Tom,
I agree that there is a need to study why there is often a discrepancy between the percentage of people who state that they have difficulty with literacy tasks vs. the percentage of people who score poorly on literacy tests. In addition to the hypothesis that you share: "some have suggested that adults may be too embarrassed to admit that they have a literacy or numeracy problem and that is why there is a
large discrepancy between the adults' test scores and their self assessments of literacy", I would like to propose two other hypotheses worth testing.
1. Is it possible that some/many struggling adult readers are okay with their level of literacy performance? Perhaps they have jobs/community support that meet their literacy abilities and needs. 
2. Is it possible that some/many struggling adult readers honestly don't know what they don't know? For example, it is not uncommon for an adult reading at the second grade level, to enroll in an adult literacy program with the expectation that within a year of weekly 2-4 hour classes s/he will be able to pass the GED exam. I think that it is human nature not to fully understand what we don't know. For example, I know that I don't know physics, but I have no clue how much I don't know compared to someone who has studied it for one semester in college. On a test asking me to self-assess my physics knowledge, I too may overestimate what I know in relation to the world of physics.
Daphne


>>> <tsticht at znet.com> 6/8/2008 5:45 PM >>>

June 8, 2008

The Great Adult Literacy Skills Debate: Tests vs. Self Assessments

Tom Sticht
International Consultant in Adult Education

In 2006, Inge Henningsen of the Department of Statistics in the University
of Copenhagen presented a paper entitled: "Adults just don't know how
stupid they are: Dubious statistics in studies of adult literacy and
numeracy."
(online at www.alm online.org/ALM13/programma%20alm13.pdf ).

In this paper Henningsen comments on the many problems, conceptual,
methodological, and statistical, with the International Adult Literacy
Survey (IALS) of the mid-1990s and the Adult Literacy and Lifeskills (ALL)
survey of 2003-06. One of the major factors in these assessments that
Henningsen addresses is the finding in various nations of a wide gap
between the literacy and numeracy
skills of adults when the literacy test scores are taken as indicators of
people's skill, and the skills that adults assign to themselves when asked
to self-assess their literacy or numeracy skills.

Though Henningsen focuses primarily on the data from these international
adult literacy surveys for Denmark, similar gaps are found in various other
nations. For instance, in Australia, based on the IALS test scores the
report writers declared some 46 percent of adults to possess low literacy
skills, whereas only 4 percent of the adults themselves thought they had
low skills.

In Canada and the United States, similar discrepancies were found, with 42
percent of Canadians and 47 percent of U.S. adults being declared low in
literacy based on the test scores, while only 5 percent of Canadian and 7
percent of U.S. adults rated their literacy low.

In New Zealand, using just the numeracy data from the ALL, 51 percent of
adults were declared low in numeracy based on test scores, while only 19
percent rated their numeracy skills as low. Finally, in England, using a
special test developed for the Skills for Life strategy in that nation, 16
percent of adults were declared low in literacy based on their test scores
but only 4 percent thought they had low test scores.

Henningsen  noted that the discrepancy between the self assessed proficiency
and the conclusions based on test scores is not treated seriously in the
reports and asks, "Is it ethically defensible to disregard the opinions and
statements of the adults regarding their own skills and "narrate" big groups
of adults in the labour market as excluded from society and lacking in basic
skills." Answering this rhetorical question, Henningsen goes on to say, "I
find it disturbing that the reports send the message that the experiences
and assessments of the test persons themselves have no validity compared to
the test results. Is it a viable for the adult education community to let
surveys convey the impression that "adults just don't know how stupid they
are."?

One important consequence of adults' thinking that their literacy and
numeracy skills are pretty good is that they will choose to not participate
in language, literacy, and numeracy (LLN) provision to improve their skills.
For instance, from various sources I can make rough estimates of the
percentage of adults that the government says are in need of LLN provision
that actually take part in LLN provision in a given year. In  Australia the
percentage of those the governments say are in need of LLN provision who
actually enroll in LLN provision in a given year is around 4%, in Canada
10%, England 5%, New Zealand 11%, and the U.S. 3%. These (admittedly
roughly estimated) percentages of participation are more in line with the
self assessed needs of adults than the needs based on the paper and pencil
tests.

In the United States, the National Center for Education Statistics reports
in the 2008 Conditions of Education that the percentages of adults aged 16
or older who participated in adult education activities consisting of basic
skills, English as a second language, or apprenticeships in 1995, 1999, 2001
and 2005 were 3, 4, 4, and 3 percent respectively. Without the category of
"apprenticeships" the percentages would be even lower. These low
percentages of self reported participation in LLN are again more in line
with the self assessments of adults regarding their literacy and numeracy
skills than with the percentages declared to be "at risk" for low literacy
based on the adult literacy survey tests.

The large discrepancies between the percentages of adults needing basic
skills education as given by governments based on the international adult
literacy surveys, and the much smaller percentages of adults who perceive
their literacy and numeracy skills to be so low that they are unable to
progress in the societies in which they live pose problems for adult
education. Some have suggested that adults may be too embarrassed to admit
that they have a literacy or numeracy problem and that is why there is a
large discrepancy between the adults' test scores and their self
assessments of literacy. If this is so, then research is needed to
establish that this is so. In general, major efforts are needed to better
understand the genuine needs of adults for LLN provision, what sorts of
educational programs would best meet these needs,  and the sorts of
activities that are needed to let
adults understand the educational opportunities available to them.


Thomas G. Sticht
International Consultant in Adult Education
2062 Valley View Blvd.
El Cajon, CA 92019-2059
Tel/fax: (619)444-9595,
Email tsticht at aznet.net 




-------------------------------
National Institute for Literacy
Assessment mailing list
Assessment at nifl.gov 
To unsubscribe or change your subscription settings, please go to http://www.nifl.gov/mailman/listinfo/assessment 
Email delivered to alcdgg at langate.gsu.edu