Q: Can you help me find articles and books on my topic?
A: No, the DLS has not expertise with bibliographic
databases, only with statistics and data. For help with selecting and accessing
bibliographic databases, you should ask at the Reference Desk
of your favourite branch library.
Q: So when should I use the Data Library Service?
A: when you need to do statistical analysis, for example, to
test a hypothesis using existing data files or data bases, or
to replicate another researcher's statistical analysis, or simply require
statistics or data that are not available in printed form.
Similarly,
if you have collected data yourself, and have or are going to
publish research based on that data, the DLS can assist you in
making the data available to other researcher for secondary
analysis.
Alternatively, if you have received a SSHRCC grant for research involving
the collection of new data, the DLS can assist you to fulfill
the deposit requirement in the SSHRCC grant application. Similarly,
if you have collected original research data with MRC, NSWERC, or other
grant funding, and wish to have your data archived and made available
to other researchers, we can provide long-term archival management, as well as
dissemination of the data as per your direction.
See our web pages on data deposit and preservation.
Q: How do I find out what data files the Data Library has?
A:
Unfortunately, at present, you can only search for words in the title or the
name of the principal investigator(s). If you don't find any files of interest, please
contact us, as we may be able to recommend an alternative.
Q: How do I get the data files?
A: You have a number of alternatives:
Q:Can't I just get the data on a floppy disk?
A: Most data files in our collection are very large and complex, in the order of
hundreds of megabytes, and in some cases gigabytes in size. For most users,
especially those working on desktop workstations, doing
any kind of analysis with very large data files would be slow and
inefficient. It is much more efficient for you if we create
a subset with just those variables and cases that you want to analyse, in a format that your
software can easily read, thus allowing you to spend more of your time on analysis and less
on data management.
Q: But I just need a couple of figures to put in my report.
Can't I just get them printed out?
A:
Most of the statistics that we have that are ready to be used in this way, you can access from our
web page on:
Finding Canadian statistics or
Finding U.S. and international statistics.
If you can't find the statistics you need there, then in some cases, it may be simpler and faster to locate the 'couple
of figures' in a printed source, than to attempt to generate them
from the data files. Most of our data files are simply flat ascii files, with no retrieval software, and
extracting data from them requires that you have access to statistical software such as SAS, SPSS,
SHAZAM, Stata, etc., and write a program using that statistical software to generate the statistics
you need.
Q: Will you do my statistical analysis for me?
A: No, we are not able to provide statistical consultation. Your alternatives are:
Q: You don't have the data file I need. What should I do?
A: If the data are available from one of our normal data suppliers, such as
ICPSR, Statistics
Canada through the
Data Liberation Initiative,
the Roper Center, etc.), we will be able
to acquire it for you quickly, or help you download it yourself, at no cost.
Even if the data are not available from one of our suppliers, the Data Library Service
does have a budget for data acquisition. If the
data are likely to be of interest to other researchers or students, and if the DLS budget can
afford it, it is very possible that we may acquire it for you.
In either case, please contact Laine G.M. Ruus to
enquire about getting the data.
The more information you can provide about the data you need, the better.
Q: How do I read an RTF format file?
A: RTF is a text file format, and the acronym stands for 'Rich Text Format'. The format can be read by most major
word processors.
To read an RTF file in Netscape:
To read an RTF file in Internet Explorer:
Q: How do I read a Beyond 20/20 file?
A: Beyond 20/20 is a commercial data browser tool.
See our guide to getting and using the Beyond 20/20 software at:
<http://www.chass.utoronto.ca/datalib/caq/b2020.htm>.
Q: How do I read the output from the CHASS Census Analyzer Microdata file output into
SPSS?
A: This involves two steps: first, selecting and saving the appropriate output from
the Census Analyzer, and secondly, reading it correctly into SPSS.
Step 1: select the appropriate output formats.
Step 2: read the file in SPSS for Windows (these instructions are for version 9.0)
Alternative Step 2: read the file in SPSS for Unix (These instructions are for version 6.0)
Q: When I try to access the ___ database, it tells me I don't have permission.
A: Most databases acquired by the Data Library Service are restricted to access by University of Toronto
faculty, students, and staff, for academic research and teaching purposes only. Therefore, access to the data is restricted to University of Toronto IP addresses
only. This is a requirement of our contracts with our data suppliers.
If you are currently a University of Toronto faculty, student, or staff member, and are working off-campus, you will need to log in to your
UTORid before attempting to access restricted data.
Alternatively, prepend the following character string at the beginning of the URL you are attempting to access:
http://myaccess.library.utoronto.ca/login?url=
E.g. http://myaccess.library.utoronto.ca/login?url=http://estat.statcan.ca
This will take you via the myaccess login to the URL you have requested.
Q: Where do I get the Beyond 20/20 software?
A: If you are a University of Toronto, Ryerson University, or York University faculty, student, or staff member,
you can download the Beyond 20/20 software from:
<http://www.chass.utoronto.ca/datalib/caq/b2020.htm>.
We do have a University-wide license to use Beyond 20/20 for academic research and teaching purposes, under the provisions
of the Statistics Canada Data Liberation Initiative.
Q: Where do I get the Shazam software?
A: If you are a University of Toronto faculty, student, or staff member, you can get the Shazam software from the Economics Department,
which holds a University site licence for the software.
Contact: Ursula Gutenburg, the Economics Department Librarian, Office: 150 St. George Street, N107. Phone: 416-978-8623. Fax:
416-978-6713.
Email address: ecolib@chass.utoronto.ca. The cost is minimal ... about $10.
For users with CHASS accounts, Shazam is also available on the 'bebop' platform.
If you are not a University of Toronto faculty, student, or staff member, the Shazam home page is at:
<http://shazam.econ.ubc.ca/>.
Q: How do I use the CPI to convert dollar amounts?
A: See:
Q: How do I compute a fertility rate from aggregate data?
A: See:
See also:
Q: How do I compute an unemployment rate?
A: Statistics Canada computes the unemployment rate as follows:
unemployment rate = (unemployed labour force / total labour force) * 100
Q: How do I cite the data in my bibliography?
A: See:
Citing other kinds of material:
Q: How do I evaluate the quality of the statistics/data I found?
A: See:
Q: Do I have access to E-stat through UofT?
A:
Yes, as of approximately 2009, the E-stat web site is completely unrestricted. However, you must agree
to Statistics Canada's licence agreement in order to access the resources on the E-stat server. [2009-11-18]
Q: What's the difference between statistics and data?
A:
I use these two words as follows. Statistics are pre-processed numbers, often
called 'descriptive statistics' or 'aggregate statistics', that summarize the characteristics of a
group of observations, such as a population, phenomena (eg the weather,
stock prices), etc. Examples of descriptive statistics:
(1) the population of Toronto in 2001 was 4,682,900 [a count];
(2) the average cost of running a car in Alberta in 1996 was $6,041 [an average or mean];
(3) the cost of Internet access in Newfoundland rose by 2.7% between January 2003 and February 2004 [computed from an index, the CPI];
(4) the average temperature in Toronto in February 2008 [another average], etc.
Descriptive (aggregate) statistics often take the form of counts (totals), percentages, means (averages) and medians, rates, or indices. These are the statistics generally published in statistical publications of various kinds and media.
Descriptive statistics are usually appropriate answers to questions that ask 'how many/how much'.
Data, on the other hand, are the raw material from
which statistics are created, and more often used to answer questions that ask 'how' or 'why'.
Data are raw, unsummarized characteristics as originally collected.
The data (or microdata) from which the above examples of statistics were created are:
(1) the database of records (one per person) of all persons in Toronto on census day 2001;
(2) the actual expenditures on automobile maintenance by each household in Alberta that was interviewed in the Survey of family expenditures, 1996;
(3) the actual prices of the goods in the basket used to compute the CPI, in each month in Newfoundland;
(4) the temperature as recorded each hour of each day, at all environmental data stations in Toronto, each day in February 2008.
Data by themselves don't really mean much, and need to be
statistically manipulated usually by specialized software, in order to
generate descriptive or inferential statistics. The answer that I gave to
the 'age' question on my 2001 census questionnaire is a 'datum' and in and
of itself is only of interest to myself and the Canada
Pension Plan, but in conjunction with the combined economic assets, recreational habits, and general health of my age cohort,
become a research resource for planning retirement policy, senior support services in municipalities across Canada, new housing construction, or alternatively,
the abolition of compulsory retirement.
Inferential statistics on the other hand, are statistical measures of significance, of direction, and of magnitude. They are generally used to answer such questions as,
"If, based on the available data, I conclude that on average, the more education you have the more income you will have, what are my chances of having drawn an incorrect conclusion, ie what are my chances of being right or wrong?
And if I further conclude that each additional university degree adds approximately $10,000 to your annual income, how accurate an estimate is that?"
Q: How do I rebase an index to another index year?
A: Use the following formula to compute a ratio:
ratio = [new index at time1]/[old index at time1]
Then multiply all data points on the old index by this ratio.
Eg. A new index has base year 1995=100; the previous version of the index had
base year 1985. The index value for 1995 (on the previous version of the index)
was 127.5. Multiply values on the old index by 100/127.5=.7843 to convert them
to the new base year.
Q: What's the difference between the unemployment rate, the participation rate and the employment rate?
A: Statistics Canada computes these rates as follows:
unemployment rate = ((unemployed labour force / total labour force) * 100)
participation rate = ((labour force / total population 15 and over) * 100)
employment rate = ((employed labour force / total population 15 and over) * 100)
Q: What's the difference between nominal and real GDP?
A: Nominal GDP is GDP at current prices. Real GDP is GDP at
constant prices, in effect, ie adjusted by a price deflator such as the CPI.
Q: How do I read a hierarchical file into SPSS or Stata?
A: A useful outline of the process of working with hierarchical microdata files in SPSS and Stata is available in:
Q: I need to analyze change over time. Can you point me to some available data resources to get me started?
A: You will find links to resources on repeating cross-sectional and panel surveys at: <http://www.chass.utoronto.ca/datalib/major/long.htm>
Q: Where can I find some resources on 'statistical literacy'?
A: Some resources we have found useful are:
Q: What's the difference between public use microdata file (pumf) and the RDC versions of the 1996, 2001, and 2006 census of population microdata?
A: According to a dlilist message dated June 5, 2007:
The 2001 Census data will be available in
the RDCs at the end of June 2007; the 1996 and 2006 census data will be
available in the RDCs by December 2008. The census tract will be the
minimum level of geography available for urban areas and the census
subdivision is the lower level for rural areas. Researchers who wish to
access these detailed microdata for their research program need to submit
their proposals to SSHRC.
The RDC program has also provided us with the following overview of the
major differences between RDC file and the 2001 PUMF:
1. RDC file one file with four census universes (population, family,
household, dwelling) linked to permit analysis at all levels and
multilevel.
To request that additional files be made available via SDA, contact <laine.ruus@utoronto.ca>.
- if you are an undergraduate student, ask your professor or class TA,
- if you are a graduate student, faculty member or researcher, you may wish to consult the
Statistical Consulting Service offered
by the Department of Statistics.
From the menu bar select Options/General Preferences/Helpers
Find RTF in the list of file types
Click on and edit that file type to open
the file in a word processor of your choice (that can read RTF format, of course). If you need
to add RTF to the list of helper applications, 'description' is any text you wish, 'mime' is 'application/rtf', and
'suffix' is 'rtf'.
From the menu bar select View/Options/Programs
Find RTF in the list of file types
Click on and edit that file type to open
the file in a word processor of your choice (that can read RTF format, of course). If you need
to add RTF to the list of helper applications, 'description' is any text you wish, 'mime' is 'application/rtf', and
'suffix' is 'rtf'.
[rev. 2009-10-22]
See also:
Contains clear, concise definitions of all major demographic measures.
Working with survey files: using hierarchical data, matching files and pooling data Rafferty, Anthony & Jo Wathan, ESDS, March 2008.
Note that the focus of this document is on British data, not Canadian.
2. It is the full 2B file which is a 20% sample (6 080 919 records). The
file excludes institutional residents, residents of incompletely
enumerated Indian reserves or Indian settlements, and foreign residents,
namely foreign diplomats, members of the Armed Forces of another country
who are stationed in Canada, and residents of another country who are
visiting Canada temporarily.
3. It has geography at the CT or CSD level where CT is not available
(i.e. rural areas)
4. Provides data directly from the questionnaire and allows researchers
to build their own variables.
5. There is no capping or categorization of continuous variables.
6. Access to RDC file needs an approval proposal study. Census enhanced
confidentiality rules are applied.
Last updated: 2010-02-03