Alternate Title: MET 4 - District-Wide Files
Principal Investigator(s): Bill and Melinda Gates Foundation
The Measures of Effective Teaching Project (MET)
The MET project is based on two premises: First, a teacher's evaluation should depend to a significant extent on his/her students' achievement gains; second, any additional components of the evaluation (e.g., classroom observations) should be valid predictors of student achievement gain.
Student achievement was measured in two ways -- through existing state assessments, designed to assess student progress on the state curriculum for accountability purposes, and supplemental assessments, designed to assess higher-order conceptual understanding. The supplemental assessments used were Stanford 9 Open-Ended Reading Assessment in grades 4 through 8, Balanced Assessment in Mathematics (BAM) in grades 4 through 8, and the ACT QualityCore series for Algebra I, English 9, and Biology.
Panoramic digital video of classroom sessions were taken of participating teachers and students, teachers submitted commentary on their lessons (e.g., specifying the learning objective) and then trained raters scored the lesson based on classroom observation protocols using the following five observation protocols:
- Classroom Assessment Scoring System (CLASS), developed by Robert Pianta, University of Virginia
- Framework for Teaching, developed by Charlotte Danielson
- Mathematical Quality of Instruction (MQI), developed by Heather Hill, Harvard University, and Deborah Loewenberg Ball, University of Michigan
- Protocol for Language Arts Teaching Observations (PLATO), developed by Pam Grossman, Stanford University
- Quality Science Teaching (QST) Instrument, developed by Raymond Pecheone, Stanford University
A subset of the videos also are being scored using an observational protocol developed by the National Board for Professional Teaching Standards (NBPTS)
Close to 3,000 teacher volunteers from across the following six, predominantly urban, school districts participated in the MET project: Charlotte-Mecklenburg Schools, Dallas Independent School District, Denver Public Schools, Hillsborough County Public Schools, Memphis City Schools, and the New York City Department of Education. Participants teach math and English language arts (ELA) in grades 4-8, Algebra I, grade 9 English, and high school biology.
The District-Wide Files
The district wide files are comprised of one data file per district for each of the 2008-2009, 2009-2010, and 2010-2011 school years, for a total of eighteen data files. Each file contains information on each student in the school district including student demographic variables, such as race, age and gender, specialty student status variables such as free lunch, English language learner, and gifted and talented program participation, and student-level test rankings for math and reading. Also included are the aggregate means of those student demographic, specialty status, and test score variables for each teacher.
Two versions of the SAS code used for creating the original value-added measures for teachers in each district are also provided with the data files. The non-aggregated version of the code uses the individual math and reading test rankings for each of a teacher's students to calculate predictive score estimates. The aggregated version of the code uses the mean math and reading test rankings for all of that teacher's students to calculate estimates.
One or more files in this study are not available for download due to special restrictions; consult the restrictions note to learn more.
The Measures of Effective Teaching Longitudinal Database (MET LDB) is restricted from general dissemination; a Confidential Data Use Agreement must be established prior to access. Researchers interested in gaining access to the data can submit their applications via ICPSR's online Restricted Contracting System, linked above.
Applicants will be required to:
- Submit IRB approval/exemption documentation;
- Scan and email the completed Confidential Data Use Agreement, signed by the Primary Investigator and an Institutional Representative;
- Pay the access fee ($500 for up to five users, and $250 for each additional user beyond five)
Please visit the MET LDB Web site for more information.
WARNING: Because this study has many datasets, the download all files option has been suppressed, and you will need to download one dataset at a time.
Bill and Melinda Gates Foundation. Measures of Effective Teaching: 4 - District-Wide Files, 2008-2011. ICPSR34798-v1. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2014-03-06. doi:10.3886/ICPSR34798.v1
Persistent URL: http://doi.org/10.3886/ICPSR34798.v1
Scope of Study
Smallest Geographic Unit: School District
Unit of Observation: Teachers, Students
Universe: Teachers and students within the six participating school districts.
Data Types: administrative records data
Data Collection Notes:
Value-added estimates from the MET LDB will not exactly match the value-added scores appearing in the MET analytical files. In response to concerns of the school districts, ICPSR has modified the district-wide data to protect the confidentiality of individual students. All changes to the data were designed solely to mask the identities of students, and their impact on value-added estimates is very small. When the RAND scripts are used to compute value-added estimates, the correlation coefficients between estimates derived from the original data and the modified data are 0.98 or higher.
The MET Study addressed several related research questions: How reliable and valid are the specific measures of teaching effectiveness under study? Do the various measures identify distinctive dimensions of teaching effectiveness, and if so, what dimensions are identified? What measures of effective teaching are empirically related to student learning gains? What does effective teaching look like, and how does it compare to less effective teaching? For example, what is the distribution of teacher scores on measures of effective teaching, and how much difference is there in teacher knowledge scores, teaching practice scores, and student outcome scores among teachers at different points in the distribution of measures of effective teaching? Can multiple sources of data on teachers and their teaching be combined to develop a set of fair, valid, and reliable indicators of teaching quality for use in teacher evaluation systems intended to rank teachers for personnel decision making and to promote teachers' professional learning and development?
Teachers at targeted grade levels (4th-9th grade) and in targeted subject areas (English Language Arts and Math with some Biology) were recruited from schools in six school districts. Video was captured of these teachers' classroom sessions and that video was scored using multiple measure to rate teaching method and classroom environment. Multiple assessment instruments and standardized test scores were used to gather student achievement data. Multiple surveys were administered to gather student, teacher and principal opinion on topics such as classroom instruction as well as classroom, school and working environment. Year two of the study included a teacher randomization.
The MET Study began with a process of "opportunity" sampling that took place over the period July - November 2009 and that resulted in six, large school districts volunteering to participate in the study. The process of opportunistic sampling then continued as elementary, middle, and high schools within each district were recruited into the study. Once schools were recruited, opportunity sampling continued as teachers (at targeted grade levels and subject areas) within these schools volunteered for the study. The sampling process resulted in 2,741 teachers from 317 schools in six large school districts being recruited into the first year of the study. Attrition in the teacher sample in year two of the study resulted from schools that dropped out of the study (11 schools; 60 teachers). Additionally, individual teachers dropped out when they left their school or district, began teaching a different subject or grade, lost interest in the study, or became ill. Overall, the year two sample of teachers included 2086 teachers in 310 schools. Of the 582 4th and 5th grade teachers in year two, the majority continued to be subject-matter generalists who taught English Language Arts (ELA) and Mathematics to a single class of students, although the sample also included a small number of subject matter specialists (who taught ELA or Mathematics to more than one class section of students) and teachers who volunteered only to have their teaching of a single subject be studied. Of the 841 middle grades teachers in year two, about half continued to be teachers of ELA in grades 6-8, and the other half teachers of Mathematics at these grades. Of the 479 9th grade teachers in year two, about a third were teachers of 9th grade English, another third were teachers of 9th grade Algebra I, and another third were teachers of 9th grade Biology.
Mode of Data Collection: cognitive assessment test
Administrative data were gathered from each of the six participating school districts.
Response Rates: 2746 teachers began the year 1 of the MET project, 1868 completed year 2 of the MET project.
Extent of Processing: ICPSR data undergo a confidentiality review and are altered when necessary to limit the risk of disclosure. ICPSR also routinely creates ready-to-go data files along with setups in the major statistical software formats as well as standard codebooks to accompany the data. In addition to these procedures, ICPSR performed the following processing steps for this data collection:
- Created variable labels and/or value labels.
Original ICPSR Release: 2013-09-01
- 2014-03-13 Added the 2010-2011 value added code files and made updated the readme file.
- 2014-03-06 Released the 2010-2011 District-Wide Files for all six districts, as well as the 2010-2011 value-added code files.
- 2013-10-31 Fixed some labeling issues in the value-added SAS code files.
- 2013-10-17 2013-09-25 Released value-added code files.
- 2013-09-24 2013-09-23 Standardized missing values in all data files. Clarified variable labels and value labels in all data files.
- Citations exports are provided above.
Export Study-level metadata (does not include variable-level metadata)