How NSF-backed AdaptivePosit boosts student learning



tl;dr: With AdaptivePosit, students who were significantly behind at the start of the unit, scored on average the same as their peers at the end.

Background:
PlayPosit is currently working on AdaptivePosit, an NSF-funded platform to deliver adaptive interactive video content. These are preliminary results for one class. If you're a teacher or school administrator and would be interested in participating in the study, please email Seijin at playposit.org.

Case study:
Mr. Butcher is a 7th grade instructor of science in Los Angeles Unified School District. He’s a first year instructor who has just started using PlayPosit to leverage student achievement in his classroom.

Classroom statistics:
34 students (9 special education students, 25 classified fluent)

Student aptitude:
General student aptitude was determined by Mr. Butcher in an interview after the lesson was delivered. There were 4 students with high general aptitude, 18 medium, and 12 low.

Time period:
1 week of implementation in four 52-minute classes and one 39-minute class. The class completed the pre-diagnostic, 3 lessons, and the post-diagnostic.

Student reviews:
“The video had a lot of information that I did not know about... the information was really good.”

Study details:
Students were randomly separated into three groups: (1) Interactive Adaptive, (2) Interactive, and (3) Non-interactive.

Interactive Adaptive students were taken down an adaptive content course which adjusted the language used to explain scientific principles into more simple or complex terms. This was based on the student’s performance answering PlayPosit questions. Students who did well were taken down a track that delved deeper into principles, and students who did not perform as well were taken down a track that used simpler language to break content down into its fundamental principles and concepts.

Interactive students were taken down a track with PlayPosit’s interactive content without adaptivity, and Non-interactive students were taken down a track with no interaction and no adaptivity; they only watched video content. The small sample size resulted in an uneven distribution of performance on the pre-test diagnostic (Table 1).

Findings:
Consistent across all aptitude levels, Adaptive students made the largest net gain in points. Non-interactive students made the smallest. Students who took the adaptive content more than doubled their post-diagnostic scores, and scored 25% higher than Interactive students. The end result was that students who were significantly behind at the start of the unit, scored on average the same as their peers at the end. Special education status  and score differences with peers were negligible, suggesting that the content could be used by all groups of students.

Get involved:
Our initial findings indicate that AdaptivePosit has an overwhelmingly positive effect on classrooms, but we need more data to verify this! We would love to partner with you, and instructors at your school for a focus group in the next month. If selected, we’ll fly out to facilitate this program with you over the course of a couple days. Please contact us at Seijin at playposit.org and we’ll reach out!

Figure 1: Improvement in post-test



Table 1: Diagnostic report

Pre-test pointsPost-test pointsChange in points% change
Interactive1834+16+88
Interactive Adaptive1636+20+125%
Non-interactive2737+10+37%

Observations: 
Adaptive students on average scored the lowest in the pre-diagnostic, but caught up to peers in the post-diagnostic. 


Table 2: Change in points for students at varying aptitude levels

Pre-test pointsPost-test pointsChange in points% change
High102818180%
Medium1216433%
Low6149150%

Observations: 
High level students continued to score highest scores, outpacing the rest of their peers. Medium level students made an incremental gain. Low level students more than doubled their initial scores, and ended almost on par with their medium level peers.


Table 3: Special Education Status

Pre-test pointsPost-test pointsChange in points% change
Special Ed.1017770%
Non-Special Ed.1017770%

Observations: 
There was a negligible difference between students who were classified as special education students, and their peers.