Thursday, April 28, 2005

DLI Access for Visiting Professor


I have received a request for Census 2001 individuals PUMF from a professor from Economics Department at the University of Winnipeg who is visiting McMaster University for the academic year (September 2004-August 2005). The data is going to be used for research purposes only. He normally access the DLI material through the "Landru" portal of the University of Winnipeg but because of IP restrictions he has not been able to access the data from McMaster. Is this a legitimate request? Is it ok if I give him access to the Census data?


Yes, visiting professors are considered to be members of the academic community in which they are working. His request is legitimate as long as the use is allowed under the DLI.

Wednesday, April 27, 2005

Time Series Statistics on Depression, Stress, Anxiety


A patron wants time series on this topic going back roughly 30 years. The NPHS, CCHS and GSS don't provide such a long time stretch, and neither does CANSIM. Any suggestions?


1) Others might have a better idea, but we've used the hospitalization statistics in 83-245 for depression and affective disorders. The data are not inclusive by any means, but offer a bit of a stable indicator (okay, that might be a stretch). The data are continued on CIHI's site under hospitalizations as an interactive table, if memory serves.

2) The other alternative is to go with the microdata. The Canada health survey 1978/79 (available from DLI) contains standard depression indicators, as does the Social Change in Canada survey, 1977-1979-1981 (available from ISR/York), which would take you back to almost 30 years ago.

Monday, April 25, 2005

Tracking Cited Survey Data


A student here is working on a literature review and wants to compile a complete-as-possible list of published academic studies using data from Canada's Alcohol and Other Drugs Survey (1994). Is there any tracking done of research projects that use STC survey data? Perhaps through the RDCs? And if so, is this information accessible/public?


Upon initial discussions with Sage, she confirmed that Statistics Canada does not keep track of all literature produced outside of HQ using our data sets.

Sage did however wish to relate to you an excellent article written by Elizabeth Hamilton and Chuck Humphrey titled "DLI and NPHS: a Study in Compatibility". You can retrieve this article from the DLI web site by following these simple steps:

1) Visit our website (
2) Select "Reports and Reference Materials"
3) Click on "DLI Update"
4) Download Volume 5, issue 2 (Fall 2002).
** The Background section of the paper tells you how they found evidence of secondary data analysis

Income Status in 1981


I am looking at the definition of income status in the 2001 census dicitonary:

and I am puzzled.

The detailed definition states: "Refers to the position of an economic family or an unattached individual 15 years of age and over in relation to Statistics Canada's low income cut-offs (LICOs)," followed by "Censuses: 2001 (1/5 sample), 1996 (1/5 sample), 1991 (1/5 sample), 1986 (1/5 sample), 1981 (1/5 sample), 1971 (1/3 sample)".

The latter implies that this variable or concept is covered in every census since 1971; however, examination of the PUMF codebooks reveals coverage only since 1991. A researcher here is looking at the relationship between immigrant status, age and
position above or below the LICO line. The above definition implies that he should be able to use the 1981 census, but the concept does not appear to be present there. Can someone explain this to me? Was it in the master file, but left out of the PUMF?


Low income cut-off variable is not included in the 1981 PUMF, however it does exists on the source database. It is important to remember that the content of the PUMF is a selection of variables only.

LICO is a derived variable based on a number of non-census resources including FAMEX/SHS and CPI, see:

Using the incomes listed on pages 22 and in that document, your user should be able to code his/her own LICO in those pumfs which are currently missing such a variable - using either the 1986 or 1992 base years specified. Alternatively, the formulae are given from which one could derive LICOs for earlier base years working from earlier versions of FAMEX. It would be a bit of work, but as far as I can tell, it could be done.