Guide to Teacher Data

Improving State Need Assessments of Secondary Science and Mathematics Teachers: Challenges, Possibilities, and Recommendations

Michael B. Allen
March 2010

Teacher Data and Data Systems

Improving Teacher Data

The possibility of reliable state-level or district-level projections of the need for science and mathematics teachers depends upon the adequacy of two essential elements:

1. The available primary data on teacher supply and demand
2. The statistical methods used to enlarge upon those data and derive the projections desired.

Two other units of this project, Establishing a State’s Current Need for Science and Mathematics Teachers and Projecting a State’s Future Need for Science and Mathematics Teachers, discuss in some detail the methodological issues involved in generating estimates of current and future teacher need. Although a number of these methodological challenges remain largely unresolved from a theoretical standpoint, the project nevertheless attempts to provide reasonable, if appropriately qualified guidance to individuals who have the practical task of evaluating existing efforts to generate estimates of teacher need or developing new ones. That guidance also includes outlines of the kinds of data required for such estimates, both the data needed for the most basic calculations and the data that would be necessary to develop estimates that are more sophisticated and possibly more precise. The Research Analysis unit of this project provides a more extended discussion of the statistical-methodological issues involved. This includes some discussion of important data considerations that is, however, far from comprehensive. This Guide to Teacher Data is intended to provide a more extensive treatment, albeit one that is more practical than theoretical, of the data requirements for generating solid estimates of teacher supply and demand.

Over the last decade, states in the U.S. have made significant progress in building sophisticated data systems that contain detailed information on each student’s academic history and performance – in some cases, not only for students’ K-12 education but also for their post-secondary careers. This progress can be attributed to the emphasis in No Child Left Behind on measuring student achievement and to the specific efforts of the Data Quality Campaign¹ and others, including, most recently, encouragement from the U.S. Department of Education’s Race to the Top initiative for states to develop robust student data systems. Increasingly, these data systems also include the assignment of unique identifiers to all K-12 teachers in the states so that students’ progress can be matched with their assigned teachers and the effectiveness of those teachers thus measured in terms of their students’ performance. Although the assignment of unique identifiers to teachers is also an essential precondition for tracking the path of teachers’ careers, only a relatively few states have given the same kind of attention to assembling systematic information on their K-12 teachers that they have given to increasing their capacity to track the progress of their students.

This is all the more surprising and unfortunate in view of the fact that excellent guidance for developing sophisticated teacher data systems has been available for several years. Particularly important and helpful in this respect is the 2003 report written by Richard Voorhees and Gary Barnes and published by the State Higher Education Executive Officers (SHEEO), Data Systems to Enhance Teacher Quality.² This study summarizes the elements required for a comprehensive teacher data system, discusses the obstacles that currently hinder many states in their efforts to develop such a comprehensive system, and offers suggestions for overcoming these challenges.

Indeed, for the limited purposes of our project, we do not believe we can improve upon the effort of Voorhees and Barnes. And, so, rather than attempt to reinvent what already has been accomplished quite adequately, we refer the reader to the Voorhees and Barnes study and include here, with the permission of the study’s publisher, a brief summary of the study’s recommendations regarding the data elements that a state should track. On occasion, we have amended those recommendations slightly for the sake of greater consistency with our specific purpose in this project of facilitating estimates of the need for science and mathematics teachers. Similarly, we have omitted several data elements identified by Voorhees and Barnes that we believe are unnecessary for our specific purpose. Although the recommendations in this Guide focus on the content of the data, Voorhees and Barnes additionally discuss considerations of data quality, which are also critically important for those interested in using and developing teacher databases.

A Teacher Data System

The overarching recommendation of the Voorhees and Barnes study is for states to create a data system that draws upon a multitude of data sources. Without an effort at systematization, these sources are often administered independently, are of varying quality, and likely contain data that may not be completely consistent or compatible between sources. These include databases maintained by a number of state agencies: the certification and licensure authority, retirement system, and department of employment. They include district-level databases on teacher employment and assignments, school conditions, and teacher retirement (when district-based retirement systems are in effect). And they include data maintained by colleges, universities, and any other providers of teacher preparation on their students and their preparation programs. Creating a comprehensive data system requires that such sources of data be made as complete and consistent as possible and that they be linked to one another by assigning a common unique identifier – such as a Social Security number – to each individual for whom data exist. Collecting the categories of data listed below to track the careers of individual teachers enables states to assess the basic quality of their teacher workforce, more accurately understand their teacher supply and demand picture, identify disparities in teacher need and quality between districts, and more confidently evaluate the effectiveness of the various programs and policies in place to prepare, recruit, retain, and develop teachers.

Although we present only an outline of the recommendations in the Voorhees and Barnes study, the outline nevertheless includes a substantial number of data elements. As the full study indicates, some of these data elements are much easier to obtain than others. And the task does not only involve obtaining the data once but of continuing to collect and update them over time. Ultimately, of course, each individual state must decide how extensive it wants its teacher data to be. This is a decision that involves issues of cost, privacy and access, and jurisdiction. From the standpoint of analyzing teacher supply and demand, it is also a decision that affects the degree of thoroughness and precision such analyses can have. A state that seriously limits the quality and extent of the data it collects on its teacher workforce will pay the price of limiting the state’s understanding of the strengths and weaknesses of that workforce and of compromising the state’s ability to craft effective policies and practices to improve it.

Essential Data Elements

1. Employment Data on Individual Teachers

Many of these data elements and those identified in 2. below are the most essential to state efforts to track the supply, demand, and basic adequacy of their teacher workforce and to identify attrition and retention patterns. The salary data facilitate the comparison of educational costs between schools and districts across the state. Unless otherwise indicated, all of these data are assumed to be longitudinal – i.e., to be collected on an ongoing basis over time.

• Current year’s teaching status, including school and specific course assignments
• Previous years’ teaching status, including school and specific course assignments for each individual year
• Status as mentor or master teacher (if not already included in licensure categories)
• Percentage FTE assigned to teaching and non-teaching responsibilities
• Compensation – including base salary and additions
• Current and historical employment outside of teaching³ (including salary)

2. Licensure and Certification Data on Individual Teachers

Maintaining these data is essential for tracking the state-issued credentials of teachers – a baseline indicator⁴ of teacher adequacy – and comparing the baseline quality of teachers between schools and districts.

• Licensure examination history and results⁵, including attempts and scores for all general and specialty area licensure examinations
• Date of initial licensure
• Licensure renewals, additions, advancements, and dates these were received
• Subject certification(s) and/or endorsements and dates received and renewed
• National Board for Professional Teaching Standards certification and dates received and renewed
• Out-of-state transfer information:⁶ Years teaching out-of-state; undergraduate and graduate institution(s) and degree(s) received; subject major(s) and minor(s); preparation program attended and completion date; prior licenses, certifications, and endorsements (including dates received and renewed)

3. Professional Development Data on Individual Teachers

Maintaining these data facilitates understanding of the continuing education activities of a state’s teacher workforce. It may provide an indication, in particular, of whether or not a state’s science teachers are maintaining currency in their ever-changing fields – and whether there appears to be a disparity between districts in the currency of their teachers’ knowledge. Teachers’ participation in continuing education also may be linked to retention and attrition patterns, a link the collection of the data identified here and in 1. above makes it possible to probe.

• Participation in mentoring and induction programs
• Participation in continuing education and professional development
• Advanced degrees received, including coursework

4. Teacher Preparation Data

Maintaining these data – which most logically would be collected by each teacher preparation program – increases understanding of the current quality of a state’s teacher workforce and the potential for improving it. It enables states, districts, or individual teacher preparation institutions and K-12 schools to probe the possible relationship between differences in teachers’ educational backgrounds and disparities in the performance of their pupils. Aggregated by preparation program, these data also facilitate understanding of program outcomes (e.g., their success in preparing minority teachers), assessment of program effectiveness, and comparison of baseline program quality.

• Data on individual teacher candidates⁷
  • Demographic information – age, gender, race/ethnicity
  • Institution and specific program attended for teacher certification (e.g., 2-year program in the college of education, special program in the college of natural sciences, post-baccalaureate certificate program, district residency program, etc.)
  • Undergraduate major(s), and minor(s)
  • Field of graduate study if license pursued via a graduate degree program
  • GPA in major (or in field(s) of certification) and overall
  • Date of graduation from teacher preparation program
• Aggregated data on individual teacher preparation programs, whether traditional or alternative
  • Number of students who enter the program each year
  • Number and percentage of entering students, by demographic, who continue in and complete the program each year
  • Pass rate and distribution of student scores on licensure examinations
  • GPA of teacher program graduates in their science and mathematics major in comparison to GPA of other students at the institution in the same major⁸

5. Non-Teacher Data

Linking these data to data on individual teachers makes it possible to assess teachers’ effectiveness on the basis of their pupils’ academic performance. It also opens the possibility of looking for correlations between teachers’ preparation history or other factors in their background and their pupils’ performance.

• School characteristics, including student SES profiles and class size, ideally specifiable for each individual class taught
• Student performance data for each individual student
• Identification of individual students’ classrooms or teachers

Endotes

1. See the Data Quality Campaign at http://www.dataqualitycampaign.org/

2. Voorhees, R. A., & Barnes, G. T. (2003). Data Systems To Enhance Teacher Quality. Accessed at http://www.sheeo.org/resources/publications/data-systems-enhance-teacher-quality-0

3. This facilitates the identification of entry, re-entry, and attrition patterns that may provide important information for developing various kinds of incentives. It would require a state’s ability to track licensed teachers not only through the teacher employment database but also through a comprehensive employment and retirement database. Few, if any, states currently have that capacity.

4. The Teacher Quality and Teacher Licensure unit of this project contains a detailed discussion of the role of teacher licensure and certification.

5. Several relatively recent research studies indicate that teachers who have had to take licensure examinations repeatedly before passing them may be less effective than teachers who pass the examinations more readily. See Presley, J. B., White, B. R., & Gong, Y. (2005). Examining the Distribution and Impact of Teacher Quality in Illinois. Carbondale, IL: Illinois Education Research Center; and Boyd, D., Lankford, R. H., Loeb, S., Rockoff, J., & Wyckoff, J. (2007). The Narrowing Gap in New York City Teacher Qualifications and its Implications for Student Achievement in High-poverty Schools. Albany, NY: Teacher Policy Research.

6. This list of data elements is slightly expanded from the recommendations of Voorhees and Barnes. If state licensure agencies do not collect these data on teachers transferring in from other states, however, it is likely to be inaccessible.

7. Not all, candidates enrolled in preparation or certification programs will complete the programs, obtain licensure, and enter the classroom, but it is valuable to collect data on all of them.

8. This is not a data element called for in Voorhees and Barnes, but it is used by some university-based preparation programs to assess the academic talent of their students and may be an indicator of a problem if the GPA of teacher candidates is significantly lower than that of other students in the same major.