VAM, Value-Added (Assessment) Models used in part to evaluate a teacher’s performance. VAMs are used to compensate, evaluate, continue a contract, or fire a teacher. The American Statistical Association (ASA) suggests that high-level of expertise is needed to develop and interpret the models, especially if the VAMs are used for high-stakes reasons. According to the ASA executive summary, most VAM studies have found that only about 1% to 14% of the irregularity in test scores are attributed to a teacher. This means that when a child takes a test and their scores are used to evaluate a teacher, less than 14% of their scores are actually attributed to the individual teacher.
VAMs are not designed to predict future learning outcomes. A standardized test score from a norm-referenced assessment does not accurately reflect how much a student has learned. Outside factors such as culture, environment, social strata, economic-level, community, and level of basic needs being met are all external inputs that may affect a student’s score. The creativity of a student cannot be measured in a standardized test, complex thinking cannot be measured by a standardized test, and analytical thinking cannot be measured by a standardized test. Student achievement includes attributes such as participation, endurance, motivation, curiosity, self-awareness, self-discipline, civic-mindedness, resilience, resourcefulness, and integrity. I suggest these are the skills of our future, not test-taking strategies. Why do we let our children suffer through endless practice tests, exam prep, practice sitting still –yes that’s a real thing- lack of recess-arts-media time- and pre-printed text passages that rely on multiple choice answers for in-depth thinking skills like comprehension and critical analysis in order to unfairly judge their potential and the quality of their teacher.
VAMs use a form of regression model to predict scores or growth. If the regression model does not account for high level of economically disadvantaged students, outside tutoring, number of ELLS in the class, or class-size then the model loses some validity; this unfairly affects the teacher. When the VAMs contribute 50% of a teacher’s evaluation as it does in Duval County, it is promoting an invalid interpretation of the teacher’s actual ability to teach. I suggest that a student’s score, alongside other learning outcomes, be used to evaluate the curriculum, provide techniques based on positive learning gains, determine the resources needed to help future students, and improve programs already in place or provide programs needed to adjust for lack of student growth; not used to determine student promotion or graduation, teachers’ pay, or longevity of career.
Extra test prep, exclusion of subjects to work on the test, removal of recess and arts to increase time spent on the “big two” math and reading, and overall increase of stress induced by the high level of importance placed on one exam are some of the unintended consequences caused when high-stakes testing scores are reflective of a teacher’s capabilities.
“It does not provide information on how to improve the teaching” (ASA)
According to state statute 1008.22 (1)(c)(f) The purpose of the assessment program is use the data to improve instruction and determine the long-range capabilities of student achievement -the very thing standardized scores cannot do- yet we not only use them, we tout them as the best way to determine the value of our teachers.
Duval is one of 19 districts in Florida to use the (Charlotte) Danielson Framework for Teaching model to develop the teacher evaluation model. It should be noted that in 2011 The Framework for Teaching model was enhanced by changes made after being part a 2009 research project by the Bill and Melinda Gates Foundation entitled Measures of Effective Teaching (MET). The project studied videos of 23K lessons, which were then analyzed using five observation protocols. Online training and certification of hundreds of observers for the purpose of rating the quality of teaching using the Danielson Group Framework from 2007 to provide input for the study. The newly certified observers watched videos of lessons and using a rubric, rated teachers’ lessons. The 3,000 teachers used in this project were volunteers in six school districts around the country teaching math and English in grades four through eight.
Please read more about me at DernforDuval.com