The State of Personalized Learning

Examining products, practices, and an easy-to-use taxonomy.

THOUGHT LEADERSHIP | by Philip Hickman, Ken Eastwood, and Eliot Levinson

CREDIT BLEgroup image.pngPersonalized learning is the edtech phrase of the year and a key concept for how digitally delivered teaching and learning will make the most significant impact on teaching and learning, but what is it? The variety of products and practices that claim to personalize learning differ vastly, making the term hard to understand.

Schools want to provide personalized learning, a plethora of vendors are selling it, and it is required for ESSA beginning in 2018. A few BLEgroup superintendents, staff, and vendors held a discussion to see if we could define a way to understand the reality of the concept. We also conversed with vendors who believe they are providing personalized learning products and services to K-12 schools and with school districts who believe that they are far down the path of providing personalized teaching and learning.

There are currently at least three generations of personalized learning on the market.

All personalized learning is not the same: Before data collection and discussion, all stakeholders held a common definition of personalized learning as “customized curriculum targeted to each student’s level of subject knowledge and his/her learning style.” After our discussion and observation of several districts’ practices and conversations with vendors of personalized learning platforms, we came to the conclusion that there needed to be an easy-to-use taxonomy for the types of PL products and practices so that school systems can clearly understand product and practice variation.

The current use of the term “personalized learning” varies from:

  • small group instruction based on performance levels to
  • longitudinal history of all assessments students ever took to provide them with knowledge of what to assign to
  • artificial intelligence based products that assess the cognitive level and learning style of a student and provide a variety of resources based on the student’s learning style, current performance, and understanding of a subject.

We concluded that rapid new developments in the areas of adaptive assessment, data analytics, customized content, and adaptive instruction in recent years are the reason why “personalized learning” is not a well understood concept. There are currently at least three generations of personalized learning on the market.

The purpose of our piece is to provide a structure for defining the levels of personalized instruction, comments on the advantages and disadvantages of each, and recommendations to vendors and schools about where the market may be heading. 

Generation I: Many school districts provide teachers with longitudinal data on individual students, including classroom tests, standards-based tests, and formative assessments. The reason is that the assessments will be able to identify exactly which standards and cognitive level a student is achieving and find appropriate resources for that student. This approach also provides performance data on the classes, which provide data on teachers and can focus the schools on particular problems. The longitudinal data approach has been around longer than any other personalization approach.

  • The advantage of the longitudinal data is that it provides teachers with performance levels of students on each standard over time.
  • The disadvantage is that it is difficult for teachers to locate appropriate materials for every student, and the longitudinal data shows performance but does not diagnose missing skills that a student did not learn in earlier years and may be preventing her from mastering a standard now. In reality many teachers don’t have the bandwidth to individualize for a class of 30. The best they can master is small group instruction.

Generation II: Adaptive assessment is done a few times throughout the year, which identifies the exact level at which a student is performing. For example, a first grader could be reading at the 8th grade level, and a 4th grade math student could be functioning at the 2nd grade level.

The adaptive assessment identifies the skills needed to achieve different standards that a student may not have achieved because of missing skills they did not learn earlier. Some teachers and products can re-teach missing skills, while others do not.

The major providers of adaptive assessment are NWEA, Curriculum Associates, and Scantron. Adaptive assessments all identify missing skills. The tests are lengthy to deliver, and some have a 3-phase process, including the original assessment, a diagnostic structure to identify missing skills, and a re-teaching component that is used to deal with missing skills.

  • The advantage of adaptive assessments is that they identify missing skills that prevent students from mastering a standard being taught 3 years later. Some of the products have the ability to re-teach the missing skills
  • The disadvantage is that, in most cases, teachers teach the first part of the test only to find out what a student has mastered but don’t take the second step to diagnose and re-teach the missing skills necessary for moving forward, as that phase is extremely time consuming. Not doing so prevents rapid learning and mastery of standards by the student. When teachers diagnose the missing skills and re-teach them, the students do rapidly master standards.

Below are examples of two school districts that are highly successful in fully utilizing adaptive assessments. In each case they provide extra time for the diagnosis and planning for each student and hold teachers accountable for utilizing the data. One district has very little technology, and the other has a vast amount of technology, but they are both successful.

Two Eagle River School – Montana
There are interesting exceptions to the poor use of adaptive assessments that show no results. The more effective schools focus on professional development and time to analyze the missing skills.

At Two Eagle School, a native American alternative school in Montana which receives throw outs from public high schools, the superintendent provides teachers with 13 hours a month to analyze the adaptive assessment to the skill level and presents plans to the rest of the faculty. Teachers use their diagnoses to plan students’ work for the next three weeks and present and discuss them with the whole faculty.

The Two Eagle alternative school is the only school in Montana consistently graduating Native American Students from high school. They have moved from being one of the worst performing high schools in the state to one of a highly proficient level. NWEA adaptive assessment is the only technology resource used to personalize learning.

Enlarged City School District of Middletown – New York
Middletown, like Two Eagle, is an 80 percent minority district containing Hispanic students. Middletown has a large technology budget because they won a $20 million Race to the Top grant.

Middletown has made tremendous gains in performance and graduation rates due to the peer professional developers in each school that provide ongoing PD to teachers. When individual skill performance problems are identified, the district immediately provides the necessary teaching resources to the teacher.

Generation III: Adaptive instruction is the new generation of personalized learning and initially appears to be the major breakthrough for real personalized learning due to the use of artificial intelligence on cognitive and learning style issues to truly individualize instruction. Also it changes the value chain by potentially being able to leapfrog assessments and Learning Management Systems. The preliminary results on these products are promising. The main product brands in this space are Curriculum Associate, Dreambox Learning, Apex Learning, Fishtree, and MyOn reader, a specialized product that is a virtual library that can be included in the group.

The adaptive instruction products are often composed of artificial intelligence. Metrics assess performance levels and learning style preferences of students and also provide ongoing performance analytics.

Adaptive instruction products provide a customized learning experience based on the ongoing data that the system is collecting about students. The advantage of this system is that there are myriad resources that the system provides for the student. Some of these systems use open resources that reduce instructional resource costs for a district.

Adaptive instruction products have recently entered the market so we can’t yet make statements about their impact. The evidence, though positive, is insufficient to prove that these products produce the best results of the three generations of personalized learning products. However, we think adaptive instruction is the most promising form of personalized instruction products.

There are several economic benefits for using adaptive instruction products:

  • Most of the products use open source curriculum, which cuts instructional material costs.
  • They eliminate the need for formative or adaptive assessment systems as testing is integrated into the products, thereby allowing districts to cut significant costs.

We want to be explicit that these systems increase teacher productivity and allow time to track results and work individually with students. The qualitative data on the third generation demonstrates that students learn quickly and have high levels of engagement in learning.

Wrap-up: There are three generations of personalized learning products, which is confusing to the market. The three generations of are:

  • Longitudinal data
  • Adaptive assessment
  • Adaptive instruction

All three generations are “personalized learning,” but each generation appears to provide an increasing amount of personalization. Although early in the game, adaptive instruction appears to provide the most personalized learning and provides additional benefits, such as lowering instructional material, LMS, and assessment costs, while providing pedagogical advantages of providing teachers more time to work with individuals and meaningful data to understand the student’s progress. However, it is early in the third generation, and we don’t yet have all the answers.

Philip W.V. Hickman is Superintendent, Columbus (MS) Municipal School District; Kenneth Eastwood is Superintendent, Enlarged City School District of Middletown (NY); and Eliot Levinson is President, The BLEgroup a PCG company. The BLEgroup, an organization of leading edtech decision makers, works with both schools and the industry. Write to: eliot@blegroup.com.

This entry was posted in cool tools. Bookmark the permalink.

One Response to The State of Personalized Learning

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s