the study of
statistics used in
everyday arguments
| Statistical Literacy: the study of statistics used in everyday arguments |
2005
|
|
|
|
Statistical Literacy studies statistics used as evidence in everyday arguments. Statistical literacy is closely related to Numeracy, Quantitative Literacy and Statistics. Stat Lit focuses more on observational studies than experiments, more on two-group comparisons than correlation, more on context and multivariate thinking than chance, and more on ordinary-language than on algebraic-formulas. Although "association doesn't prove causation" in observational studies, the decision-related question is how much support does a statistical association give for inferring a direct causal connection. This site focuses on that evaluation. "Statistical literacy, the ability to follow and understand arguments from data..." For All Practical Purposes. COMAP "What everyone needs [from statistics] is...called statistical thinking or statistical literacy, a crucial component of quantitative literacy." Lynn Steen For background, read "Telling the Truth about More Damned Lies and Statistics" by Joel Best, Chronicle of Higher Education. Print version.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
This site was last updated 08/23/15
2005
09/23/22
Paul Rosenbaum's
classic Observational Studies is out in an expanded 2nd edition. This book is
a great reference. 
Tony Griffith, Math Instructor,
Westminster School. Online Feedback. I agree with the premise [of QL], but
..I teach high school math at a prep school.... Does the
admissions office at Amherst prefer a kid who can discuss Simpson's
Paradox or simplify a rational function? Does the SAT 2 reward a kid
who can understand Arrow's Impossibility Theorem or factor a cubic
function?
Milo
Schield participated in
Fiona
Fidler (
"IASSIST
has a significant role to play in helping nurture new
generations of data service providers who can help build a persistent
culture of statistically-literate students, researchers and
professionals. And may this wave of commitment to and
enthusiasm for the cause of statistical literacy thrive!’!’"
Wendy Watkins (right) and Louise Corti (far right).
Milo
Schield, Augsburg College, (shown with Louise Corti, IASSIST
Europe) argues that librarians interested in information literacy and data librarians interested in
data literacy should consider teaching statistical literacy since all
three involve critical thinking and since many social science arguments
involve statistics. 
At 
"The
YES Competition offers college scholarship awards to high school juniors
and seniors who conduct outstanding research projects that apply
epidemiological methods of analysis to a health-related issue. David
Fraser (past President Swarthmore) views "
Host
David Bressoud, Macalester College. In his keynote, "Core Concepts in QL", Milo Schield argued that QL must have core
concepts "that non-math faculty can understand easily, use regularly and
teach effectively." 
Danny
Kaplan, Macalester College, reviewed using everyday news stories in a
Quantitative Methods for Public Policy course . Policy
Topics: Response to Global Warming, Mad Cow Disease, Immigration and
Taxes. Quantitative topics that span policy topics: rates,
growth, trade-offs, detection and sampling. Average vs
Marginal Rates. 
Bernie
Madison, University of Arkansas, is teaching a unique QL course,
"News Math," based on articles in the everyday news. Topics:
Bernie
Madison, noted that the 
Bernie Madison
(5 Jan): Talk at MAA. This combination of word and number is
deadly! Check your math majors at the ends of their careers, check your
graduate students, check your colleagues.... See "
Carleton
Dean, Scott Bierman, hosted the two-day Associated Colleges of the
Midwest (ACM) workshop. He introduced speakers Joel Best (author
of Damned Lies and Statistics) and Bill Seagraves (Associate Dean
of Yale College). He reviewed data obtained in the use of
quantitative measures in
student portfolios.
Milo
Schield, Augsburg College, analyzed some 500 words that involve quantity
without specifying a number. E.g., many, most, few and some.
Schield argued that students avoid using more precise measures (rates
and percentages) because they did not know how to describe or compare rates
and percentages. Includes Schield's 2004 list of the
top 25 books on Statistical Literacy. 
Chris
Wild (New Zealand) organized a panel on the topic of "Statistics for
Life: What statistical ideas or wkills matter most, and why?"
Speakers included Denise Lievesley (ISI President 2007-2009),
Stephen Stigler (ISI President 2003-2005) and Niels Keiding (ISI
President 2005 - 2007), Nick Fisher (ISI Vice President 2003-2005) and
Milo Schield. Speakers were asked to be provacative, over-the-top
in making predictions about the future of statistical education. 
Denise
Lievesley (UNESCO) noted that "No modern society can function
meaningfully as a democracy if the electorate remains statistically
illiterate." Essential skills: "Are useful in multiple life
contexts; Are associated with social, economic , educational and health
outcomes; Can be learned and taught; and Improve individual’s ability to
adapt to change and to shape their environment." "Numeracy is the
knowledge and skills required to effectively manage and respond to the
mathematical demands of diverse situations."
Milo
predicted that 2005 would be the high point for the prevalence of
statisticians in MBA programs and for statistics in business programs.
Nick
Fisher (ValueMetrics Australia) argued that managing uncertainty
involves: Understand basic statistical graphics and how they can be
misused. Appreciating that statistical numbers ... aren’t really
as precise as they appear. Being conscious of the possibilities of
biases … Realising that reliable data can be obtained by surveying
a relatively tiny fraction of population. Realising that apparent
differences between two averages may not be real. 
Enriqueta
D. Reston (Univ. of San Carlos,) investigated the statistical literacy
of 56 graduate students using a 15 item scale. Statistical
literacy was measured at two levels: (1) understanding basic statistical
concepts, and (2) ability to understand and follow claims and arguments
based on data and graphs presented in newspapers and research reports.
students performed relatively poorly in the 15- point scale (mean =6.93;
SD=2.39; range=12-2). For the students, their relatively poor
performance was an “ eye-opener“. Further, it has served as a motivation
for them to own the goal of developing statistical literacy.
Fredrich
Ho presented 
The
IASE hosted a conference in Sydney April 2005 on "Statistical
Education and the Communication of Statistics. Kay Lipson
(center) and Brian Phillips (right) were co-chairs. Larry Weldon
(left) edited the
Kay
Lipson (left) and Sue Kokonis (center) talked on "
K.
Larry Weldon, Canada, proposed that those who design curricula
for early statistics courses should "provide a serious introduction to
graphical data analysis, at the expense of some traditional parametric
inference." In "
Milo
Schield, Augsburg College US, presented a paper on "
Cumming,
G. (left) , & Finch, S. (right). Inference by eye: Confidence intervals,
and how to read pictures of data.
American Psychologist, 60,
170-180. "For 95% confidence intervals and two independent means:
p £.05 when the bottom of one 95% confidence
interval overlaps the top of the other by about 50 percent." 
Jerry
Moreno's goal: "to produce liberal arts students who are
able ... to read a newspaper more intelligently." "We cannot afford to offer two intro courses." "Our
task is to define a few basic core statistics concepts ...and then build one course around those core concepts."
See also his article, "Statistics
as a Core Course in Liberal Arts" in
MAA's
John McKenzie,
Babson College, reported on core concepts: Variability (96%), Association-causation (82%),
Randomness (77%), Significance: practical / statistical (77%), Data collection (75%),
Sampling
distributions (71%) and Hypothesis tests (64%). John classified
courses as "statistical
literacy-reasoning-thinking" (concepts), applied stats (applications,
service) & math. statistics (theory).
The
8th annual session on Statistical Literacy was hosted by Milo Schield at
the ASA meeting in Minneapolis. Speakers included Lewis Cope
(left, Jane Miller (right), Marc Isaacson (below), Chris Olsen (lower
right), and Tom Burnham (below Isaacson). Some 90 people attended
the session. Lewis gave an animated presentation titled
Marc
Isaacson, Augsburg College, reviewed the Statistical Literacy course
first offered by Capella University in April 2005 using Schield's Statistical Literacy
textbook. Students found this course "very demanding" but "very
satisfying." They ranked this course 2nd among their 15 Gen Ed
courses in challenging students to think
critically about the subject matter.
Chris
Olson, Cedar Rapids Community Schools, presented a paper, "Epidemiology
as a Context for Teaching Confounding in Introductory Statistics." "surely the most needed
discussion about everyday inference is about what is probably the
greatest source of inferential error – the confusion of correlation with
causation, as well as a general ignorance of ... confounding."
Tom
Burnham, Cognitive Consulting, presented a paper titled "An
Online Program for Decoding Ordinary English Descriptions and
Comparisons of Percentages and Rates." This
Milo
Schield, Director of the W. M. Keck Statistical Literacy project at
Augsburg College, reviewed the prevalence
of chance-related terms on the web and used these prevalences to
identify those topics that should be part of a general statistical
literacy course. Includes unusual topics such as evolution, ESP and a
Bayesian approach to statistical significance.