Thursday, July 29, 2004

Products that fall through the cracks

Question

Late yesterday afternoon I received a phone call from a professor in Recreational Administration who wanted access to the average household spending on recreational items in CMAs. He had some prior assistance from the Edmonton STC Regional Office and our Reference Desk, which had identified for him a catalogue number for a table that contained just what he wanted: 62F0031. He had been told that this didn't not appear within the DSP and should be a DLI product... hence his phone call.

I went directly to the Online Catalogue of Products and Services and, even though the catalogue number that he provided seemed a bit short, searched for 62F0031 within Catalogue #. No results.

I repeated the search for 62F0031 within a page. No results.

Next I conducted a site search for 62F0031 and was presented with one result: the Labour Market and Income Data Guide
http://www.statcan.ca/english/freepub/75F0010XIE/3504Ae.htm

A search for 62F0031 on this page discovered the following information:

Products and services:
1. Spending Patterns in Canada (Catalogue no 62-202-XPB or 62-202-XIB)
2. Standard data tables (62F0031, 62F0032, 62F0033, 62F0034, 62F0035, 62F0041, 62F0042, 62F0043, 62F0044, 62F0045)
3. Public Use Microdata File (62M0004XCB)
4. Custom tabulations

The term "Standard data tables" is close to "standard data products".

My questions are:

Are standard data tables a subcategory of standard data products? If they are, can we obtain these tables for distribution within DLI?

I checked CANSIM to see if there was any connection between these standard data tables and any of the series in CANSIM. While Table 203-0010 contains household spending on recreation by province and territory, there is no table for CMAs.

Looking on the STC website under Canadian Statistics / Families, housholds and housing / Expenditures, there is a table for average household expenditures for seleced metropolitan areas. This table unfortunately has no table number or catalogue number by which to reference it. Furthermore, the patron wanted more recreational items than it contains. There are over 60 items identified for recreation in CANSIM 203-0010.

How do I know that the patron had the right standard data table number for CMA?

As a last hope search, I checked the Statistics Canada downloadable publications on the DSP site for Spending Patterns in Canada. In the 2002 edition of this pub, a search for 62F0031 found an appendix with the title, Related Products and Services containing the table number. All of the standard data table numbers listed above were identified in the appendix WITH titles. 62F0031 has the title, Detailed average household expenditure for Canada, provinces and selected metropolitan areas. It appears that I replicated someone's earlier work for this patron but was similarly deadended.

This brings me back to the question, can we get these tables through DLI?

Answer

What we have sometimes found out is that even though the name "standard tables" is given to s product or set of products they are in fact not "standard". That is they are not "off the shelf" tables ready to be disseminated but still require manual intervention based on the clients

Tony Moren here in the Library did some further checking and found this
http://www.statcan.ca:8096/bsolc/english/bsolc?catno=62F0031XDB

The product number you mentioned 61F0031 is actually 62F0031XDB. Looks like these are tables they send out on diskette. We will ask for this product as it is in neither DLI or DSP.

Wednesday, July 28, 2004

Definition of Income Class

Question

A researcher at McMaster is looking for definition on what constitutes a lower, middle and higher income class in Canada. She needs the income range of families belonging to each class (lower, middle and upper). I have looked through Census guides and many other publications but could not come across anything that she wants.

Answer

The classification of income into low, middle and high, is a very difficult concept. What is a high income in one area of the country (say an annual income of $100,000) could be considered as middle or low in another area of the country or province. Setting aside actual income dollars, even the "median" income of people in one part of the country could be considered as "high" or "low" in other parts of the country.

Similarly there is no standard as to what is poverty or the poverty line. Again a number of factors have to be taken into account when trying to identify individuals and families who are "below the poverty line". For these reasons Statistics Canada instead publishes something called the "low income cut-offs. to help users in identifying people who are in financially difficult situations. As stated in
http://www.statcan.ca/english/research/13F0027XIE/13F0027XIE.htm

"For many years, Statistics Canada has published a set of measures called the low income cut-offs. We regularly and consistently emphasize that these are quite different from measures of poverty. They reflect a well-defined methodology which identifies those who are substantially worse off than the average. Of course, being significantly worse off than the average does not necessarily mean that one is poor."

A 1994 article in the US based "Monthly Labor Review" states that: "Intuitively, income is a natural choice for a classifying variable because it is an indicator of consumers' financial ability to purchase goods and services and therefore is assumed to be a measure of their economic well-being. However, there are theoretical and practical drawbacks pertaining to income that make alternative measures more attractive, at least for some applications."

Also an article in the Winter 2002 issue of the STC publication "Perspectives on Labour and Income" states that; because the same income can affects different families in different ways " ...to make comparisons between different family types, it is necessary to look at both income pooling and economies of scale within families. An equivalency scale is the device most commonly used to level the playing field."

In any case, if you can not find an acceptable definition or classification of high, middle and low income it probably because there is no universally acceptable standard. Income Division has told us, for example, that "in SLID there are no definitions of lower, middle and upper incomes." In fact what they suggest is to use quintiles to divide the population. They also suggest that the researcher define their own groups using previous employed/published groupings, either their own or those of someone else.

Monday, July 26, 2004

Food Expenditure Survey

Question

We have a grad student who has been using the Food Expenditure Survey for expenditure at restaurants by country of birth. This variable has been dropped for the 2001 survey. Is there any other data source for this type of information? He has collected data to 1996 and now cannot find more current data.

Answer

Income division (responsible for Food Expenditures) confirmed that you are correct, there is nothing after 1996 on this and they are not aware of any other resources to be able to assist you.

Sunday, July 25, 2004

Economic families vs. Census families

Question

A student who wants to use the SLID2000 data has asked me about the differences between 'census families' and 'economic families'. As I read the definitions, I do not seem to find a major distinction, yet they are maintained as two separate files, so there must be a reason.

As I interpret the definitions, it seems that most households would be both 'economic families' and 'census families'. Only cases of siblings residing together would seem to be outside the definition of a census family.

Somehow, this does not seem to be enough to justify having two separate files -- what am I missing?

Answers

1) A census family consists of a couple and their children in a home. An economic family extends this definition to include all relatives (by blood or adoption) also living in the home. An example of this would be in the situation with grandparents living
with a couple and their children. All of them together consist of the economic family whereas using the census definition you would actually have two families; the grandparents are one and the couple with the children would be the second.

2) Many definitions (plus questionnaires, plus concepts, plus classification codes; plus survey descriptions, plus more) can be found at: http://www.statcan.ca/english/concepts/index.htm

They have the following definitions for Census and Economic family. As you will see in a Census Family only kids under 25 and without their own spouse or child are counted, but in an Economic Family there is no age restriction, they just have to be all related (so could include grandparents as well!)

Census family is defined as a now-married couple, a common-law couple or a lone-parent with a child or youth who is under the age of 25 and who does not have his or her own spouse or child living in the household. Now-married couples and common-law couples may or may not have such children and youth living with them. Now-married couples and common-law couples are classified as husband-wife families and the partners in the couple are classified as spouses. For a more detailed definition go to:
http://www.statcan.ca/english/concepts/definitions/cen-family.htm

Economic family is defined as a group of two or more persons who live in the same dwelling and are related to each other by blood, marriage, common-law or adoption. By definition, all persons who are members of a census family are also members of an economic family. Examples of the broader concept of economic family include the following: two co-resident census families who are related to one another are considered one economic family and two co-resident siblings who are not members of a census family are consider an economic family. For a more detailed definition go to Economic Family
http://www.statcan.ca/english/concepts/definitions/eco-family.htm

3) The economic family concept requires only that family members be related by blood, marriage, common-law*** or adoption. By contrast, the census family concept requires that family members be either a male or female spouse, a male or female common-law partner, a male or female lone parent, or a child with a parent present. The concept of economic family may therefore refer to a larger group of persons than does the census family concept. For example, a widowed mother living with her married son and daughter-in-law would be treated as a non-family person under the definition of a census family. That same
person would, however, be counted as a member of an economic family along with her son and daughter-in-law. Two or more related census families living together also constitute one economic family as, for example, a man and his wife living with their married son and daughter-in-law. Two or more brothers or sisters living together, apart from their parents, will form an economic family, but not a census family, since they do not meet the requirements for the latter. All census family persons are economic family persons.

Input-Output Tables

Question

What is an input-output table?

Answer

There is a basic lecture about input output tables at the following URL.
http://www.sonoma.edu/users/b/benito/econ403/units/

Also at http://www.scotland.gov.uk/about/FCSD/OCEA/00014713/index.aspx

Canada has been one of the world leaders in the building of these tables. For instance, we developed a rectangular table as opposed to the traditional square ones. Kishori Lal has published some articles on the evolution of the Canadian tables but you need the above primers to undertand what he is talking about. See
http://www.statcan.ca/english/research/13F0031MIE/13F0031MIE2001009.pdf

Basically, an I/O table allows the researcher to do 'what if analysis'.

They can be constructed at various different levels and for different sectors of the economy.

Information on the interprovincial IO tables is at:
http://www.statcan.ca:8096/bsolc/english/bsolc?catno=15F0009X

Also see the publication:
15F0077GIE - A Guide to Deflating the Input-Output Accounts: Sources and methods
http://www.statcan.ca/english/freepub/15F0077GIE/free.htm

Monday, July 19, 2004

African Professors in Canadian Universities

Question

A professor in sociology is presently conducting research on Africans scholars in Canadian universities. He is looking for recent and relevant data on African Professors in Canadian universities.

Answer

The Education Division has something called the "University and college academic staff system" which should have the information you need.
http://www.statcan.ca:8096/bsolc/english/bsolc?catno=81C0015

Please contact them directly because the data is only available by special request.

Friday, July 9, 2004

Citing maps, aerial photographs, satellite images, etc.

Question

I was wondering if anybody had any helpful hints on the following topic? Mount Royal College is attempting to use APA citation format to reference maps and graphs created by Stats Can raw data, and aerial photographs, and satellite images. We are also trying to cite maps made from data collected by the author (or person who created the data). Does anyone use this type of citation style to reference spatial data?

Response

Work is currently being done on a StatCan citation guide. The plan is to build bibliographic references examples from their standards (public) products including maps and geospatial data files. In doing the preliminary work (inventory), I found a good Canadian web site at McMaster on the subject:
http://library.mcmaster.ca/maps/mapcite.htm

Other good examples include:
http://library.owu.edu/citing222.html

http://137.122.27.99/map/guides/citation-e.pdf

http://www.brocku.ca/maplibrary/howtoref.htm

http://www.lib.uwaterloo.ca/locations/umd/digital/citation.html

More, the classic on the subject of maps citation seems to be:
Clark, Suzanne M., Mary Lynette Larsgaard, and Cynthia M. Teague.
Cartographic Citations: A Style Guide, MAGERT Circular No. 1.
Chicago: American Library Association, 1992.