Get documents about "Student Tracking Systems A Literature Review
Document Sample
P-16 Data Sharing Systems…
What is Being Done and What is Needed
Introduction
The literature review offers a summary of how students are followed through their educational
experience and into the workforce and focuses on how data can be shared among those concerned with
the improvement of teaching and learning. While there is considerable literature on the systems that
follow students through educational endeavors, there is less literature on systems that follow students
into the workforce.
Literature Review
Interest in increasing the depth and breadth of educational accomplishment in the United States
has drawn new attention to standards, data, and accountability systems. It has also engendered some
heated discussions among educators and policy makers as well as parents and other stakeholders
concerned with the kinds of standards, data, and accountability systems needed to help improve
educational achievement. As the accountability movement builds momentum in higher education,
institutional, state, and national policymakers are searching for ways to improve student outcomes.
Among these outcomes are retention, graduation, transfers from two-year to four-year institutions,
employment, and earnings. Policymakers are increasingly interested in the return on investment of
public dollars spent on higher education. Concerns about efficiency, equity, and cost effectiveness are
driving policy makers to demand more data with more accurate reporting of student outcomes.
Increasingly, federal and state leaders are using longitudinal data systems for both policymaking and
school improvement. In response, states are investing more resources in the systems design,
development and use.
Several recent trends have prompted interest in monitoring student progress from pre-school
throughout college and into their professional lives. Bers (1989) argues that increasing emphasis on
marketing, accountability, communication with students, and internal competition for students all serve
as catalysts for the development of tracking systems.
P-16 Data Sharing Systems…What is Being Done and What is Needed
Purpose of P-16 Tracking Systems
Educational systems of the 21st century are being asked to double the production of the
educational system of the mid 20th century with no compromise on quality. Student educational careers
do not begin with college admission; rather success in college is closely linked to elementary and
secondary school experiences. As society asks complicated questions about how well our educational
institutions perform, the need for better tracking systems increases.
Several states are undertaking P-16 initiatives to improve the connection among and transitions
between components of education systems. According to the Education Commission for the States
(Virginia‘s P-16 Education Council, 2006), common elements of these initiatives include:
Inclusiveness with the aim of improving the education of all students
Aligning efforts at all levels
Support standards and assessments
Establishing a logical progression across systems
Reducing the need for remediation
Identifying and removing artificial barriers to student progression and success
Promoting greater collaboration between education professionals at all levels
Advancing widespread parent, community, and student understanding of goals and expectations
Reducing dropout rates in both secondary schools and colleges
Applying these nine elements is expected to lead to higher education levels across all income and
ethnic groups, which in turn is associated with greater employment stability, increased civil engagement
and a decline in public assistance and crime rates. Statewide data systems allow coordinating and
governing boards to address complex policy questions with increasing skill and efficiency.
SHEEO (State Higher Education Executive Officers) identifies five key components of an integrated
educational system, which can be described and elaborated in many different ways. Essential
components include (Lingenfelter, 2003):
Early outreach to motivate parents and students to hold high educational aspirations and show
what is required for postsecondary educational achievement.
Curriculum and assessments systems that specify the knowledge and skills students need and
assist teachers in the assessment of academic progress
High quality teaching to enable students to achieve at higher levels
Student financial assistance to encourage and support participation in postsecondary education
Data and accountability systems to permit educators and policy makers to monitor progress and
guide efforts to promote greater achievement
Effective and comprehensive systems share several common characteristics. They inform all
stakeholders of the condition of education at various levels. They enable states to identify effective
NBH Solutions, LLC Page 2
P-16 Data Sharing Systems…What is Being Done and What is Needed
educational practices and diagnose problems. They have the potential to increase the commitment
among stakeholders to collect, analyze, and use information on student performance. Effective systems
also have the ability to identify programs, schools, and students that are successful, in addition to those
that need attention and assistance to become successful. Finally, such systems help students and
teachers focus on the curricula and content that must be mastered to be successful in postsecondary
education (SHEEO, 2003).
As state systems for data and accountability progress, the ability to track student progress overtime is
increased, enhancing capture of a wide range of educational influences and offer the promise of
improving student, teacher and school performance. When successful, accountability systems become
more than simply reporting mechanisms by focusing on student performance in relationship to criteria
established by the state and providing a common rubric for evaluating student and school performance.
Success in education can become widespread only if the entire educational system - from early
childhood, elementary, high school, and college - is geared toward preparing and enabling students to
become successful learners and workers.
The central goal in P-16 systems is to raise student achievement by getting students ready for school,
raising standards, conducting appropriate assessments, improving teacher quality, and generally
smoothing student transitions from one level of learning to the next (Rainwater and Venezia, 2003).
The level of interest in statewide P-16 systems comes from both Federal and state agencies. Interest
is stimulated by a labor market which places a high premium on knowledge and skills along with the
appetite of young Americans for postsecondary education. In 2007, the Data Quality Campaign found
22 states have the ability to match student records between P-12 and postsecondary systems (Table 1).
NBH Solutions, LLC Page 3
P-16 Data Sharing Systems…What is Being Done and What is Needed
Table 1
State of the Nation:
Survey Results on Matching Student Records
Between P–12 and Postsecondary Systems
Total Yes
Total Plan
NBH Solutions, LLC Page 4
P-16 Data Sharing Systems…What is Being Done and What is Needed
Sources: Data Quality Campaign, 2007 NCEA Survey of State P–12 Data Collection Issues Related to Longitudinal Analysis; National Center for Higher
Education Management Systems, 2006, comprehensive inventory of state-level SUR capacity; Achieve, Inc., Closing the Expectations Gap 2007
Policymakers want to determine the return gained from the investment of public dollars in P-16
educational institutions. Concerns about efficiency, equity, and cost effectiveness are driving an
increased demand for accountability coupled with accurate reporting of student outcomes from
community and four year colleges. Some of the questions from policymakers include (Data Quality
Campaign, 2007):
Are students succeeding in the P-12, public colleges and universities at an acceptable rate?
Are states getting reasonable return from their support of students who concentrate in certain
disciplines?
Are some ethnic/geographic populations doing better or worse than others in terms of
educational attainments?
Are P-12 and public two year colleges preparing students adequately for success in the upper
division of a four year institution – or are they falling short in some areas?
Are workforce programs at the state‘s higher education institutions turning out graduates who
stay in state and working in their chosen field?
Federal Interest in Student Tracking Systems
Run by the National Center for Educational Statistics (NCES) under the direction of the U.S.
Department of Education, the Integrated Postsecondary Education Data System (IPEDS) is the core
postsecondary education data collection program. Currently IPEDS captures the experiences of 1st time,
full-time students who stay in a single college or university for their undergraduate education. With
IPEDS, it is difficult to assess the curriculum or determine how well students have been educated, given
increasing numbers of nontraditional students and the mobility of students.
Education Secretary Margaret Spellings has proposed a national data tracking system, calling for
greater accountability by colleges and universities through the creation of a national database to track
how well students learn. This unit record system would track the progress of individual students over
time to better assess and compare the educational performance of institutions. The proposed system
would capture new dimensions of postsecondary education by tracking students across institutions,
providing unduplicated national headcounts, and computing net prices while accounting for student
characteristics and enrollment patterns (Dorn, 2006).
Despite increased federal reporting as required in several reauthorizations of the Higher Education
Act (including the 1992 HEA Amendments that created the National Commission on College Costs),
Students Right-to-Know legislation, and other attempts, such as the Graduation Rates Survey, insight
NBH Solutions, LLC Page 5
P-16 Data Sharing Systems…What is Being Done and What is Needed
into student enrollment and the costs of education is elusive. It is difficult, if not impossible, to measure
net prices and enrollment patterns. A 1998 redesign of IPEDS led to insignificant improvement and
prompted a congressional appeal from the American Council of Education (ACE), the Association of
State Colleges and Universities (AASCU), and the State Higher Education Executive Officers (SHEEO)
calling for feasibility study of a data system, based from individual student unit record (UR) data, to
provide data on enrollment patterns and outcome measures such as institutional persistence, completion
rates, time to degree, along with detailed information on student aid and accurate calculations of the net
price of education. The call for a feasibility study reflected a renewed interest in a UR system at the
federal level and resulted from several trends in postsecondary education during the 1990‘s (NCES,
2005) including:
Annual price increases at postsecondary institutions that have exceeded increases in inflation
indexes such as the Consumer Price Index;
Policy concerns about the impact of price increases on consumers and on student aid programs;
A growing congressional interest in holding institutions accountable for outcomes, starting with
graduation rates for student athletes and campus crime reporting;
A demand for better and more timely data to inform policymaking and consumer choices; and
The desire of many postsecondary institutions for more accurate measures of net price and
graduation rates, especially which take into account institutional mission and student mobility.
The resulting March 2005 report by NCES examined the feasibility of implementing a UR system to
replace IPEDS. If implemented, this new UR system would collect individually identifiable data
through files that are submitted electronically by institutions. Files would then be used to calculate
institutional summary totals for each school, with information about enrollment, completions, graduation
rates, financial aid, and price.
Four types of files would be submitted.
Header files to provide individually identifiable information such as name, Social Security
Number, date of birth, address, race/ethnicity, and gender
Enrollment/term files would include program information, number of courses, credits attempted,
major field of study, start and end dates, and attendance status.
Completions files with information on degree completions.
Financial aid files with information on aid received from federal, state, and institutional sources
and the cost of education.
At present, many Federal and state organizations maintain student UR systems. NCES conducts
surveys of postsecondary students (such as the National Postsecondary Student Aid Study) in which UR
information is collected for each student in the sample. The National Student Loan Data System
NBH Solutions, LLC Page 6
P-16 Data Sharing Systems…What is Being Done and What is Needed
compiles information on recipients of federal student loans. The Office of Postsecondary Education
collects student-level information on the recipients of specific program funds, such as GEAR UP,
Upward Bound, and Talent Search. The National Science Foundation conducts a Survey of Earned
Doctorates. The Internal Revenue Service requires colleges and universities to annually submit
individually identifiable student data on tuition and related expenses and scholarships/grants. More
recently, the Department of Homeland Security (DHS) has created the Student and Exchange Visitor
Information System (SEVIS) in order to maintain information on nonimmigrant students and exchange
visitors from the time they receive their visa documents until they complete their program.
State Interest in Post-secondary Student Tracking Systems
At the state level, initiatives to understand student retention in the 1980s gained visibility when
Federal legislation such as the Student Right-to-Know and the Campus Security Act of 1990 passed.
While there were early efforts (Ewell, 1995), UR systems began to evolve in the states during the early
1990s. States looked for a means to provide accurate information for policy formation and management
decisions at the district and school levels. Longitudinal systems emerged to provide information about
student growth over time with links to teachers, programs, and schools that those students. Given the
capacity to study the educational experience over time, these UR systems allow district level
information to be compiled at both the local and state levels to study student outcomes, retention and
school performance.
The full potential of statewide databases is not realized until information can be linked together
using common student identifiers – that is, until data sharing capabilities are developed. Building on
existing term-by-term student databases, many states went one step further and began to track the
academic progress of individual students. Statewide tracking systems of this kind have two major
advantages over institutional tracking systems. First, state-level systems are more efficient; rather than
every institution in a state developing its own tracking system, information can be analyzed centrally
and provided back to institutions. Second, and more importantly, these systems would allow tracking
across institutions, providing more complete information on student outcomes: which students actually
drop out and which students transfer and later graduate (Russell, 1995).
According to a 2006 inventory, there are 47 higher education UR databases in 39 states. Most
contain records only of students enrolled in public postsecondary institutions; however, some also
include data on students enrolled in independent colleges, and more states are considering moving in this
NBH Solutions, LLC Page 7
P-16 Data Sharing Systems…What is Being Done and What is Needed
direction. Collectively, such systems contain basic information on 73 percent of the students enrolled
nationwide in colleges and universities. All 47 UR systems have been in place at least long enough to
track students to the point of earning a degree (six years), and many have much longer histories. In
addition, all have sufficiently common data content and structures to support research on education
pipeline issues such as (Russell and Chisholm, 1995):
Retention and completion rates by gender and race/ethnicity
Patterns of attendance at multiple institutions
Progression from remedial coursework to collegiate study.
Table 2 summarizes twelve common uses of UR systems, in order of frequency mentioned by survey
respondents.
Table 2
Use of Statewide Student Data Bases
Use of UR System Number of States
Persistence, completion, and time-to-degree studies 33
Student transfer studies 32
Studies of minority students 32
Enrollment projections 25
IPEDS reporting (fall enrollment & completions 24
K-12 feedback reports 23
Remedial education studies 22
―Report card‖/accountability reporting 21
Student Right-to-Know Reporting(intended use) 20
Financial aid studies 18
Studies of admissions standards 17
Vocational-technical reporting 11
Source: (Russell and Chisholm, 1995)
Persistence, completion, and time-to-degree studies, student transfer studies, and studies of
minority students illustrate the added value of multi-institutional student tracking capabilities.
Many statewide data systems are now beginning to address other important policy issues such as
what happened before the student entered college and what happens after he or she leaves. One new and
rapidly growing phenomenon is the development of linkages between state higher education databases
and those maintained by other state, federal, and private agencies and organizations. Table 3
summarizes the states currently engaged in this type of activity.
NBH Solutions, LLC Page 8
P-16 Data Sharing Systems…What is Being Done and What is Needed
Table 3
Interagency Data Sharing
Data Sharing Arrangements in the
Data Sharing Arrangements in
Type of Agency Discussion/Planning Stage
Place (# of States)
(# of States)
State department of education (K-12) 10 10
Postsecondary agencies in other
3 6
states
State employment records 11 9
State corrections 3 3
Federal training and employment
3 4
records
Military 5 2
Private sector employers 2 3
Source: (Russell and Chisholm, 1995)
According to the Data Quality Campaign (2007) complete state longitudinal data systems include ten
essential elements:
1. A unique statewide student identifier
2. Student-level enrollment, demographic and program participation information
3. Ability to match individual students‘ test records from year to year to measure academic growth
4. Information on untested students
5. A teacher identifier system with the ability to match teachers to students
6. Student-level transcript information, including information on courses completed and grades
earned
7. Student-level college readiness test scores
8. Student-level graduation and dropout data
9. Ability to match student records between the Pre K-12 and higher education
10. A state data audit on quality, validity, and reliability
In developing tracking databases, a series of decisions must be made (Ewell, 1987b):
Who will be tracked? All students? Credit students who have initiated a matriculation process?
Credit students who have completed a minimum number of courses? In brief, it must be decided.
The college must decide if all students will be or only those who can be categorized as
matriculated students.
How long or over how many terms should the students be tracked? Much depends on the
mortality rate of the student cohort. Ewell suggests that a good rule of thumb is to maintain a
tracking period of sufficient length to determine the fates of at least 90 percent of the students in
the cohort. Tracking outcomes after students leave will require a longer period of operation than
tracking persistence and attainment while the students are enrolled.
How often will new cohorts be tracked? Ewell points out that while it may be possible to begin a
new tracking cohort each term, it may not be practical or necessary. Ewell reports relatively
little variation in results across cohorts from successive years. As a result; some UR systems
establish new cohorts on a periodic basis, for example, once every three years.
NBH Solutions, LLC Page 9
P-16 Data Sharing Systems…What is Being Done and What is Needed
What data elements will be tracked? The answer to this question will likely be a compromise
between what researchers would like to know and what data are available. For example, the
educators might want to compare the outcomes of students with deferring socioeconomic
backgrounds but finding a proxy measure for SES is difficult.
A final question, discussed by Ewell, Parker, and Jones (1988) focuses on data processing: how
will data from various entities be merged into a single cohort tracking file? Much of the needed
data may be in disparate systems run on incompatible computers and file formats. A major
challenge is to find ways of pulling these data together; however more recent technical
developments make synthesis of such data easier.
Many states have a growing interest in improved information on student flow and success. In
California, the State Assembly directed the California Postsecondary Education Commission (CPEC) to
―develop a feasibility plan for a study to provide comprehensive information about factors which affect
students‘ progress through California‘s educational system, from elementary school through
postgraduate education‖ (California Education Code, Chapter 4, Section 99172). The CPEC plan
proposed an ambitious modified longitudinal study using samples of students at different points in the
educational pipeline. Going far beyond the usual questions of how many students are enrolled and how
many degrees are awarded, the proposed study would assess access to education, student progress
through the educational pipeline, and subsequent student success in the job market. California has made
great strides in developing a statewide management information system that will, among other things,
link student demographic data with course data, thus allowing policymakers to track student progress
over time.
Johnson County Community College in Overland Park, Kansas, has played a central role in initiating
and coordinating longitudinal studies of the state‘s community college students. Doucette and Teeter
(1985) describe a 1984 study examining student mobility between the 19 community colleges and six
state universities in Kansas. The study was conducted cooperatively between those institutions and
involved three components: (1) coordinated analyses of student data bases to determine the demographic
and academic characteristics of community college transfer students; (2) a survey of former community
college students enrolled in Kansas universities; and (3) a retrospective longitudinal examination of
selected groups of native university and community college transfer students. Doucette and Teeter warn
of study limitations, however, created by a lack of operational data definitions and differences in the
student information databases.
The Illinois Community College Board (ICCB) has conducted several statewide longitudinal studies.
Examples include the Statewide Occupational Student Follow-up Study, involving a four-year
NBH Solutions, LLC Page
10
P-16 Data Sharing Systems…What is Being Done and What is Needed
longitudinal analysis of students entering community college occupational programs, and the Illinois
Community College Board Transfer Study, a five-year longitudinal analysis conducted to follow the
progress of 9,757 students who transferred to Illinois senior institutions. The Illinois studies
demonstrate that, with appropriate leadership at the state level, student performance data recorded in the
files of various colleges can be merged and analyzed to draw pictures of student flow and progress.
While many states collect student follow-up data gathered by individual institutions and thus provide
some statewide measure of transfer, graduate job placement, and other student outcomes, most states
have not followed through on demands for improved information on student success with the requisite
support for rigorous longitudinal research. States can do little to improve the picture of student flow and
outcomes as long as individual institutions are unable or unwilling to collect the requisite data. States
have developed uniform approaches for institutional use when collecting data [Michigan Student
Information System (MSIS), the Texas Student Information System (TSIS), Texas Longitudinal
Evaluation Student Tracking and Reporting (LONESTAR)].
Other efforts provide a sporadic picture of student success. These include follow-up studies
(Palmer, 1985), surveys of the goals of entering students (Community College of Philadelphia, Glendale
Community College in Arizona, Kirkwood Community College in Iowa, San Francisco Community
College District, and Broome Community College); however, few institutions have made longitudinal
analyses a routine part of their institutional research effort. It is one thing to ask ―How many students
were enrolled in fall 2004?‖ It is quite another to ask ―What happened to these students during the past
three years?‖ Longitudinal analyses of student flow, whether they are conducted retrospectively through
an examination of student transcripts or (more rarely) progressively through a sequential term-by-term
analysis of student progress, pose a more formidable research task than cross-sectional analyses.
Published longitudinal studies usually track student persistence through the institution without
looking at outcomes after graduation or without drawing a link between outcomes and student goals.
For example, Miami-Dade Community College draws upon its student information systems to track the
progress of entering student cohorts, particularly for special populations (Palmer, 1990). Underlying
these studies is Miami-Dade‘s commitment to the development of student success indicators that
bespeak the community college‘s nontraditional student body. Morris and Losak (1986) illustrate the
use of these measures, tracking the three-year progress of full-time, degree-seeking students entering
and argue for new definitions of student success. Longitudinal studies at other colleges found indicators
NBH Solutions, LLC Page
11
P-16 Data Sharing Systems…What is Being Done and What is Needed
of student persistence and progress and show the varied enrollment patterns of non-traditional students
(Lucas 1986).
One way to make longitudinal studies more manageable is to specify in advance what data, or
indicators, the longitudinal database will yield. While there is a temptation to measure student flow
against any number of academic, demographic, or socioeconomic variables, limitations in financial
resources and staff time may dictate that only selected variables be examined. The process of building a
tracking system, then, ideally begins with the question ―What indictors of student flow will we use to
assess effectiveness?‖ As the Morris and Losak (1986) study at Miami-Dade Community College
indicates the selection of these indicators is a difficult process, requiring researchers to specify variables
that yield accurate indicators of student success without presenting an exaggerated and self-serving
picture of the institution.
Several student tracking models do exist and are well described in the literature (Wilkinson,
1985; Voorhees and Hart, 1989; Ewell, Parker, and Jones, 1988). Wilkinson (1985) identifies a division
of labor related to tracking systems involving:
A steering/advisory committee designed to assure broad participation and to ensure that the data
collected is of use for educational improvement and policymaking
Institutional research staff to provide direct support to the advisory committee
Staff charged with the data collection task and other day-to-day operations.
Table 4 details a simple student tracking system (Wilkinson, 1985). While Wilkinson‘s system may
be too simple for non-traditional educational institutions, the three-part division of labor seems
promising.
NBH Solutions, LLC Page
12
P-16 Data Sharing Systems…What is Being Done and What is Needed
Table 4
Student Tracking Flow Chart
Entering
Applicants Accepted freshmen Sophomores
1 2
1
Drop-outs/
Rejected No-Shows Non-returning Graduates
5 3
1. Entering Student Survey
2. Student Opinion Survey Alumni
3. Student Opinion Survey 4
4. Alumni Survey
5. Withdrawing/Non-
returning Survey
Source: Wilkinson, 1985
Voorhees and Hart (1989) call longitudinal systems a challenging task as even a simple model
requires voluminous amounts of data.
Perhaps the most widely recognized student tracking system is the LONESTAR system, developed
for the Texas community colleges by the National Center for Higher Education Management System
(NCHEMS). Exceptionally detailed, LONESTAR (Longitudinal Student Tracking and Reporting) was
designed to meet several objectives: to provide management information improvement; to provide a
framework for the uniform reporting of institutional effectiveness data; and to evaluate and report on the
effectiveness of the remediation function of community colleges as a primary access point to higher
education in the state (Ewell, Parker, and Jones, 1988). Implementing LONESTAR involved the
development of a:
Common methodology for identifying the types of students to be included in the system;
Common procedures for determining how individual tracking records were to be constructed,
handled, and maintained;
List of commonly defined data elements that all institutions were to include in the system;
Set of optional data elements that institutions might include at their discretion;
Set of recommended reports for local institutional use and for submitting information to the
Coordinating Board; and
NBH Solutions, LLC Page
13
P-16 Data Sharing Systems…What is Being Done and What is Needed
Set of recommended procedures for installing and operating the system locally using hardware
and software.
While somewhat dated in terms of technology, Establishing a Longitudinal Student Tracking
System: An Implementation Handbook (Ewell, Parker, and Jones, 1988) is the single most
comprehensive technical resource available on longitudinal student tracking at community colleges,
detailing both recommended data elements, and outlining methodologies for constructing the database
itself. The data elements summarized in Table 5 allow researchers to compare term-by-term student
progress against many attributes, including age ethnicity, physical disability, economic
disadvantagement, academic ability, and educational background. Besides specifying the data elements,
the LONESTAR model provides operational definitions for each, drawing upon several sources,
including the Texas Educational Data System; the U.S. Department of Education‘s Integrated
Postsecondary Education Data System; the Higher Education General Information Survey; the Council
on Postsecondary Accreditation; the National Center for Education Statistics; and the National Center
for Higher Education Management Systems.
Table 5
Selected LONESTAR Data Elements
Follow-Up Indicators for
Attributes Term-by-Term Indicators
Graduates/Leavers (All Optional)
Student Identification No. Term Identification Transcripts Requested
Demographics: Credit Hours Attempted Transferred to Another Institution
Gender Credit Hours Attempted for which Credit Hours Accepted by transfer
Date of birth Grades were received Institution
Ethnicity Credit Hours Successfully Completed First-term Enrolled in Transfer
Citizenship Grade Point Average Institution
Residence (in-state, out-of-state, etc.) Credit Hours Attempted for Non- Program Enrolled at Transfer
Physical/Learning Disability remedial Classes Institution
Economic/Academic Disadvantagement Credit Hours for Which Grades were First Degree Awarded
Current Employment Status Received for Non-remedial Classes Employment Status at Time of
Educational Background Credit Hours Successfully Completed Follow-up
last high school attended for Non-remedial Classes Employment in Field for Which
type of high school certificate Grade Point Average for Non-remedial Trained
awarded Classes Average Hourly Wage
date of high school graduation Academic Standing (good, probation, Employer Ratings
high school grade point average suspension) technical knowledge
last college attended Remediation Attempted work attitude
pervious college-level academic reading work quality
experience writing
Remediation Status at Time of Entry computation
reading Remediation Attained
writing reading
computation writing
English proficiency computation
Enrollment Status Programs Enrolled In
first term of academic history Degree/Certificate Awarded
admission status (full, provisional) GED Activity in Term
basis of admission (high school ESL Activity in Term
graduate, individual approval, etc.) Term Non-credit Activity
NBH Solutions, LLC Page
14
P-16 Data Sharing Systems…What is Being Done and What is Needed
Financial Aid Status
Time of Attendance (day, evening, etc.)
Location of Instruction (on-campus,
off-campus)
Initial Program at Time of Entry
Program Track (vocational, academic,
unclassified)
Student Objective (primary reason for
enrolling)
Intended Duration (one term only,
two term, etc.)
Term of Enrollment in First College-
Level English Class
Performance in First College-Level
English Class
Term of Enrollment in First College-
Level Math Class
Performance in First College-level
Math Class
The LONESTAR implementation manual specifies a three-step process, illustrated in Table 6,
for constructing the database. The process presupposes that all student records include a common
identification number for each individual, thus making it possible to merge student files from different
databases. Required data are extracted from existing student records and placed into source files. Each
source file contains the data elements specific to a given portion of the student longitudinal enrollment
record. Furthermore, each source file is generally keyed to a single location in the institution‘s master
student record system (Ewell, Parker, and Jones, 1988.) The LONESTAR system assumes that updates
will occur at two points in each term: (1) a ―beginning-of-term‖ update, just after an official reporting or
census date, which captures such elements as credit hours attempted, student‘s academic standing, and
student‘s program of study; and (2) an ―end-of-term‖ update containing data for credit hours
successfully completed, grade point average earned, etc. The final step, often omitted from other
discussions of student tracking, involves using the cohort files to generate required reports and ad hoc
analyses. These reports take the form of cross-tabulations that compare indicators of academic progress
for students with different attributes.
NBH Solutions, LLC Page
15
P-16 Data Sharing Systems…What is Being Done and What is Needed
Table 6
Basic Procedures for Creating Longitudinal Tracking Files
I. Extract Student data from College Files and Other Sources
Student Assessment Feedback Research
Records Center from 4-year Office
Colleges
II. Download Fields and
Recode as Necessary into
Source File
SSN Cohort
Student A F 87
Student B F 88
Student C F 88
Student D S 89
0
0
Student ZZZ F 89
III. Add to Cohort Files
Term since Entry
Cohort 1 2 3 4 5
5
Students Entering in Fall ‘87 X
Students Entering in Spring ‘88 X
Students Entering in Fall ‘88 X
Students Entering in Spring ‘89 X
Students Entering in Fall ‘89 X
Source: Palmer, 1990
NBH Solutions, LLC Page
16
P-16 Data Sharing Systems…What is Being Done and What is Needed
In 1988, drawing heavily on LONESTAR, the American Association of
Community and Junior Colleges convened to suggest the components of a model student
tracking system for community, technical, and junior colleges (Palmer, 1990) . The
members drew heavily upon the LONESTAR effort, and created a system with
resemblance to the LONESTAR system.
Examination of these designs reveals tracking systems as promising research tools
that replace ad hoc longitudinal studies with a centralized and on-going data collection
system. Yet the success of such systems will depend on the ability of institutions and
agencies to implement and use them. Several inter-related issues are central to
implementation (Palmer, 1990).
Merging Data Files is a major task requiring the organization of data around a
common student identification number so that data from different sources can be
compiled.
Centralization is key to coordinated collection of data.
Tying a central system to program and institution improvement makes a tracking
system effective
State Interest in Post-secondary Student Tracking Systems
According to the National Center for Education Accountability‘s 2006 Survey of
State Data Collection, 44 states have compiled sufficient student information on public
P-12 students to generate useful and informative analysis (www.DataQualityCampaign.org).
Many of these states have invested in their P-12 information systems to meet the
requirements of the No Child Left Behind Act. Each state‘s database contains
information on basic enrollment, and 43 states report having the data system in place to
match student records from year to year to measure academic growth.
California Longitudinal Pupil Achievement Data System (CALPADS) will
become the K-12 student-level data system for California. Under development, it is not
expected to be completed until 2009. CALPADS was authorized by California SB 1453
(2002) and SB 257 (2003) mainly to meet the federal, state, district, and school reporting
requirements under the No Child Left Behind Act of 2001 (Vernez, Krop, Vuollo,
Hansen, 2006). The authorizing legislation calls for a database to access the long-term
value of educational investments and programs and provide a research base for improving
pupil performance. The purposes of the system are to provide (1) a way to evaluate
NBH Solutions, LLC
Page 17
P-16 Data Sharing Systems…What is Being Done and What is Needed
California‘s educational progress and investments over time, (2) information to LEAs
that they use to improve pupil achievement, and (3) an efficient, flexible, and secure way
to maintain longitudinal pupil-level data statewide (Vernez, Krop, Vuollo, and Hansen,
2006).
In preparation for CALPADS, in school year 2005-06 every student in California‘s
public K-12 education system was assigned a unique, yet non-identifiable SSID that will
follow the student throughout his or her enrollment in the system. CALPADS will
contain only non-identifiable student data. Student identifiable information will be
available only at local education agency level will be able to match an SSID with a
student. It is expected that this data system will be accessible to (Vernez, Krop, Vuollo,
and Hansen, 2006):
Educational agencies in California
Legislative and executive overseers
Researchers from established research organizations, on approval of the state
Department of Education
The public will not have direct access to CALPADS.
The system enables data sharing among P-12 schools, community colleges and
universities. The goal is to help educators understand performance and transitions,
improve instruction, and increase student success by addressing questions such as:
How do our students do when they move on?
Were they well prepared?
Are changes in curriculum necessary to help others?
How many students earned degrees at the next level?
How many students earned degrees at the next level?
How long did it take?
Potential uses of the data include program review, cohort tracking and identification
of successful course-taking patterns. Information on student cohort is provided to cross-
sector, discipline-based faculty to examine curricula and instructional practices.
Recommendations for improvement are provided to the appropriate agency, with a goal
of developing more seamless curriculum and improved instructional strategies.
The Florida Department of Education (FDOE) oversees 67 county-wide public school
districts, encompassing almost 4,000 schools, in which over 2.6 million students were
NBH Solutions, LLC
Page 18
P-16 Data Sharing Systems…What is Being Done and What is Needed
enrolled in the fall of the 2005-06 school year. Florida has a long history of collecting a
multitude of data on its public education system and was one of the country‘s pioneers in
collecting student-level data.
Early efforts to evaluate and hold educational programs accountable in Florida began
with legislation passed in 1968 instructing the Department of Education to improve
educational effectiveness. The Florida Statewide Assessment Program was created as a
result of the 1971 Educational Accountability Act. An important element in the state's
accountability effort, the program was designed to assess students' academic strengths
and weaknesses, particularly in the basic skills. Since 1984, accountability for career and
technical education, especially at the postsecondary level, has also been a focus in
Florida. Accountability systems for community colleges and the state university system
have been required by state statute since 1991. The state legislature in Florida has
historically been supportive of implementing and enhancing statewide student
longitudinal data systems for informing and improving public education. Currently, in
every legislative budget a portion of the funding allocated to school districts must be used
for data and information services.
In 1986-87, Florida piloted collecting individual student-level data through the
Florida Information Resource Network (FIRN). In 1990, the FDOE began to use the data
collected through FIRN for reporting on the P-12 education system. In 1988 the Florida
Education and Training Placement Information Program (FETPIP) was implemented.
FETPIP is a data collection system that obtains follow-up information on students after
they exit the P-12 system and includes employment, postsecondary education, military,
public assistance participation, and incarceration data. FETPIP uses data files from P-12,
community colleges, universities, the Florida Department of Labor, and the Florida
Department of Corrections to identify and match students to determine their workforce
placement and earnings. FETPIP findings are used to identify the success of students,
employee earnings and the impact of education offerings on the economy.
An electronic transcript system, the Florida Automated System for Transferring
Educational Records (FASTER) has been in place since 1988-89. By 1994, Florida had
one of the most progressive, comprehensive and efficient systems for transferring student
records in the nation. In 2001, over 900,000 electronic transcripts were exchanged. At
NBH Solutions, LLC
Page 19
P-16 Data Sharing Systems…What is Being Done and What is Needed
about the same time that Florida's FASTER system was going into production, Florida
began to work with other states to develop a nationwide student record transfer system,
now known as SPEEDE/ExPRESS (Standardization of Postsecondary Education
Electronic Data Exchange/Exchange of Permanent Records Electronically for Students
and Schools). The Florida Comprehensive Assessment Test (FCAT) is part of Florida‘s
overall plan to increase student achievement by implementing higher standards. The
FCAT, administered to students in Grades 3-11, contains two basic components:
criterion-referenced tests (CRT), measuring selected benchmarks in Mathematics,
Reading, Science, and Writing from the Sunshine State Standards (SSS); and norm-
referenced tests (NRT) in Reading and Mathematics, measuring individual student
performance against national norms. Since 2002, the Florida Education Data Warehouse
(EDW) has provided a single repository of data extracted from multiple sources available
at the state level on students, education facilities, curriculum, and instructional staff in the
P-20 public education system. The EDW allows longitudinal data analysis at the student
and staff levels from 1995-96 forward. Student level data include demographics,
enrollment, course completion, assessment results, financial aid, and employment.
Future plans include collecting SAT, ACT, and advanced placement data, and obtaining
information on private school students (Data Quality Campaign, 2006).
A new learning and teaching environment tool called Sunshine Connections is under
development to provide information to educators, administrators, parents, and students.
Teachers will be provided with interactive access to classroom management tools, student
performance data, and interactive capabilities with other teachers, curricular materials,
and professional development opportunities. Benefits include (Data Quality Campaign,
2007):
Increases in accuracy and efficiencies in data collection have been realized over
time.
There are efficiencies inherent in state department efforts to define elements such
as course numbers so that districts do not have to develop these. Standardization
allows information to transfer from district to district and to higher education with
shared understanding of how elements are defined.
The state has the capability to cross reference data files submitted by the districts
and identify errors and anomalies, a process not all districts can do locally.
Reports provided to districts allow errors to be identified and corrected before
final submission.
NBH Solutions, LLC
Page 20
P-16 Data Sharing Systems…What is Being Done and What is Needed
The FDOE provides data to federal offices (such as the Office for Civil Rights)
instead of having the districts submit the data directly as is done in most states.
Uses include:
Data have been used for accountability and reporting in Florida for many years.
Data are used for reporting almost immediately after the 2- to 3-week submission
period.
Districts are provided files containing data on their own students who are included
in calculations for accountability purposes by the state.
There is a very high use of data by teachers and administrators.
FDOE staff work with legislative staff to ensure or strengthen understanding of
the data used by legislators.
The Florida Office of Program Policy Analysis and Government Accountability
uses student level data to examine performance in various areas in the context of
costs of education.
The FDOE uses a quid pro quo system for negotiating with researchers requesting
education data. In exchange for access to the data, the research question either has
to be one of interest to the state, or fees are assessed for acquiring the data. The
FDOE requires that agency staff be allowed to review research reports before they
are released.
Florida has learned many lessons when it comes to systems design:
Start with an effort sufficiently focused and useful to have an early win, then
build on it rather than having a scope that is too broad to be manageable.
Do not wait until things are perfect to start. Plan on mid-course corrections and
phased implementation.
Know that all the issues associated with these efforts (matching records across
systems, confidentiality, demonstrating viability, selling the products,
maintaining the quality of operations, watching for pitfalls and opportunities)
will need to be addressed continually.
Identify all the information you want to obtain at the beginning and plan on
midcourse corrections as dynamics change. Put it in a format you can change
because it will never remain static.
Expect to spend a year in discovery for an effort like a data warehouse and know
that no shrink-wrapped answer is available. Be aware that it will take staff time
for development and contract monitoring activities.
Keep ultimate goals in sight to maintain expectations of where you need to be.
Transparency of procedures and processes is helpful, so that all parties involved
can see the benefits of better quality data.
Document systems and how and why they were designed.
In order to improve the quality of data and eliminate confusion, determine one
source of data for all reports and evaluations.
NBH Solutions, LLC
Page 21
P-16 Data Sharing Systems…What is Being Done and What is Needed
Be aware that once data are linked to other areas, perhaps in ways that were not
originally intended, this will change the scope of how the data are collected and
reported and can have unintended consequences.
Florida‘s K-20 Education Data Warehouse provides another example of how a state
is building the infrastructure to align kindergarten through postsecondary education
policies and practices. The warehouse provides stakeholders in public education
including administrators, educators, parents, students, state leadership and professional
organizations with information on Florida‗s public school students from kindergarten
through their graduate-level studies, as well as some workforce information (L‘Orange,
2007). This is accomplished by extracting and integrating data from existing systems,
including a robust K–12 data system that has been in place for more than 10 years, the
community college system, the university system, the Florida workforce development
information system and student assessment files. The result is a single repository of data
on all students served by the public K–20 education system, as well as information on
facilities, curriculum, and staff.
The system provides (L‘Orange and Ewell, 2007):
K–20 public education data integration;
Longitudinal analysis;
A student-centered perspective;
Historical and current data;
Confidentiality (personally identifiable information is removed); and
State-of-the-art analytical capabilities.
Other state P-12 UR systems track students through their education. Kansas
Individual Data on Students (KIDS) (http://kids.ksde.org) is a student-level data collection
system which consists of two different systems (Gosa, 2007):
The Assignment System is based on a vendor software package. The software is
used to collect a set of core data elements for every student in Kansas accredited
preK-12 schools; assign a unique randomly generated state number to each of
these students; and track the students as he/she moves between Kansas public
schools.
The Collection System was developed by KSDE to work in conjunction with the
Assignment System. The Collection System collects additional data elements on
every student in Kansas. This additional data is used for district funding, student
assessments, school accountability, and state and federal reporting requirements
such as enrollment, graduation, attendance, and truancy information.
NBH Solutions, LLC
Page 22
P-16 Data Sharing Systems…What is Being Done and What is Needed
KIDS collects 84 individual data elements in the following categories (Gosa, 2007):
Demographics (first name, last name, date of birth, gender, race/ethnicity, etc.)
Enrollment Data (Building, Funding Building, School/District/State Entry Date,
Exit Information, etc.)
Attendance Data (cumulative days in attendance, cumulative days in membership,
truancy, etc.)
Program/Services Participation Data- (National School Lunch Program, IEP, Title
1, etc.)
Assessment Data (test subject, test type, grouping indicators, etc.)
KIDS data is used for State Funding, Federal Funding (ESL, Title 1, etc.), Student
Assessments, and School Accountability.
To supplement KIDS, Kansas has embarked on a three year plan to integrate
education data in an enterprise data warehouse using business intelligence tools. The
major components include: Enterprise Data Warehouse, Master Data Management,
Enterprise Metadata Repository and Data Marts. This system is scheduled for
completion in 2009. Data access will be available to stakeholders via:
Data Marts (reports, business intelligence, research, data mining, etc.)
Metadata (self service and enhanced stakeholder understanding of data)
Training (specific to stakeholder needs, includes use of Metadata and use of
Business intelligence solutions and focus on how to use the data)
This system will be used link P-12 to Postsecondary education through a unique
student ID number.
Linking to Employment
Unemployment insurance wage and related files are valuable resources for all sectors
of higher education. Researchers can use these files to describe and compare students‘
workplace status before, during, and after enrollment. Since the 80‘s state governments
have been collecting information from employers about employment and earnings.
State‘s department of labor or employment security (ES) collects and maintains this
unemployment insurance (UI) wage file. The UI wage file contains data for each
employed worker in the state, on such matters as total earnings, employment status, and
industry of employment. Given these contents, the UI wage file also has considerable
potential value for researchers in determining the labor force status of students and
former students.
NBH Solutions, LLC
Page 23
P-16 Data Sharing Systems…What is Being Done and What is Needed
States collect UI wage records under federal guidelines, but these allow states some
choice regarding the specific data elements collected. Florida, which has the nation‘s
most comprehensive data-linking system, matches data from postsecondary institutional
enrollment records with military enlistment records from the U.S. Department of
Defense, postal career service data from the U.S. Postal service, employment records for
federal career employees from the Federal Office of Personal Management, inmate
records contained in the state‘s Department of Corrections database, and records covering
participation in the major welfare payment programs available from the State Public
Assistance Agency (Seppanen, 1995).
Most states that have engaged in administrative data-linking to determine labor force
status have done so for two reasons: research and accountability. The University of
Maryland, for example, used the UI match to determine the economic value of university
education to the state. A Washington State University researcher used UI data to
compare returns on investment for community college transfers versus those who begin
their education at WSU. The Washington State Board for Community and Technical
Colleges links college data files with other administrative records from the Employment
Security Department. Through the assistance of the Workforce Training and Education
Coordinating Board, it creates partnerships with the managers of the Job Training
Partnership Act and other programs to defray the cost of data linking. The partnership
results in the Data Linking for Outcomes Assessment (DLOA) program. DLOA contains
one record for each student for every three-month period with information on firms for
which the student worked, and colleges attended for a six-year time frame. Uniquely,
DLOA contains the number of hours a student has worked in a quarter (Seppanen, 1998).
Florida gathers performance measurements from several systems broken into a
common three-tier measurement system. Measurements are applied across all workforce
education and development programs at progressively detailed levels. Tier 1 measures
outcomes for all workforce education system-wide. Tier 2 looks at program level
measurements such as postsecondary education. Tier 3 examines program operations and
management and outcome measurements demanded by federal and state agencies
(Pfeiffer, 1994). Also, Florida uses such data to study the rate-of-return by educational
sector and to look at recidivism rates following prison-based education.
NBH Solutions, LLC
Page 24
P-16 Data Sharing Systems…What is Being Done and What is Needed
At least twenty states now link higher education enrollment data with available UI
wage files; leaders include Colorado, Florida, Illinois, Texas, etc. File linking started in
the mid 70‘s related to accountability reporting for the Cooperative Education and
training Act federal job training program. Many two-year college systems use such
methods to replace traditional vocational education follow-up surveys. With a survey,
the basic assumption is that the student‘s employment status after college is a useful
indicator of the quality of education received. The data-linking process is simply a cost-
effective way to determine subsequent employment.
A number of benefits have led states to increased use of administrative data-links for
student tracking. The primary benefit is the low cost of such procedures when compared
to surveys. The major costs of data-links are the fixed costs associated with initial
programming and of management and analytical staff time. Data-links also eliminate
survey non-response bias. Lastly, administrative records allow researchers to examine
directly the complex dynamics of employment (Seppanen, 1995).
Since higher education is financed partly by state funds, it is an investment in human
capital that at the present time is perhaps not being fully realized by states. An example
of this is out-migration of graduates due to the lack of suitable employment opportunities.
Wyoming and the University of Wyoming put together a study that focuses on graduates
of UW and their relationship to the Wyoming work-force by identifying how many
students are showing-up in the Wyoming work-force and in what job classifications they
are finding work in (Butler, 1995).
The population under the study was 11 years of University of Wyoming graduates
earning a degree from 1983-1993. Information was provided by the UW Office of
Registration and Records and the UW Office of Institutional Analysis. Also a second set
of data was provided by Wyoming‘s Research & Planning Section of the Employment
Resource Division which contained all members of the Unemployment Insurance (UI)
covered work-force in the state of Wyoming from 1992 through 1994.
By matching the UW data set, graduates of UW 1983-1993, against the employment
and wage data for 1992-1994, they were able to determine how many UW graduates of
the study period were employed in the UI covered work-force of the target years. With
NBH Solutions, LLC
Page 25
P-16 Data Sharing Systems…What is Being Done and What is Needed
this information they also could pinpoint the industry in which they worked and their
quarterly wages.
This study concluded that UW graduates are filtering out of the work-force over time.
As this is taking place the loss in human capital could be negatively impacting the entire
economy of the state. This raises serious questions both for the state and the educational
system. Over half, 55.6% of UW‘s graduates from 1983-1993 do not show up in the UI
covered work-force in the years 1992-94.
Butler (1995) notes that some questions of concern to the state, the policymakers, and
the University include:
1. Is Wyoming falling short in creating white collar career positions?
2. Are white collar wages in Wyoming uncompetitive with other states?
3. Is this kind of out-migration true of other states and other universities?
4. Is the low cost of university education at UW being exploited by both residents and non-
residents?
5. What kind of an in-migration of college graduates from other states does Wyoming
experience?
6. Should state boundaries even be a part of the discussion of higher education?
Some other conclusions that Butler discovered were that some majors are more likely
to show up in the employment data than others. Social Work majors are more than twice
as likely to show up in Wyoming as not. While Teaching, Psychology, Sociology,
Nursing, etc. are being found in the state at the greatest frequency. Education, Social
Science, and Humanities majors are overall more likely to appear in the employment data
then are Business, Professional, and ―Hard Science‖ majors. These conclusions also
bring up the questions of how much money is being spent by the state on any given
major, and hence, what kind of return does it get on the investment. Through this study it
has shown the importance for linking student data with employment data to find out the
investment in human capital being made by states.
UI wage files and related data bases are being used effectively for accountability
reporting and research. Effective use of these data files requires a re-examination of
many assumptions about the connection between education and work. Among them are
the linear nature of this relationship and the degree to which any degree of relatedness
can really be established given increasing job mobility and the rapidly changing nature of
workplace skills. All researchers would benefit from further national dialogue on issues
of non-coverage and how to develop better indictors of quality training (Seppanen, 1995).
NBH Solutions, LLC
Page 26
P-16 Data Sharing Systems…What is Being Done and What is Needed
Barriers to P-16/20 Student Tracking Systems
Several barriers inhibit the development of P-16/20 tracking systems (Laird, 2008):
1. Lack of common identifiers.
There is no single means for sharing data between elementary, secondary
and postsecondary levels of education. Some states are developing data
warehouses that allow data from all systems to be linked, analyzed and shared
appropriately.
2. Incompatible systems.
States are surmounting incompatible systems by employing immediate
solutions such as placing the unique K–12 student identifier on high school
transcripts. In other states, postsecondary institutions provide annual feedback
reports to individual high schools on the success of their graduates in their first
year of college-credit coursework, which gives high schools valuable information
for improving the rigor and effectiveness of high school curricula and instruction.
3. State law prohibitions.
Although federal privacy laws place some restrictions on the exchange of
individual records, they do not prohibit states from sharing student records.
Several states have worked out ways to make this exchange possible. Minnesota
and Virginia are recent examples of states pursuing changes to their state laws to
allow these data exchanges.
4. Lack of coordination.
As states increasingly focus on the need to align policies and practices
across elementary, secondary and postsecondary education, there is corresponding
growth in the development of collaborative bodies that span the sectors. Having
these governance structures that define a common vision and the data needed to
achieve these goals is a common characteristic of states that are having success
aligning P–20 data systems. In 2006, the Education Commission of the States
(ECS) found that 30 states were engaged in some kind of P–20 activity. These
initiatives varied widely from major governance changes, such as those in Florida,
to establishing P–20 councils. ECS reported in 2006 that five states (Florida,
NBH Solutions, LLC
Page 27
P-16 Data Sharing Systems…What is Being Done and What is Needed
Indiana, North Carolina, Ohio and Texas) passed legislation to establish a P–20
council or initiative and that three more (Missouri, Nevada and Illinois)
legislatively mandated the creation of such councils since the report‘s release
(Dounay, 2008). Most P–20 councils are formed voluntarily or through executive
order. Delaware created its P–20 Council through executive order and then
enacted legislation to codify it into law (Ed. Com. Of the States, 2006). In 2007,
36 states reported on the DQC survey that they have a statewide group or council
that discusses policies related to P–20 data systems. Six states reported that these
discussions occurred between two separate groups (P–12 and postsecondary), and
29 states reported that a P–20 council addresses alignment within the state (Data
Quality Campaign, 2007). Although the expansion of P–20 councils is
encouraging, these entities must be granted authority over the issues they discuss.
5. Lack of resources.
As stakeholders realize the purpose and potential of data sharing systems,
it is easier to free up resources of time and money. In fact, various state models,
such as Louisiana, show that aligning P–20 data systems does not have to be
expensive, and it does not require waiting to build a data warehouse to start
having P–20 conversations with supporting data. Therefore, state policymakers
— legislators, governors, state board members, higher education officials,
attorneys general and chief state school officers — must work together to ensure
that there is the political will to build these state longitudinal data systems, along
with the resources, legal clarity around privacy issues, and increased capacity
throughout the system to use these data for policymaking and decision making.
6. Funding.
Other related issues revolve around the costs, financial and otherwise, of
creating systems in states where sectors are funded separately. A shared system
requires shared responsibility; successful systems have either created a separate
funding mechanism or explicitly mandated how cost sharing will occur. There
also can be a tendency to provide one-time funds; however, these efforts require
sustained support, which can be a challenge in difficult economic times and is just
NBH Solutions, LLC
Page 28
P-16 Data Sharing Systems…What is Being Done and What is Needed
one of the reasons why addressing long-term state goals at the beginning of a
project is so critical.
Incorporating Quality Teacher Education Programs In New P-16 Systems
Since 1976, national policymakers, state officials, and local educators have struggled
to measure and improve teaching but ―America is still a very long way from realizing that
future‖ (National Commission on Teaching and America‘s Future, 1996). Work in the
states is being supported by a number of national initiatives aimed at reforming the
teaching profession. For example, recent efforts of the National Commission on
Teaching and America's Future (NCTAF) and Teacher Quality Enhancement Program of
the U.S. Department of Education stress quality at each point of a teacher's career – from
recruitment to initial preparation, to the transition of the beginning years of teaching, and
throughout continuing professional development. These national blueprints for achieving
quality in teacher education serve to involve interested states as partners in the design and
implementation of effective strategies and programs.
A review of state strategies aimed at incorporating quality teacher education and
professional development programs in P-16 UR systems led to initiatives in three states -
Georgia, Maryland, and Ohio.
Georgia's P-16 initiative began in 1995 and is a collaborative development of the
Office of School Readiness (OSR), the Department of Education (DOE), the Department
of Technical and Adult Education (DTAE), and the University System of Georgia (USG)
(Zimpher, 1999). Beyond the broad-based involvement of these agencies, individuals
were involved representing P-12 and postsecondary educators, school board members,
youth advocacy organizations, community members, and legislative and business leaders.
Georgia's P-16 initiative has identified five goals:
To improve student achievement from preschool through postsecondary
education;
To help students move smoothly from one education system to another;
To ensure that all students who enter postsecondary education are prepared to
succeed;
To increase access and success of all students in postsecondary education,
especially from minority and low income groups;
To focus reform of all education organizations on practices that result in children
and youth meeting high academic standards.
NBH Solutions, LLC
Page 29
P-16 Data Sharing Systems…What is Being Done and What is Needed
The last goal focuses specifically on the "reform of teacher education, advanced
educator preparation programs, and public schools toward practices that result in all
children meeting high academic standards" (Zimpher, 1999). Georgia's P-16 initiative is
formed around a statewide P-16 council and several local and regional councils.
Leadership has been critical to the successful establishment of P-16 as a statewide
priority. The P-16 Council set overarching goals for P-16 to improve student success.
The successful student has met high standards and demonstrated achievement at each
level, and is ready to advance to the next level - of work, of occupational training, of
education - resulting in productive employment and responsible citizenship. This goal
helped give local and regional P-16 councils direction for creating curricular reforms and
initial work plans.
In 1996, the Georgia P-16 Council targeted teacher quality as a priority. A P-16
Teachers and Teacher Education Sub-Committee were appointed to assess what needed
to change in Georgia in order to improve teacher quality, and to develop
recommendations for change. Early work of the Teachers and Teacher Education Sub-
Committee resulted in:
An over-all framework for change;
Recommendations to increase the availability of alternative teacher preparation
programs and to strengthen traditional programs;
Completion of ―The Status of Teaching in Georgia,‖ a 1998 state status report of
each of the following NCTAF recommendations:
1. Establish standards for both students and teachers,
2. Enhance teacher preparation and professional development,
3. Put a qualified teacher in every classroom,
4. Encourage and reward knowledge and skills,
5. Create schools that are genuine learning organizations.
Georgia's Professional Standards Commission and Board of Regents took
immediate action on these recommendations. In 1997, the Professional Standards
Commission put in place the Innovative Program Rule to expand alternative teacher
preparation programs, and in 1998, the commission approved the first alternative teacher
preparation program (Zimpher, 1999). Following a full year of study, the Board of
Regents adopted the Policy on Teacher Preparation to be phased in at all public
NBH Solutions, LLC
Page 30
P-16 Data Sharing Systems…What is Being Done and What is Needed
universities that prepare teachers. This policy includes ten principles that address quality
assurance, collaboration, and responsiveness.
The ability to link individual students to their specific teachers has led to
numerous insights into the value of various educational interventions, and revealed the
key importance of teacher quality as a determining factor in student learning. Student
tracking systems are a useful way to support teacher quality. They offer comprehensive
research on teacher effects; a teacher‘s name is associated with each student‘s test record,
which allows for the match of students to their teachers and to the characteristics of their
classes. Which in turn, students and teachers can be tracked over multiple years, which
researchers need in order to estimate the type of statistical models (student fixed-effects
models) that account for the likely non-random match between teachers and students that
arises from classroom assignment (Goldhaber, 2005).
According to the Data Quality Campaign (2007) there are five major benefits of
linked teacher and student data systems:
1. Linked data systems can be used to meet the NCLB reporting requirement for
highly qualified teachers.
2. Linked data systems allow states to evaluate teacher preparation programs.
3. Linked data systems provide student achievement information to districts,
schools, and teachers.
4. Linked data systems can make the connection between quality and student
learning.
5. Linked student and teacher data systems can be used for feedback on teacher
performance.
An example of these benefits can be seen in Tennessee through its VAM (value-
added methodology); Tennessee has the greatest opportunity to use student achievement
data for teacher evaluation. When the model was introduced, the state promised that the
information would be confidential between the teacher and his or her principal, and it has
honored that commitment. The SEA (State Education Agencies) leaves the evaluation of
teachers up to school districts. Hamilton County (Chattanooga) uses the value-added
data to identify high-performing teachers and recruit them to work in high-poverty
schools. Tennessee offers researchers rich possibilities for investigation at the aggregate
level but allows each district to determine how VAM fits into its teacher assessments
(Bergner, Steiny, and Armstrong, 2007). VAM use shows how the student tracking
NBH Solutions, LLC
Page 31
P-16 Data Sharing Systems…What is Being Done and What is Needed
system can be used to support teacher quality by tracking individual students‘ academic
growth over several years and different subjects, researchers can estimate the
contributions that teachers make to that growth (Berry, 2007).
Action Steps for Policymakers
A unified and deliberate approach with support across all levels is critical to
developing a shared system. There is often a tendency to look for quick technical
solutions, but addressing the policy environment and political issues must come first. To
start, the state must determine what its goals are to prevent education sectors from acting
independently and at cross-purposes from one another. Providing a means for the
elementary, secondary and postsecondary sectors to work together will address some of
the inevitable ―turf‖ issues. Many states also have found that fostering voluntary and
cooperative relationships between state boards of education and higher education systems
is one of the keys to success. In fact, a single database may not be necessary to achieve a
state‗s goals; finding procedures to share data and informing policies across sectors may
prove to be the most practical and ultimately beneficial outcome. Some of the action
steps policymakers can take include (Laird, 2008):
Again, opening the lines of communication between P–12 and higher education is
critical to ensuring that students succeed at both the secondary and postsecondary levels.
NBH Solutions, LLC
Page 32
P-16 Data Sharing Systems…What is Being Done and What is Needed
Connecting student performance in college to what happens in high school will give high
schools the information they need to align curriculum and instruction to ensure that
graduates are better prepared for college and work. Most states have the technical know-
how to ensure that data can be shared between the systems. Moving forward, building
the political will to invest in the technology and people to ensure that these systems are
linked and used to improve student achievement outcomes remains an urgent and
ongoing priority.
Conclusion
Despite growing national interest in better aligning the P–12 and postsecondary
education systems to ensure all students leave high school college ready, conversations
on this issue will be limited until all states have better data provided by aligned
longitudinal data systems. Educators and policymakers need to not only collect data, but
also use the information to improve education policy and practice. To do this, they need
to know whether schools are preparing students for long-term success in college,
postsecondary training and the workplace. With the ability to match student records
between P–12 and postsecondary systems — element 9 of the 10 essential elements of a
longitudinal data system — policymakers and educators can know how graduates are
faring in postsecondary education, including (Data Quality Campaign, 2008):
The percentage of each district‘s high school graduates who enrolled in college
within 15 months after graduation;
The percentage of the previous year‘s graduates from each high school or school
district who needed remediation in college and how this percentage varied by
family income and ethnicity;
The percentage of students who met the proficiency standard on the state high
school test and still needed remediation in the same subject in college; and
How students‘ ability to stay in and complete college is related to their high
school courses, grades and test scores.
In addition, as more data are available, the power of predictive analysis will help
educators tailor the academic courses, programs and teaching practices that are proving to
be effective for helping all kids graduate from high school ready for college success.
Most states today do not have data systems that enable this two-way communication
between P–12 and postsecondary. They often have two separate data systems, and while
possible, connecting these data systems takes open communication, common goals, and
NBH Solutions, LLC
Page 33
P-16 Data Sharing Systems…What is Being Done and What is Needed
planning among the various system stakeholders. Therefore, policymakers should
consider a range of efforts that will help reduce barriers and increase support for the
building and use of longitudinal data systems, and states must continue to build, maintain,
and align them (Laird, 2008).
Currently, there is no silver bullet model for states seeking to align P–20 data
systems. The commonality across states is that they convened diverse stakeholders from
P–12 and postsecondary systems to define what they are trying to accomplish and how
aligning P–20 data systems will help achieve these goals. States such as Texas, Florida,
and California have promising systems that merit further study.
NBH Solutions, LLC
Page 34
P-16 Data Sharing Systems…What is Being Done and What is Needed
BIBLIOGRAPHY
Adelman, Clifford. (2006). The Toolbox revisited: Paths to Degree Completion from High School Trough
College. Washington, DC: U.S. Department of Education, Office of Educational Research and
Improvement (OERI).
Adelman, Stanley I. (1995). Amarillo College: Tracking in a Two-Year College Context. In P.T. Ewell
(ed), Student Tracking: New Techniques, New Demands (New Directions in Institutional Research #87,
pp. 31-43. San Francisco: Jossey-Bass.
Bergner, Terry; Steiny, Julia and Armstrong, Jane. (2007). Benefits of and Lessons Learned from Linking
Teacher and Student Data. National Center for Educational Accountability, Data Quality
Campaign. www.DataQualityCampaign.org
Berry, Barnett. (2007). Connecting Teacher and Student Data: Benefits, Challenges and Lessons Learned.
Center for Teaching Quality. www.teachingquality.org
Borden, Victor M.H. (1995). Harnessing New Technologies for Student Tracking. In P.T. Ewell (ed),
Student Tracking: New Techniques, New Demands (New Directions in Institutional Research #87,
pp.55-67. San Francisco: Jossey-Bass.
Christie, Kathy. (2008). State Data Systems. Denver, CO: Education Commission of the States.
www.esc.org
Cunningham, A. F., and Milam, J. (2005). Feasibility of a Student Unit Record System Within the
Integrated Postsecondary Education Data System (NCES 2005-160). U.S. Department of Education,
National Center for Education Statistics, Washington, DC: U.S. Government Printing Office.
Data Quality Campaign Case Study. (2006). Building Student-Level Longitudinal Data Systems: Lessons
Learned from Four States. National Center for Educational Accountability, Data Quality Campaign.
www.DataQualityCampaign.org
Daun-Barnett, Nathan, Molina, Danielle, and Walsh, Kat. (2006). Data Tracking Systems: Rethinking the
Purposes and Benefits. University of Michigan: Center for the Study of Higher and Postsecondary
Education. www.thenationalforum.org
Dorn, Sherman. (2006). A Compromise on Unit Records. Inside Higher Ed. www.insidehighered.com
Ewell, Peter T. (1995). Working over Time: The Evolution of Longitudinal Student Tracking Data Bases.
In P.T. Ewell (ed), Student Tracking: New Techniques, New Demands (New Directions in Institutional
Research #87, pp. 7-21. San Francisco: Jossey-Bass.
Ewell, Peter T.; Schild, Paula R.; Paulson, Karen. (2003). Following the Mobile Student: Can We Develop
the Capacity for a Comprehensive Database To Assess Student Progression? Lumina Foundation for
Education, Indianapolis, IN. www.luminafoundation.org
Esch, Camille E.; Sheilds, Patrick, M.; and Young, Viki, M. (2002). Strengthening California’s Teacher
Information System. Santa Cruz: CA: The Center for the Future of Teaching and Learning.
www.cftl.org
Goldhaber, Dan. (2005). Key Attributes of State Data Systems That Allow for Sophisticated Research on
Teachers. Center for Teaching Quality. www.teachingquality.org
Gosa, Kathy. (2007). Will Kansas have a K-20 Statewide Database? KBOR Data Conference: Kansas State
Department of Education.
NBH Solutions, LLC
Page 35
P-16 Data Sharing Systems…What is Being Done and What is Needed
Holub, Tamara. (2003). College Student Records: Legal Issues, Privacy, and Security Concerns.
Washington, DC: ERIC Clearinghouse on Higher Education.
Kugle, Cathy and Smith, Nancy. (2006). Florida Case Study: Building a Student-Level Longitudinal Data
System. National Center for Educational Accountability, Data Quality Campaign.
www.DataQualityCampaign.org
Laird, Elizabeth (2008, February). Developing and Supporting P-20 Education Data Systems: Different
States, Different Models. National Center for Education Achievement, Data Quality Campaign.
www.DataQualityCampaign.org
L‘Orange, Hans and Ewell, Peter. (2007). P-20 Data Systems: An Alignment Status Report. National Center
for Education Accountability, Data Quality Campaign. www.DataQualityCampaign.org
L‘Orange, Hans and Voorhees, Rick. (2003). P-16 Data and Accountability Systems. Denver, CO: State
Higher Education Executive Officers (SHEEO).
L‘Orange, Hans and Voorhees, Rick. (2003). Partnerships and Promising State Practices. Denver, CO:
State Higher Education Executive Officers (SHEEO).
Mills, Jack. (2005). State Data Systems and Privacy Concerns: Strategies for Balancing Public Interest. An
Achieving the Dream policy Brief. Boston, MA: Jobs for the Future.
Palmer, Jim (1990). Accountability through Student Tracking: A Review of the Literature. American
Association of Community and Junior Colleges, Washington, DC.
Pfeiffer, Jay J. (1994). Student Follow-Up Using Automated Record Linkage Techniques: Lessons from
Florida’s Education and Training Placement Information Program (FETPIP). Educational Resources
Information Center (ERIC).
Quimbita, Grace (1989). Student Tracking Systems in Community Colleges. ERIC Clearinghouse for Junior
Colleges, Los Angeles,CA. www.eric.ed.gov
Report to the Governor and General Assembly. (2006). Virginia’s P-16 Education Council.
www.education.virginia.gov
Romano, Lois. (2006). College Student Tracking Assailed; Opponents Cite Survey Showing Private
Schools Oppose Idea. http://www.lexisnexis.com
Russell, Alene Bycer (1995). Advances in Statewide Higher Education Data Systems. State Higher
Education Executive Officers Association, Denver, CO.
Russell, A.B. and Chisholm, M.P. (1995). Tracking in Multi-Institutional Contexts. In P.T. Ewell (ed),
Student Tracking: New Techniques, New Demands (New Directions in Institutional Research #87, pp.
43-54. San Francisco: Jossey-Bass.
Seppanen, Loretta J. (1995). Linkages to the World of Employment. In P.T. Ewell (ed), Student Tracking:
New Techniques, New Demands (New Directions in Institutional Research #87, pp. 77-93. San
Francisco: Jossey-Bass.
State Higher Education Executive Officers. (2003). Student Success: Statewide P-16 Systems. Retrieved
February 8, 2008, from www.sheeo.org/k16/p16.pdf
Venezia, A.; Kirst, M.W.; and Antonio, A.L. (2002). Betraying the College Dream: How Disconnected K-
12 and Postsecondary Education Systems Undermine Student Aspirations. Stanford, CA: The Bridge
Project, The Stanford Institute for Higher Education. http://bridgeproject.stanford.edu
NBH Solutions, LLC
Page 36
P-16 Data Sharing Systems…What is Being Done and What is Needed
Vernez, G., Krop, C.; Vuollo, M.; and Hansen, J.S. (2007). Toward a K-20 Student Unit Record Data
System for California. Rand Education.
Will, Katherine. (2006). Big Brother on Campus. Washington Post. http://www.lexisnexis.com
Zimpher, Nancy. (1999). Teacher Quality and P-16 Reform: The State Policy Context. Denver, CO: State
Higher education Executive Officers (SHEEO).
NBH Solutions, LLC
Page 37
