blog posts and news stories

New Project with ALSDE to Study AMSTI

Empirical Education is excited to announce a new study of the Alabama Math, Science, and Technology Initiative (AMSTI) commissioned by the Alabama legislature. AMSTI is the Alabama State Department of Education’s initiative to improve math and science teaching statewide; the program, which started over 20 years ago and now operates in over 900 schools across the state, has been validated multiple times by external evaluators, including Empirical’s own 2012 study directed by Chief Scientist Andrew Jaciw. That cluster-randomized trial, which involved 82 schools and ~700 teachers, assessed the efficacy of AMSTI over a three year period and showed an overall positive effect (Newman et al., 2012).

The new study will use a quasi-experimental matched comparison group design and take advantage of existing data available from the Alabama State Department of Education and the AMSTI program. Schools using AMSTI will be compared to matched schools not using AMSTI to determine the impact of the program on math and science achievement for students in grades 3 through 8, as well as differential impacts of the program on important student subgroups. We will also make use of Improvement Science principles to examine school climates where the program may have greater or reduced impact.

At the conclusion of the study, Empirical Education’s report will be made available to select committees of the Alabama state legislature, the Governor and the Alabama State Board of Education, and the Alabama State Department of Education. Researchers from Empirical Education will also travel to Montgomery, Alabama to present the study findings and recommendations for improvement to the Alabama legislature.

2018-07-13

How Are Edtech Companies Thinking About Data and Research?

Forces of the rebellion were actively at work at SIIA’s Annual Conference last week in San Francisco. Snippets of conversation revealed a common theme of harnessing and leveraging data in order to better understand and serve the needs of schools and districts.

This theme was explored in depth during one panel session, “Efficacy and Research: Why It Matters So Much in the Education Market”, where edtech executives discussed the phases and roles of research as it relates to product improvement and marketing. Moderated by Pearson’s Gary Mainor, session panelists included Andrew Coulson of the MIND Research Institute, Kelli Hill of Kahn Academy, and Shawn Mahoney of McGraw Hill Education.

Coulson, who was one of the contributing reviewers of our Research Guidelines, stated that all signs are pointing to an “exponential increase” of school district customers asking for usage data. He advised fellow edtech entrepreneurs to start paying attention to fine-grained usage data, as it is becoming necessary to provide this for customers. Panelist Kelli Hill agreed with the importance of making data visible, adding that Kahn Academy proactively provides users with monthly usage reports.

In addition to providing helpful advice for edtech sales and marketing teams, the session also addressed a pervasive misconception that that all it takes is “one good study” to validate and prove the effectiveness of a program. A company could commission one rigorous randomized trial reporting positive results and obtaining endorsement from the What Works Clearinghouse, but that study might be outdated, and more importantly, not relevant to what schools and districts are looking for. Panelist Shawn Mahoney, Chief Academic Officer of McGraw-Hill Education, affirmed that school districts are interested in “super contextualized research” and look for recent and multiple studies when evaluating a product. Q&A discussions with the panelists revealed that school decision makers are quick to claim “what works for someone else might not work for us”, supporting the notion that the conduct of multiple research studies, reporting effects for various subgroups and populations of students, is much more useful and reflective of district needs.

SIIA’s gathering proved to be a fruitful event, allowing us to reconnect with old colleagues and meet new ones, and leaving us with a number of useful insights and optimistic possibilities for new directions in research.

2018-06-22

A Rebellion Against the Current Research Regime

Finally! There is a movement to make education research more relevant to educators and edtech providers alike.

At various conferences, we’ve been hearing about a rebellion against the “business as usual” of research, which fails to answer the question of, “Will this product work in this particular school or community?” For educators, the motive is to find edtech products that best serve their students’ unique needs. For edtech vendors, it’s an issue of whether research can be cost-effective, while still identifying a product’s impact, as well as helping to maximize product/market fit.

The “business as usual” approach against which folks are rebelling is that of the U.S. Education Department (ED). We’ll call it the regime. As established by the Education Sciences Reform Act of 2002 and the Institute of Education Sciences (IES), the regime anointed the randomized control trial (or RCT) as the gold standard for demonstrating that a product, program, or policy caused an outcome.

Let us illustrate two ways in which the regime fails edtech stakeholders.

First, the regime is concerned with the purity of the research design, but not whether a product is a good fit for a school given its population, resources, etc. For example, in an 80-school RCT that the Empirical team conducted under an IES contract on a statewide STEM program, we were required to report the average effect, which showed a small but significant improvement in math scores (Newman et al., 2012). The table on page 104 of the report shows that while the program improved math scores on average across all students, it didn’t improve math scores for minority students. The graph that we provide here illustrates the numbers from the table and was presented later at a research conference.

bar graph representing math, science, and reading scores for minority vs non-minority students

IES had reasons couched in experimental design for downplaying anything but the primary, average finding, however this ignores the needs of educators with large minority student populations, as well as of edtech vendors that wish to better serve minority communities.

Our RCT was also expensive and took many years, which illustrates the second failing of the regime: conventional research is too slow for the fast-moving innovative edtech development cycles, as well as too expensive to conduct enough research to address the thousands of products out there.

These issues of irrelevance and impracticality were highlighted last year in an “academic symposium” of 275 researchers, edtech innovators, funders, and others convened by the organization now called Jefferson Education Exchange (JEX). A popular rallying cry coming out of the symposium is to eschew the regime’s brand of research and begin collecting product reviews from front-line educators. This would become a Consumer Reports for edtech. Factors associated with differences in implementation are cited as a major target for data collection. Bart Epstein, JEX’s CEO, points out: “Variability among and between school cultures, priorities, preferences, professional development, and technical factors tend to affect the outcomes associated with education technology. A district leader once put it to me this way: ‘a bad intervention implemented well can produce far better outcomes than a good intervention implemented poorly’.”

Here’s why the Consumer Reports idea won’t work. Good implementation of a program can translate into gains on outcomes of interest, such as improved achievement, reduction in discipline referrals, and retention of staff, but only if the program is effective. Evidence that the product caused a gain on the outcome of interest is needed or else all you measure is the ease of implementation and student engagement. You wouldn’t know if the teachers and students were wasting their time with a product that doesn’t work.

We at Empirical Education are joining the rebellion. The guidelines for research on edtech products we recently prepared for the industry and made available here is a step toward showing an alternative to the regime while adopting important advances in the Every Student Succeeds Act (ESSA).

We share the basic concern that established ways of conducting research do not answer the basic question that educators and edtech providers have: “Is this product likely to work in this school?” But we have a different way of understanding the problem. From years of working on federal contracts (often as a small business subcontractor), we understand that ED cannot afford to oversee a large number of small contracts. When there is a policy or program to evaluate, they find it necessary to put out multi-million-dollar, multi-year contracts. These large contracts suit university researchers, who are not in a rush, and large research companies that have adjusted their overhead rates and staffing to perform on these contracts. As a consequence, the regime becomes focused on the perfection in the design, conduct, and reporting of the single study that is intended to give the product, program, or policy a thumbs-up or thumbs-down.

photo of students in a classroom on computers

There’s still a need for a causal research design that can link conditions such as resources, demographics, or teacher effectiveness with educational outcomes of interest. In research terminology, these conditions are called “moderators,” and in most causal study designs, their impact can be measured.

The rebellion should be driving an increase the number of studies by lowering their cost and turn-around time. Given our recent experience with studies of edtech products, this reduction can reach a factor of 100. Instead of one study that costs $3 million and takes 5 years, think in terms of a hundred studies that cost $30,000 each and are completed in less than a month. If for each product, there are 5 to 10 studies that are combined, they would provide enough variation and numbers of students and schools to detect differences in kinds of schools, kinds of students, and patterns of implementation so as to find where it works best. As each new study is added, our understanding of how it works and with whom improves.

It won’t be enough to have reviews of product implementation. We need an independent measure of whether—when implemented well—the intervention is capable of a positive outcome. We need to know that it can make (i.e., cause) a difference AND under what conditions. We don’t want to throw out research designs that can detect and measure effect sizes, but we should stop paying for studies that are slow and expensive.

Our guidelines for edtech research detail multiple ways that edtech providers can adapt research to better work for them, especially in the era of ESSA. Many of the key recommendations are consistent with the goals of the rebellion:

  • The usage data collected by edtech products from students and teachers gives researchers very precise information on how well the program was implemented in each school and class. It identifies the schools and classes where implementation met the threshold for which the product was designed. This is a key to lowering cost and turn-around time.
  • ESSA offers four levels of evidence which form a developmental sequence, where the base level is based on existing learning science and provides a rationale for why a school should try it. The next level looks for a correlation between an important element in the rationale (measured through usage of that part of the product) and a relevant outcome. This is accepted by ESSA as evidence of promise, informs the developers how the product works, and helps product marketing teams get the right fit to the market. a pyramid representing the 4 levels of ESSA
  • The ESSA level that provides moderate evidence that the product caused the observed impact requires a comparison group matched to the students or schools that were identified as the users. The regime requires researchers to report only the difference between the user and comparison groups on average. Our guidelines insist that researchers must also estimate the extent to which an intervention is differentially effective for different demographic categories or implementation conditions.

From the point of view of the regime, nothing in these guidelines actually breaks the rules and regulations of ESSA’s evidence standards. Educators, developers, and researchers should feel empowered to collect data on implementation, calculate subgroup impacts, and use their own data to generate evidence sufficient for their own decisions.

A version of this article was published in the Edmarket Essentials magazine.

2018-05-09

AERA 2018 Recap: The Possibilities and Necessity of a Rigorous Education Research Community

This year’s AERA annual meeting on “The Dreams, Possibilities, and Necessity of Public Education,” was fittingly held in the city with the largest number of public school students in the country—New York. Against this radically diverse backdrop, presenters were encouraged to diversify both the format and topics of presentations in order to inspire thinking and “confront the struggles for public education.”

AERA’s sheer size may risk overwhelming its attendees, but in other ways, it came as a relief. At a time when educators and education remain under-resourced, it was heartening to be reminded that a large, vibrant community of dedicated and intelligent people exists to improve educational opportunities for all students.

One theme that particularly stood out is that researchers are finding increasingly creative ways to use existing usage data from education technology products to measure impact and implementation. This is a good thing when it comes to reducing the cost of research and making it more accessible to smaller businesses and nonprofits. For example, in a presentation on a software-based knowledge competition for nursing students, researchers used usage data to identify components of player styles and determine whether these styles had a significant effect on student performance. In our Edtech Research Guidelines, Empirical similarly recommends that edtech companies take advantage of their existing usage data to run impact and implementation analyses, without using more expensive data collection methods. This can help significantly reduce the cost of research studies—rather than one study that costs $3 million, companies can consider multiple lower-cost studies that leverage usage data and give the company a picture of how the product performs in a greater diversity of contexts.

Empirical staff themselves presented on a variety of topics, including quasi-experiments on edtech products; teacher recruitment, evaluation, and retention; and long-term impact evaluations. In all cases, Empirical reinforced its commitment to innovative, low-cost, and rigorous research. You can read more about the research projects we presented in our previous AERA post.

photo of Denis Newman presenting at AERA 2018

Finally, Empirical was delighted to co-host the Division H AERA Reception at the Supernova bar at Novotel Hotel. If you ever wondered if Empirical knows how to throw a party, wonder no more! A few pictures from the event are below. View all of the pictures from our event on facebook!


We had a great time and look forward to seeing everyone at the next AERA annual meeting!

2018-05-03

Updated Research Guidelines Will Improve Education Technology Products and Provide More Value to Schools

Recommendations include 16 best practices for the design, implementation, and reporting of Usable Evidence for Educators

Palo Alto, CA (April 25, 2018) – Empirical Education Inc. and the Education Technology Industry Network (ETIN) of SIIA released an important update to the “Guidelines for Conducting and Reporting Edtech Impact Research in U.S. K-12 Schools” today.

Authored by Empirical Education researchers, Drs. Denis Newman, Andrew Jaciw, and Valeriy Lazarev, the Guidelines detail 16 best practices for the design, implementation, and reporting of efficacy research of education technology. Recommendations range from completing the product’s logic model before fielding it to disseminating a study’s results in accessible and non-technical language.

The Guidelines were first introduced in July 2017 at ETIN’s Edtech Impact Symposium to address the changing demand for research. They served to address new challenges driven by the accelerated pace of edtech development and product releases, the movement of new software to the cloud, and the passage of the Every Student Succeeds Act (ESSA). The authors committed to making regular updates to keep pace with technical advances in edtech and research methods.

“Our collaboration with ETIN brought the right mix of practical expertise to this important document,” said Denis Newman, CEO of Empirical Education and lead author of the Guidelines. “ETIN provided valuable expertise in edtech marketing, policy, and development. With over a decade of experience evaluating policies, programs, and products for the U.S. Department of Education, major research organizations, and publishers, Empirical Education brought a deep understanding of how studies are traditionally performed and how they can be improved in the future. Our experience with our Evidence as a Service™ offering to investors and developers of edtech products also informed the guidelines.”

The current edition advocates for analysis of usage patterns in the data collected routinely by edtech applications. These patterns help to identify classrooms and schools with adequate implementation and lead to lower-cost faster turn-around research. So rather than investing hundreds of thousands of dollars in a single large-scale study, developers should consider multiple small-scale studies. The authors point to the advantages of looking at subgroup analysis to better understand how and for whom the product works best, thus more directly answering common educator questions. Issues with quality of implementation are addressed in greater depth, and the visual design of the Guidelines has been refined for improved readability.

“These guidelines may spark a rebellion against the research business as usual, which doesn’t help educators know whether an edtech product will work for their specific populations. They also provide a basis for schools and developers to partner to make products better,” said Mitch Weisburgh, Managing Partner of Academic Business Advisors, LLC and President of ETIN, who has moderated panels and webinars on edtech research.

Empirical Education, in partnership with a variety of organizations, is conducting webinars to help explain the updates to the Guidelines, as well as to discuss the importance of these best practices in the age of ESSA. The updated Guidelines are available here: https://www.empiricaleducation.com/research-guidelines/.

2018-04-25

Where's Denis?

It’s been a busy month for Empirical CEO Denis Newman, who’s been in absentia at our Palo Alto office as he jet-sets around the country to spread the good word of rigorous evidence in education research.

His first stop was Washington, DC and the conference of the Society for Research on Educational Effectiveness (SREE). This was an opportunity to get together with collaborators, as well as plot proposal writing, blog postings, webinars, and revisions to our research guidelines for edtech impact studies. Andrew Jaciw, Empirical’s Chief Scientist, kept up the company’s methodological reputation with a paper presentation on “Leveraging Fidelity Data to Make Sense of Impact Results.” For Denis, a highlight was dinner with Peg Griffin, a longtime friend and his co-author on The Construction Zone. Then it was on to Austin, TX, for a very different kind of meeting—more of a festival, really.

At this year’s SXSWEDU, Denis was one of three speakers on the panel, “Can Evidence Even Keep Up with Edtech?” The problem presented by the panel was that edtech, as a rapidly moving field, seems to be outpacing the rate of research that stakeholders may want to use to evaluate these products. How, then, could education stakeholders make informed decisions about whether to use edtech products?

According to Denis, the most important thing is for a district to have enough information to know whether a given edtech product may or may not work for that district’s unique population and context. Therefore, researchers may need to adapt their methods both to be able to differentiate a product’s impact between subgroups, as well as to meet the faster timelines of edtech product development. Empirical’s own solution to this quandry, Evidence as a ServiceTM, offers quick-turnaround research reports that can examine impact and outcomes for specific student subgroups, with methodology that is flexible but rigorous enough to meet ESSA standards.

Denis praised the panel, stating, “In the festival’s spirit of invention, our moderator, Mitch Weisberg, masterfully engaged the audience from the beginning to pose the questions for the panel. Great questions, too. I got to cover all of my prepared talking points!”

You can read more coverage of our SXSWEDU panel on EdSurge.

After the panel, a string of meetings and parties kept the energy high and continued to show the growing interest in efficacy. The ISTE meetup was particularly important following this theme. The concern raised by the ISTE leadership and its members—which are school-based technology users—was that traditional research doesn’t tell the practitioners whether a product is likely to work in their school, given its resources and student demographics. Users are faced with hundreds of choices in any product category and have little information for narrowing down the choice to a few that are worth piloting.

Following SXSWEDU, it was back to DC for the Consortium for School Networking (CoSN) conference. Denis participated in the annual Feedback Forum hosted by CoSN and the Software & Information Industry Association (SIIA), where SIIA—representing edtech developers—looked for feedback from the CIOs and other school district leaders. This year, SIIA was looking for feedback that would help the Empirical team improve the edtech research guidelines, which are sponsored by SIIA’s Education Technology Industry Network (ETIN). Linda Winter moderated and ran the session like a focus group, asking questions such as:

  • What data do you need from products to gauge engagement?
  • How can the relationship of engagement and achievement indicate that a product is working?
  • What is the role of pilots in measuring success?
  • And before a pilot decision is made, what do CoSN members need to know about edtech products to decide if they are likely to work?

The CoSN members were brutally honest, pointing out that as the leaders responsible for the infrastructure, they were concerned with implementability, bandwidth requirements, and standards such as single sign-on. Whether the software improved learning was secondary—if teachers couldn’t get the program to work, it hardly mattered how effective it may be in other districts.

Now, Denis is preparing for the rest of the spring conference season. Next stop will be New York City and the American Education Research Association (AERA) conference, which attracts over 20,000 researchers annually. The Empirical team will be presenting four studies, as well as co-hosting a cocktail reception with AERA’s school research division. Then, it’s back on the plane for ASU-GSV in San Diego.

For more information about Evidence as a Service, the edtech research guidelines, or to invite Denis to speak at your event, please email rmeans@empiricaleducation.com

2018-03-26

Presenting at AERA 2018

We will again be presenting at the annual meeting of the American Educational Research Association (AERA). Join the Empirical Education team in New York City from April 13-17, 2018.

Research presentations will include the following.

For Quasi-Experiments on EdTech Products, What Counts as Being Treated?
Authors: Val Lazarev, Denis Newman, & Malvika Bhagwat
In Roundtable Session: Examining the Impact of Accountability Systems on Both Teachers and Students
Friday, April 13 - 2:15 to 3:45pm
New York Marriott Marquis, Fifth Floor, Westside Ballroom Salon 3

Abstract: Edtech products are becoming increasingly prevalent in K-12 schools and the needs of schools to evaluate their value for students calls for a program of rigorous research, at least at the level 2 of the ESSA standards for evidence. This paper draws on our experience conducting a large scale quasi-experiment in California schools. The nature of the product’s wide-ranging intensity of implementation presented a challenge in identifying schools that had used the product adequately enough to be considered part of the treatment group.


Planning Impact Evaluations Over the Long Term: The Art of Anticipating and Adapting
Authors: Andrew P Jaciw & Thanh Thi Nguyen
In Session: The Challenges and Successes of Conducting Large-Scale Educational Research
Saturday, April 14 - 2:15 to 3:45pm
Sheraton New York Times Square, Second Floor, Central Park East Room

Abstract: Perspective. It is good practice to identify core research questions and important elements of study designs a-priori, to prevent post-hoc “fishing” exercises and reduce the role of drawing false-positive conclusions [16,19]. However, programs in education, and evaluations of them, evolve [6] making it difficult to follow a charted course. For example, in the lifetime of a program and its evaluation, new curricular content or evidence standards for evaluations may be introduced and thus drive changes in program implementation and evaluation.

Objectives. This work presents three cases from program impact evaluations conducted through the Department of Education. In each case, unanticipated results or changes in study context had significant consequences for program recipients, developers and evaluators. We discuss responses, either enacted or envisioned, for addressing these challenges. The work is intended to serve as a practical guide for researchers and evaluators who encounter similar issues.

Methods/Data Sources/Results. The first case concerns the problem of outcome measures keeping pace with evolving content standards. For example, in assessing impacts of science programs, program developers and evaluators are challenged to find assessments that align with Next Generation Science Standards (NGSS). Existing NGSS-aligned assessments are largely untested or in development, resulting in the evaluator having to find, adapt or develop instruments with strong reliability, and construct and face validity – ones that will be accepted by independent review and not considered over-aligned to the interventions. We describe a hands-on approach to working with a state testing agency to develop forms to assess impacts on science generally, and on constructs more-specifically aligned to the program evaluated. The second case concerns the problem of reprioritizing research questions mid-study. As noted above, researchers often identify primary (confirmatory) research questions at the outset of a study. Such questions are held to high evidence standards, and are differentiated from exploratory questions, which often originate after examining the data, and must be replicated to be considered reliable [16]. However, sometimes, exploratory analyses produce unanticipated results that may be highly consequential. The evaluator must grapple with the dilemma of whether to re-prioritize the result, or attempt to proceed with replication. We discuss this issue with reference to an RCT in which the dilemma arose. The third addresses the problem of designing and implementing a study that meets one set of evidence standards, when the results will be reviewed according to a later version of those standards. A practical question is what to do when this happens and consequently the study falls under a lower tier of the new evidence standard. With reference to an actual case, we consider several response options, including assessing the consequence of this reclassification for future funding of the program, and augmenting the research design to satisfy the new standards of evidence.

Significance. Responding to demands of changing contexts, programs in the social sciences are moving targets. They demand a flexible but well-reasoned and justified approach to evaluation. This session provides practical examples and is intended to promote discussion for generating solutions to challenges of this kind.


Indicators of Successful Teacher Recruitment and Retention in Oklahoma Rural Schools
Authors: Val Lazarev, Megan Toby, Jenna Lynn Zacamy, Denis Newman, & Li Lin
In Session: Teacher Effectiveness, Retention, and Coaching
Saturday, April 14 - 4:05 to 6:05pm
New York Marriott Marquis, Fifth Floor, Booth

Abstract: The purpose of this study was to identify factors associated with successful recruitment and retention of teachers in Oklahoma rural school districts, in order to highlight potential strategies to address Oklahoma’s teaching shortage. The study was designed to identify teacher-level, district-level, and community characteristics that predict which teachers are most likely to be successfully recruited and retained. A key finding is that for teachers in rural schools, total compensation and increased responsibilities in job assignment are positively associated with successful recruitment and retention. Evidence provided by this study can be used to inform incentive schemes to help retain certain groups of teachers and increase retention rates overall.


Teacher Evaluation Rubric Properties and Associations with School Characteristics: Evidence from the Texas Evaluation System
Authors: Val Lazarev, Thanh Thi Nguyen, Denis Newman, Jenna Lynn Zacamy, Li Lin
In Session: Teacher Evaluation Under the Microscope
Tuesday, April 17 - 12:25 to 1:55pm
New York Marriott Marquis, Seventh Floor, Astor Ballroom

Abstract: A 2009 seminal report, The Widget Effect, alerted the nation to the tendency of traditional teacher evaluation systems to treat teachers like widgets, undifferentiated in their level of effectiveness. Since then, a growing body of research, coupled with new federal initiatives, has catalyzed the reform of such systems. In 2014-15, Texas piloted its reformed evaluation system, collecting classroom observation rubric ratings from over 8000 teachers across 51 school districts. This study analyzed that large dataset and found that 26.5 percent, compared to 2 percent under previous measures, of teachers were rated below proficient. The study also found a promising indication of low bias in the rubric ratings stemming from school characteristics, given that they were minimally associated with observation ratings.

We look forward to seeing you at our sessions to discuss our research. We’re also co-hosting a cocktail reception with Division H! If you’d like an invite, let us know.

2018-03-06

Jefferson Education Accelerator Contracts with Empirical for Evidence as a Service™

Jefferson Education Accelerator (JEA) has contracted with Empirical Education Inc. for research services that will provide evidence of the impact of education technology products developed by their portfolio companies. JEA’s mission is to support and evaluate promising edtech solutions in order to help educators make more informed decisions about the products they invest in. The study is designed to meet level 2 or “moderate” evidence as defined by the Every Student Succeeds Act. Empirical will provide a Student Impact Report under its Evidence as a Service offering, which combines student-level product usage data and a school district’s administrative data to conduct a comparison group study. Denis Newman, Empirical’s CEO stated, “This is a perfect application of our Evidence as a Service product, which provides fast answers to questions about which kids will benefit the most from any particular learning program.” Todd Bloom, JEA’s Chief Academic Officer and Research Associate Professor at UVA’s Curry School of Education, commented: “Empirical Education is a highly respected research firm and offers the type of aggressive timeline that is sorely needed in the fast-paced edtech market.” A report on impact in the school year 2017-2018 is expected to be completed in July.

2018-02-20

IES Published Our REL Southwest Study on Trends in Teacher Mobility

The U.S. Department of Education’s Institute of Education Sciences published a report of a study we conducted for REL Southwest! We are thankful for the support and engagement we received from the Educator Effectiveness Research Alliance throughout the study.

The study was published in December 2017 and provides updated information regarding teacher mobility for Texas public schools during the 201112 through 201516 school years. Teacher mobility is defined as teachers changing schools or leaving the public school system.

In the report, descriptive information on mobility rates is presented at the regional and state levels for each school year. Mobility rates are disaggregated further into destination proportions to describe the proportion of teacher mobility due to within-district movement, between-district movement, and leaving Texas public schools. This study leverages data collected by the Texas Education Agency during the pilot of the Texas Teacher Evaluation and Support System (T-TESS) in 57 school districts in 201415. Analyses examine how components of the T-TESS observation rubric are related to school-level teacher mobility rates.

During the 2011-12 school year, 18.7% of Texas teachers moved schools within a district, moved between districts, or left the Texas Public School system. By 2015-16, this mobility rate had increased to 22%. Moving between districts was the primary driver of the increase in mobility rates. Results indicate significant links between mobility and teacher, student, and school demographic characteristics. Teachers with special education certifications left Texas public schools at nearly twice the rate of teachers with other teaching certifications. School-level mobility rates showed significant positive correlations with the proportion of special education, economically disadvantaged, low-performing, and minority students. School-level mobility rates were negatively correlated with the proportion of English learner students. Schools with higher overall observation ratings on the T-TESS rubric tended to have lower mobility rates.

Findings from this study will provide state and district policymakers in Texas with updated information about trends and correlates of mobility in the teaching workforce, and offer a systematic baseline for monitoring and planning for future changes. Informed by these findings, policymakers can formulate a more strategic and targeted approach for recruiting and retaining teachers. For instance, instead of using generic approaches to enhance the overall supply of teachers or improve recruitment, more targeted efforts to attract and retain teachers in specific subject areas (for example, special education), in certain stages of their career (for example, novice teachers), and in certain geographic areas are likely to be more productive. Moreover, this analysis may enrich the existing knowledge base about schools’ teacher retention and mobility in relation to the quality of their teaching force, or may inform policy discussions about the importance of a stable teaching force for teaching effectiveness.

2018-02-01

How Efficacy Studies Can Help Decision-makers Decide if a Product is Likely to Work in Their Schools

We and our colleagues have been working on translating the results of rigorous studies of the impact of educational products, programs, and policies for people in school districts who are making the decisions whether to purchase or even just try out—pilot—the product. We are influenced by Stanford University Methodologist Lee Cronbach, especially his seminal book (1982) and article (1975) where he concludes “When we give proper weight to local conditions, any generalization is a working hypothesis, not a conclusion…positive results obtained with a new procedure for early education in one community warrant another community trying it. But instead of trusting that those results generalize, the next community needs its own local evaluation” (p. 125). In other words, we consider even the best designed experiment to be like a case study, as much about the local and moderating role of context, as about the treatment when interpreting the causal effect of the program.

Following the focus on context, we can consider characteristics of the people and of the institution where the experiment was conducted to be co-causes of the result that deserve full attention—even though, technically, only the treatment, which was randomly assigned was controlled. Here we argue that any generalization from a rigorous study, where the question is whether the product is likely to be worth trying in a new district, must consider the full context of the study.

Technically, in the language of evaluation research, these differences in who or where the product or “treatment” works are called “interaction effects” between the treatment and the characteristic of interest (e.g., subgroups of students by demographic category or achievement level, teachers with different skills, or bandwidth available in the building). The characteristic of interest can be called a “moderator”, since it changes, or moderates, the impact of the treatment. An interaction reveals if there is differential impact and whether a group with a particular characteristic is advantaged, disadvantaged, or unaffected by the product.

The rules set out by The Department of Education’s What Works Clearinghouse (WWC) focus on the validity of the experimental conclusion: Did the program work on average compared to a control group? Whether it works better for poor kids than for middle class kids, works better for uncertified teachers versus veteran teachers, increases or closes a gap between English learners and those who are proficient, are not part of the information provided in their reviews. But these differences are exactly what buyers need in order to understand whether the product is a good candidate for a population like theirs. If a program works substantially better for English proficient students than for English learners, and the purchasing school has largely the latter type of student, it is important that the school administrator know the context for the research and the result.

The accuracy of an experimental finding depends on it not being moderated by conditions. This is recognized with recent methods of generalization (Tipton, 2013) that essentially apply non-experimental adjustments to experimental results to make them more accurate and more relevant to specific local contexts.

Work by Jaciw (2016a, 2016b) takes this one step further.

First, he confirms the result that if the impact of the program is moderated, and if moderators are distributed differently between sites, then an experimental result from one site will yield a biased inference for another site. This would be the case, for example, if the impact of a program depends on individual socioeconomic status, and there is a difference between the study and inference sites in the proportion of individuals with low socioeconomic status. Conditions for this “external validity bias” are well understood, but the consequences are addressed much less often than the usual selection bias. Experiments can yield accurate results about the efficacy of a program for the sample studied, but that average may not apply either to a subgroup within the sample or to a population outside the study.

Second, he uses results from a multisite trial to show empirically that there is potential for significant bias when inferring experimental results from one subset of sites to other inference sites within the study; however, moderators can account for much of the variation in impact across sites. Average impact findings from experiments provide a summary of whether a program works, but leaves the consumer guessing about the boundary conditions for that effect—the limits beyond which the average effect ceases to apply. Cronbach was highly aware of this, titling a chapter in his 1982 book “The Limited Reach of Internal Validity”. Using terms like “unbiased” to describe impact findings from experiments is correct in a technical sense (i.e., the point estimate, on hypothetical repeated sampling, is centered on the true average effect for the sample studied), but it can impart an incorrect sense of the external validity of the result: that it applies beyond the instance of the study.

Implications of the work cited, are, first, that it is possible to unpack marginal impact estimates through subgroup and moderator analyses to arrive at more-accurate inferences for individuals. Second, that we should do so—why obscure differences by paying attention to only the grand mean impact estimate for the sample? And third, that we should be planful in deciding which subgroups to assess impacts for in the context of individual experiments.

Local decision-makers’ primary concern should be with whether a program will work with their specific population, and to ask for causal evidence that considers local conditions through the moderating role of student, teacher, and school attributes. Looking at finer differences in impact may elicit criticism that it introduces another type of uncertainty—specifically from random sampling error—which may be minimal with gross impacts and large samples, but influential when looking at differences in impact with more and smaller samples. This is a fair criticism, but differential effects may be less susceptible to random perturbations (low power) than assumed, especially if subgroups are identified at individual levels in the context of cluster randomized trials (e.g., individual student-level SES, as opposed to school average SES) (Bloom, 2005; Jaciw, Lin, & Ma, 2016).

References:
Bloom, H. S. (2005). Randomizing groups to evaluate place-based programs. In H. S. Bloom (Ed.), Learning more from social experiments. New York: Russell Sage Foundation.

Cronbach, L. J. (1975). Beyond the two disciplines of scientific psychology. American Psychologist, 116-127.

Cronbach, L. J. (1982). Designing evaluations of educational and social programs. San Francisco, CA: Jossey-Bass.

Jaciw, A. P. (2016). Applications of a within-study comparison approach for evaluating bias in generalized causal inferences from comparison group studies. Evaluation Review, (40)3, 241-276. Retrieved from http://erx.sagepub.com/content/40/3/241.abstract

Jaciw, A. P. (2016). Assessing the accuracy of generalized inferences from comparison group studies using a within-study comparison approach: The methodology. Evaluation Review, (40)3, 199-240. Retrieved from http://erx.sagepub.com/content/40/3/199.abstract

Jaciw, A., Lin, L., & Ma, B. (2016). An empirical study of design parameters for assessing differential impacts for students in group randomized trials. Evaluation Review. Retrieved from http://erx.sagepub.com/content/early/2016/10/14/0193841X16659600.abstract

Tipton, E. (2013). Improving generalizations from experiments using propensity score subclassification: Assumptions, properties, and contexts. Journal of Educational and Behavioral Statistics, 38, 239-266.

2018-01-16
Archive