The critique of modelling and sensitivity analysis in ...

Description

• cours - matière potentielle : at the modeller
• expression écrite

The critique of modelling and sensitivity analysis in the scientific discourse. An overview of good practices Transatlantic Uncertainty Colloquium (TAUC) Washington, October 10-11 Andrea Saltelli October 5, 2006

• manmade systems into sets of mathematical rules
• bootstrapping of the modelling process
• part of the scientific method
• sensitivity analysis
• parameters
• input
• model
• models
• uncertainty
• system

Sujets

• Analysis
• System
• Uncertainty
• Parameter
• Mathematics

Informations

• Publié par: istyo
• Nombre de lectures: 17
• Langue: English

Extrait

Best Practices in Civic Education:
Changes in Students’ Civic Outcomes
Amy K. Syvertsen*, Constance A. Flanagan, &
Michael D. Stout
The Pennsylvania State University
* [e-mail]
CIRCLE WORKING PAPER 57
AUGUST 2007CIRCLE Working Paper 57: August 2007 Best Practices in Civic Education Best Practices in Civic EducationCIRCLE Working Paper 57: August 2007
The power of civic education to elicit METHODOLOGY
positive student outcomes has been empirically
documented. However, the feld is only now The data used to the present study were
beginning to understand the causal processes gathered from two waves of surveys with 1,670
that bring about these positive changes in young students ages 14-19 from 80 social studies classes
people. The Civic Mission of Schools report in the United States. Classes were recruited from
commissioned by the Carnegie Corporation a pool of teachers throughout one mid-Atlantic
of New York and CIRCLE (2003) lays out six state who had expressed interest in training in an
“promising approaches” to civic education. These election-based curriculum. Teachers and students
practices emphasize the need for instruction that completed a pre-test at the beginning of the
is relevant to young people (i.e., links young semester (early to mid-September) and a post-
people’s interests to political contexts), provides test at the end (late-November to mid-December)
opportunities for practice, and that moves beyond leaving a 2.5 to 3.5 month lapse between the
rote learning praxis. A growing body of evidence points of data collection over the course of which
suggests that these approaches to civic education the national election occurred. The student and
yield positive, lasting outcomes in young people teacher survey instruments used at both time
(see CIRCLE, 2007). For example, research has points consisted of a combination of open-ended
found that deliberative classroom discussions and Likert-type items. Items on the student
are positively associated with interest in politics, survey were counterbalanced to ensure that all
political knowledge, and feelings of political effcacy questions had an equal chance of being answered.
(Feldman, Pask, Romer, & Hall Jamieson, in press; Questionnaires were distributed to students during
McDevitt, Kiousis, Wu, Losch, & Ripley, 2003). The a 45-minute class period at each occasion of
goal of this project is to examine the association measurement.
between activities regularly used in civic education The study was originally designed as a
courses (e.g., staging a mock election) and their randomized evaluation of a civics curriculum
impact on key student outcomes. By linking called Student Voices in the Campaign (for
classroom praxis to outcomes, we intend to additional information on the program evaluation,
provide evidence supporting best practices in civic see Syvertsen, Flanagan, & Stout, 2007). The
education. current study, however, does not evaluate that
particular program. Rather, we look at a range of
usual practices that social studies teachers report
Figure 1. Breakdown of sample by grade.
14%
10th Grade
11th Grade
12th Grade
25%
61%
www.civicyouth.org 2 www.civicyouth.org 3CIRCLE Working Paper 57: August 2007 Best Practices in Civic Education Best Practices in Civic EducationCIRCLE Working Paper 57: August 2007
using and assess whether various practices had a MEASURES
demonstrable impact on targeted civic outcomes
for students. The practices which we assess are Teachers reported (using a long list) on
common practices and activities that are regularly which practices they used in their classes over
found in standard civic education curricula and the course of the Fall 2004 semester. In order to
classrooms. match exposure to a specifc activity with student
outcomes, the teachers’ reports were related to
the students’ reports of various civic outcomes. PARTICIPANTS
Rather than matching each individual activity
with a specifc student outcome, scales were In total 80 teachers and 1,670 students
created (i.e., like items were grouped to create completed surveys at the beginning and end of the
a single measure). For example, teachers were semester. The mean age of the students was 16.63
asked four separate questions concerning whether years (SD = .85). Students were in grades 10, 11,
the students in their class participated in feld and 12. As illustrated in Figure 1, the majority of
th trips to Washington, DC, the state capitol, local students were in the 12 grade, which research has
government offces, and polling stations. Instead shown to be the optimal grade for civic education
of testing whether a feld trip to each unique locale (Niemi & Junn, 1998). Fifty percent of the students
made students more trusting of elected offcials, were female. The ethnic background of the
feld trips to all four locations were grouped into a participants was 92% European-, 6% African-, 3%
single Field Trip measure which, in turn, was used Hispanic-, 3% Native-, and 2% Asian-American. An
to measure the impact of feld trips on students’ additional 2% of the participants identifed as being
trust for elected offcials. of some other ethnicity. Note that percentages do
The research team created the instructional not add up to 100% as several students indicated
practice scales based on theory and the common multiple ethnicities. These ethnic breakdowns
underlying skill or disposition targeted by the refect the overall student population in the school
activity. To get a score on each measure, the districts in the study.
number of activities for which the teacher Adolescents’ socioeconomic status (SES)
responded “yes” was summed. The various was calculated based on their reports of mother
instructional practice scales ft under three broad / female guardian’s highest educational level.
categories which seek to enhance: Students reported that their mother / female
Civic Skillsguardian’s highest level of education was: high
* Communication Skillsschool or less (43%), technical or vocational
* Democratic Deliberationtraining (7%), 2-year college degree (12%), 4-year
* Critical Analysis of Political Informationcollege degree (24%), graduate degree (14%).
Civic Engagement Teachers in the sample had a wide-range
* Election Simulationof teaching experience with careers ranging from
* Electoral Engagement2 months to 37 years (M ≈ 14 years; Mode = 5
* Alternative Engagementyears). Thirty-eight percent of the teachers were
Awareness of Civic Issues and Conceptsfemale. The majority of teachers described the
* Local Issuesclass participating in this study as having mixed
* Youth Issuesabilities (83%), while 16% were identifed as AP/
* Civic Education ConceptsGifted and 1% were considered remedial. Eight
* International Issuespercent of the teachers indicated that the class in
* Contested Issueswhich this study took place was required to fulfll a
* Current National Events.graduation requirement.
These measures are described in more
detail in Tables 1, 3, and 5 later in this report.
www.civicyouth.org 2 www.civicyouth.org 3CIRCLE Working Paper 57: August 2007 Best Practices in Civic Education Best Practices in Civic EducationCIRCLE Working Paper 57: August 2007
When there was an explicit hypothesis about and interest. Similarly, mother’s education was
the outcome of a specifc activity, the activity included as a covariate to control for differences
was entered into the analysis independently. For in civic interest and opportunities associated with
example, if we thought – over and above all of the parental education.
other feld trips – that a trip to Washington would The combination of teacher and student
make students more likely to express interest in a reports, the multilevel design, and inclusion of
political career, we tested the affect of a “feld trip prior political interest and mother’s education as
to Washington” independently from all of the other covariates refect considerable rigor in the analyses
feld trip locations. reported in this paper. The results of these analyses
The measures used to assess students’ can be used to inform civic education standards
civic outcomes have been discussed in detail in and practices.
the CIRCLE Working Paper, Civic Measurement We have organized the results into four
Models: Tapping Adolescents’ Civic Engagement sections: (1) Civic Skill, (2) Civic Engagement,
(Flanagan, Syvertsen, & Stout, 2007). Readers are (3) Awareness of Civic Issues and Concepts,
directed to this piece for detailed information on and (4) Specifc Activities. Each section of the
the psychometric properties (e.g., alpha coeffcient, report includes a table outlining the instructional
individual item factor loadings) and th