Wednesday, June 27, 2007

FED Profiles for 2003 Representation Order

Question

We have a professor interested in efficient access to Census 2001 profile data by FED for the 2003 representation order.

It is possible to find the data, riding by riding, for these 'adjusted' FED's on the Stats Can website:
http://www12.statcan.ca/english/census01/products/standard/fedprofile/index.cfm
but are these calculated profiles made available as a set, to the DLI?

Answer

The Census 2001 Profiles by FED for the 2003 representation order are available on the DLI FTP site in Beyond 20/20 format at:
/census/2001/B2020/Profiles/cumulative-electronic-profiles-b2020/can
filename: 95f0495xcb01011-can-ivt.zip

Monday, June 18, 2007

Query about CRIC

Question

The current government has dismantled the Centre for Research and Information on Canada (CRIC), which used to do regular public opinion surveys as well as collecting other kinds of data. It all used to be available to Canadians through their website (including the data sets), until just the last couple of days when their website disappeared as well. Do you know who might have received it?

Answer

Both CORA at Kingston and Carleton University have received some of the data files. I'm about to contact Bob Burge at Queen's to see if we can share what we have that they don't and vice versa. We have a mandate from the Canadian Unity Council to try to track down what we can.

You can get the metadata from the CRIC collection that we house if you go to
http://www.library.carleton.ca/ssdata/surveys/pop_cric.html

Click on Carleton University Data Centre and then CRIC. We don't have security in place at the individual file level to allow access at this time, but we will be moving the files to a more open server in the future.

Wednesday, June 6, 2007

Concordances for SOC to NOC

Question

We have a researcher who needs to convert the 1991 SOC codes for occupations in the NPHS to NOC codes. I have not been able to find any direct concordance from 1991 SOC to the NOC. Have I missed something?

My searches have led me to conclude that this will actually involve two steps:

1 -- convert from the 1991 SOC to the 1980 SOC via 12F0028XDE

2 -- then convert from the 1980 SOC to the NOC 12F0029XDE

If there is a more direct method for doing this, I would appreciate being introduced to it!

Answer

There's a concordance on the DLI ftp side, in
/standard_classifications/concordance

As far as I can tell, the socnoce.xls and nocsoce.xls files are NOC to/from the 1991 SOC, and the soc1980_noc.txt and noc_soc1980.txt files are to/from the 1980 SOC.

Also, a tool from Standard Division will help the conversion from NOC to NOC-S. The following URL:
http://www.statcan.ca/english/Subjects/Standard/soc/2001/nocs01-index.htm explains the difference between NOC and NOC-S.

The tool helps map NOC and NOC-S at the unit group level for 2006. This should save researchers some time, but note that some adjustment will be required when applying this to older data due to changes in NOC-S over time. (2006 does not add or delete any unit groups from the 2001 version, though there are some changes in unit group titles).

Question about Survey of Household Spending (SHS) Variables

Question

Is it safe to assume that a code of 0 in either variable PURPRICP or SELPRICE means that no house was purchased or sold in the year? If so, we can code 0 to designate a missing value (not applicable), and allow users to make meaningful estimates of house prices. As it stands right now, the average purchase price of a home is under $9,000, and the average selling price under $5,000.

Answer

Yes, it is safe to say no house was purchased or sold in the year.

Tuesday, June 5, 2007

New Product - Teacher's Guide to Data Discovery

Teacher's Guide to Data Discovery as released in today's Daily.
http://www.statcan.ca/english/freepub/12-593-XIE/12-593-XIE2007001.htm
Teachers are finding that basic statistical skills “reading and creating tables, constructing graphs and calculating measures of central tendency (mean, median and mode)" are required across the curricula. Students practise these skills not only in the traditionally number-based subjects like mathematics and economics, but also in history, geography and other subjects.

This guide was written to support both elementary and secondary teachers in helping students develop basic statistical skills. It provides teachers with specific instructions on:

- Finding interesting and grade-appropriate Canadian datasets
- Choosing appropriate graphs for different kinds of data
- Calculating basic statistical measures, with or without statistical software

Very positive feedback already from the education community with the Ontario Ministry of Education including a version on 1,000 teacher training CD-ROMs and linking to it from their OCUP website

Monday, June 4, 2007

LICO by CSD

Question

A researcher is hoping to prepare maps for the incidence of low income by CSD from 1980 forward. She has found that the 1996 and 2001 Censuses publish the incidence rate of low-income at the census division level. Is this available for previous to 1996?

Answer

There is certainly 'Incidence of low income' in the 1991 profile files at the CSD level - it's right down at the bottom of the list of characteristics in the 2B file, immediately before the household income categories. Ditto 1986. And it seems to be available in the 1981 census tract level profile file, so the prospect is good it's in the CSD-level file too. For LICO prior to 1981, it would have to be obtained via a custom tabulation.

Friday, June 1, 2007

NYSE Trade & Quote Database

Question

A faculty member here wishes the Library to subscribe to the New York Stock Exchange Trade and Quote database; I've never heard of it before. I gather this is something that not very many universities have and that most often it is the business school that pays for it.

Any comments about this that I can share with the Collections Development Unit?

Answer

The NYSE Trade and Quote data is available directly from the NYSE on DVD or via the WRDS service (must pay for the WRDS membership).

They do offer an academic discount. You may buy the data in monthly increments or by the year for historical data.

Your collections folks (or your business librarian, if you have one) need to contact the NYSE for a detailed quote.

The contact information can be found on their website.

We looked at subscribing to TAQ but because you have to pay an annual rate for the current year plus purchase, for a flat fee, each historical year (goes back to 1993) we decided we couldn't afford a long enough backfile to allow for data analysis. If you want to learn more you can visit this website.