errors can be calculated from CVs by multiplying the estimated figure and then dividing by 100. In the fall, there will be a special release of CVs at the national, provincial and territorial levels only for key variables.
Q. Can standard errors be calculated on a custom basis?
A. Standard errors will not be calculated for, nor provided with, custom tabulations.
Q. Can standard errors be calculated on a custom basis?
A. Standard errors will not be calculated for, nor provided with, custom tabulations.
