IALSS 2003: Update from Statistics Canada
The International Adult Literacy Survey (IALS) is
part of a continuing tradition of attempts to measure literacy levels in the adult population by means of surveys and to
produce international comparisons. Such research is
driven by the search for universals in the relationship
between literacy, education and prosperity, which can be
used to further the goal of global development.
The IALS draws on a particular
discipline – the
psychometric measurement tradition. It uses an
information processing model of literacy and attempts to
identify levels of literacy skill that are independent of the
context of use – the literacy counterpart of the generic
and transferable labour skills supposedly possessed by
the flexible worker...
The IALS Surveys... are re-defining literacy to fit in with
the projected needs of an ideal, consumer-oriented citizen
who is responsive to multiple new contexts for literacy
use. They justify a vision of what literacy should be, rather
than being based on people’s lived experiences. This is an
institutional vision that has little to do with supporting
people to use and control literacy for their own purposes.
It privileges some literacies and deletes other, vernacular
practices and then presents its findings as the ‘truth’
excerpt from “Privileged Literacies: Policy, Institutional
Process and the Life
of the IALS” by Mary Hamilton, Language and Education, 2001, Vol. 15, Nos.
The second round of the IALS survey happened
earlier this year. We interviewed Jean Pignal, Chief
of the Literacy Section at the Special Surveys Division
of Statistics Canada, by e-mail. Below are his
responses. Jean will also be joining us on the listserv
discussion (see page 11) to address questions about
how the IALSS works.
We understand that the report from the new
IALS – ALL – is due out in early 2004. What will
The name of the Canadian survey is the
International Adult Literacy and Skills Survey
(IALSS). While this is known internationally as the
Adult Literacy and Lifeskills Survey (ALL), Canada
preferred to keep the IALS cachet and recognition.
The first scheduled report will examine the measured
skill domains and their distribution across the
participating countries. Very little sub-national
analysis will be included in this report. Its flavour
will very much be like the 1994 publication Literacy,
Economy and Society: Results of the first International
Adult Literacy Survey.
The results will be ready by the spring, but the
first international report won’t be ready until the fall
or winter of 2004. National results will be available
in September 2005, as will a series of provincial,
territorial and sub-population reports. We should also
have a public use Microdata file ready by then.
Another update for the IALS
will be an interactive web site. It allows people to develop their
own tables using the IALS data.
I have looked at the Educational Testing Service
(ETS) web site (www.ets.org/all/)
and read the frameworks. I understand that the ALL will
measure teamwork and analytical problem-solving. Why did you include these
After piloting the instruments
in 2001/02, we found that the teamwork framework was not
successful as a psychometric measure, nor could we
use it to produce valid and reliable scales. We did
manage to produce a problem solving test which
measured analytical reasoning through a series of
scenarios. The reasons for expanding our focus was
simple – while literacy and numeracy are basic skills
that have impacts on individual well-being and
socio-economic success, other skills must also be considered if
we are to better understand the impact of skills on our labour market
and social fabric. The
development of these test and measures can only be
related by following a long road. In 1998, we began
to develop and conceive of the theoretical
frameworks and the framework and items were
finally chosen in 2002 – four years of trial and error,
but the domains that have been measured will
undoubtedly add to our understanding of skills
distribution in Canada.
IALS Methodology and Validity
In the IALS, the performance scales and the self-assessments represent
two fundamentally different approaches to assessing literacy
abilities. In the
performance assessments, literacy is construed as acognitive ability (latent
trait) that makes possible the use of printed materials in various
contexts. It is
considered that some people have more of this capacity
than others, although how much people have or lack
may not be consciously apparent to them. Nonetheless,
it is assumed that these differences in the amount of
capacity can be inferred using people’s performance on
various real-world tasks that incorporate the latent trait
that is theorized to make possible each person’s
In the self-assessment approach to assessing literacy,
literacy is considered as an ability or set of abilities (as in
reading, writing, and numeracy in the IALS) that adults
are consciously aware of and can perceive well enough to
estimate how well their literacy skills permit them to
n e g o t i a te the literacy demands of different sets of
activities at work or in their daily life. This requires that
adults are aware both of the demands for literacy in the
different contexts that they encounter and of how well
their literacy abilities permit them to meet these demands
on a recurrent basis.
Clearly, these two different approaches to assessing
literacy are based on different implicit theories about
literacy and different procedures for measuring literacy. It
is also evident from the discrepancies in data that these
approaches produce different estimates of how many
adults are at risk because of literacy in the various
nations that participated in the IALS. These findings raise
serious questions about the validity of the different
assessments. Is each assessment equally valid as a
means of representing the literacy abilities of the adult
population? If so, then how should the different results of
each method be used?
reprinted with permission from "The International Adult
How well does it represent the literacy abilities of adults?" by Thomas G.
Sticht, The Canadian Journal for the Study of Adult Education, Vol. 15,
November, 2001, pp19-36.
I have also heard that, in Canada, the ALL will
look at some specific communities more in
depth or separately from the general
The ALL framework requires each country to field a
sample of respondents aged sixteen and sixty-five. In
Canada, the IALSS obtains this sample and
supplements it with a sample of individuals over
sixty-five (seniors). Moreover, Aboriginal respondents
living off reserve were sampled in urban areas of
Manitoba and Saskatchewan, as well as the three
territories. We also had an augmented sample for
immigrants (recent and established), linguistic
minorities (in New Brunswick, Quebec, Ontario and
Manitoba) and youth (British Columbia and Quebec).
Some discussions have taken
place regarding on-reserve aboriginal and in-prison samples (to the point
where estimates have been made and reports written
on approximate costs and methodologies for these two
sub-populations), but funding has yet to materialize.
If so, what are those communities
and how/why were they chosen?
The IALSS samples were taken from the Census.
We randomly chose households with a high
probability of containing a respondent with the
desired demographics (i.e. sixteen and older for the
base sample, or with the added constraint of having
to be of aboriginal descent, or of a linguistic minority
(French in Manitoba, Ontario and New Brunswick,
English in Quebec), or screened on some other
demographic criteria). Thus we had a sample of
randomly chosen households that could then be
screened at the door by the interviewer to see if
anyone in the household fulfilled the conditions
imposed for inclusion in the required IALSS samples.
From this list of eligible residents, a person was then
randomly chosen as a respondent. We currently have
over 23,000 respondents from across Canada with
sufficient numbers in every targeted sub-population to
produce reliable estimates of proficiency in the four
measured Domains (Prose Literacy, Document Literacy,
Numeracy and Problem Solving).
Why did Canada choose to look at seniors and
While adult literacy may be a factor in the
workplace, it is also a quality of life issue. In order
to provide proficiency estimates for the entire
Canadian adult population, we needed to
supplement the sample with older Canadians. This
sample will also be comparable with the 1994 IALS
allowing us, in a limited manner, to track change in
the profiles across Canada.
Why do certain provinces have different
As with our Federal partners, each Province and
Territory was given the opportunity to enrich their
sample. This offer was taken up by every province
except New Brunswick (which nevertless received
an augmented sample of Francophones through the
federally funded linguistic minority sample) and
Prince Edward Island (which was allocated 650
responses in the base sample). In addition, all three
territorial governments funded a northern sample.