Tuesday, January 31, 2012

A Coherent Network of Assessments is a Key to Learning for All


With the emphasis on accountability demanded by Race to the Top (RTT) current assessment systems need to change.  This change will need to begin with annual testing.  Thereafter, classroom assessment will need to follow. 

In an Assessment and Accountability Comprehensive Center (AACC) report, Herman (2010) detailed how a better assessment network needs to begin with the conception of assessment not as a single test but as a coherent system of measures. Coherent systems must be composed of valid measures of learning and be horizontally, developmentally, and vertically aligned to serve classroom, school, and district improvement (p. 1)

A sole multiple choice test, administered in an hour or two, cannot cover the full range of year-long standards representing what students should know and be able to do.  In contrast, a system composed of multiple assessments can illuminate a broader, deeper perspective of student knowledge and skills. A second assessment for example, cannot only assess more content knowledge, but, if designed to measure applied knowledge, can evaluate different types of skills (p.2).

An assessment system comprised of multiple types of measures can provide a more thorough picture of student learning. Such systems also can be more responsive to the diverse decision-making needs for those who need data to support improvement—teachers, administrators parents, students. A solitary, end-of-year test simply cannot provide sufficient formative information to guide teaching and learning throughout the year (pp. 2-3).

Coherent assessment systems are comprised of component measures that each reflect significant learning goals and provide accurate information for intended purposes.  Drawing from the KnowingWhat Students Know National Research Council conception (National Research Council [NRC], 2001), coherence starts with a clear specification of the goal(s) to be measured.  Next, assessment tasks are specially designed or selected to reflect the learning goal(s).  Finally, an appropriate interpretation framework is applied to student responses to reach valid conclusions about student learning—for example, a score of “proficient” on a state test or an inference about the source of a student’s misunderstandings in teachers’ formative practice (Herman, 2010, p. 3).

Furthermore, Herman (2010) notes this creates “a more fragile base for classroom teaching and learning, the emphasis on a system of assessments by the addition of through-course exams to complement end-of-year assessments is very promising” (p.6).  How does this promote coherence?  Herman (2010) notes that using “through-course exams—more extended, performance-oriented assessments conducted during the course of instruction—provide rich opportunities to assess students’ thinking and reasoning as well as their ability to apply and communicate their knowledge and skills in solving complex problems.  Performance assessments also provide useful models of effective teaching while supporting authentic instruction and student learning" (p. 6).  Coherence in assessment networks could create data-based accountability systems that “support educational improvement, better education for all students, so that every student is prepared for college and success in life” (p.7).  Learning for all must be the goal.  A coherent, data-based accountability system is a correlate of Effective Schools.  Therefore, it becomes inherent upon true educators to embrace this concept. 

References:

Herman, J. L. (2010). Coherence: Key to Next Generation Assessment Success (AACC Report).

            Los Angeles, CA: University of California.

Thursday, January 5, 2012

Appropriate Disaggregation of Student Performance Data Increases Learning for All


The use of data for school reform is well explicated in Effective Schools research.  Researching this correlate, Bambrick-Santoyo (2008) found given “the proper interplay among interim assessments, analysis, action and data-driven culture, schools can be transformed, and a new standard can be set for student learning” (p. 46).  As a result of using data-driven instruction teacher buy-in is actually created. 
No Child Left Behind Act of 2001 (NCLB) requirements advanced the use of disaggregated data as a powerful motivator for change and school improvement.  In a 2008 press release, then USDE Secretary Spellings noted, disaggregation data is especially important for closing the achievement gap between poor and minority students and their peers.  Spellings contended, “the more information consumers have, the better equipped they are to demand, and achieve, lasting school improvement. And the more information teachers have, the better able they are to customize and improve instruction” (April 2008).
While the NCLB included benefits to students with disabilities (SWD) it also created some barriers that might prevent students with disabilities from enjoying all of the opportunities in the law.  While NCLB requires schools, school districts, and states disaggregate test results for several subgroups of students in an effort to increase the accountability of at-risk groups of students and thereby close the achievement gap.  Students with disabilities are one of the subgroups that must be disaggregated.  NCLB allows states to set a minimum number of students that each subgroup must contain before the data for that particular subgroup can be used for purposes of determining achievement.  Cortiella (2010) notes “the minimum number of students in each subgroup is to be based on what would be sufficient to yield statistically reliable information as well as to make sure that disclosing the results for a particular small subgroup would not, in fact, result in revealing the identity of the students in that subgroup” (para 9). 

A possible unintended consequence of allowing flexibility in SWD subgroup size is that public reporting of any subgroup’s performance is intended to highlight achievement gaps and thereby motivate schools to improve when necessary.  However, schools can escape scrutiny because of subgroup size that may not focus the same level of effort on students whose results aren't reported on the school or district level due to subgroup size.  These schools and districts can avoid a "needs improvement" rating under NCLB if the SWD subgroup size doesn't meet the state minimum.  Therefore, Cortiella (2010) finds “this provision could result in schools attempting to limit the number of students with learning difficulties it qualifies for special education services” (para 10).  For this reason, as an ethical and moral imperative, the appropriate use of disaggregated data must be required and ensured to truly promote learning for all.

References:
Bambrick-Santoyo, P. (2007). Data in the driver's seat. Educational Leadership, 65(4), 43-46.
Cortiella, C. (2010)  No Child Left Behind and Students With Learning Disabilities:
            Opportunities and Obstacles
Spelling. M, (2008) U.S. Secretary of Education Margaret Spellings Announces Department Will
Move to a Uniform Graduation Rate, Require Disaggregation of Data.