高等学校におけるアダプティブラーニング実証実験の経過レポート 〜苦手克服レコメンドの活用パターンを可視化〜:Intermediate Progress Report on High School Adaptive Learning Validation Experiment: Visualizing the Usage Patterns of Recommendations to Overcome Weaknesses

2016. 10. 2


Progress report of the Adaptive Learning Implementation Experiment research program conducted by the Recruit Institute of Next Generation Education, Director Rieko KOMIYAMA.

Adaptive Learning Implementation Experiment


This implementation experiment is conducted with the cooperation and help of the schools which have implemented Study Sapuri High School Courses / Study Sapuri University Courses ( to validate the effectiveness of adaptive learning in a public education environment.


Understanding every student’s stumbling blocks and providing an individualized instruction to all students is an unreasonable burden on instructors. With a goal of providing effective learning using adaptive learning, study sapuri worksheet logs were analyzed on a regular basis to provide a “recommended lecture” thought to be effective to overcome weaknesses.

  • 実施校数(Implementation Schools):高等学校5校(『スタディサプリ』導入校)(5 High schools ( Schools implementing “Study Sapuri”))
  • 期間(Period):2016年4月より順次(April 2016 and onwards)
  • レコメンド講義配布頻度 (Recommended Lecture Frequency): およそ2週間おきに各5講義(5 lectures approximately every 2 weeks)
  • 対象教科(Target subjects):英語・数学(English, Mathematics)

80% of students responded the Recommended Lecture was helpful in overcoming weaknesses, and periodic recommendation helping self recognize academic ability


A intermediate survey was conducted 3 months after the start of the validation experiment targeting 5 classrooms of 10th grade students at one of the cooperating schools. 49.67% of students have responded that “It is now more clear what I need to do”. Reasons of “I now understand what I didn’t understand”, “I learned what my weaknesses are”, and “I was able to better understand where I stand” were given. We have found that the recommended lectures were not simply assignments, but were serving as useful reference to the student’s self recognition of their own academic ability.

続いて『危機感を感じた』(16.99%)『やる気になった』(13.07%)と多くの生徒が学事の中でレコメンド講義を受け取ることに好意的な印象を抱いている一方で、『やる気にならなかった』という回答は 15.03% にとどまりました。

In addition, many students are showing a positive impression of the recommended lecture with 16.99% responding “I now recognize the risk I am under” and 13.77% with “I am more motivated”. 15.03% of students responded that “I do not feel like doing the lecture”.

またレコメンドされた講義内容についても『苦手な部分を復習するのに役立った』と答えた生徒は約8割(79.02%)という一方で『すでに理解している内容だった』と回答された方は 20.98% となり、更なる精度の向上が求められるものの苦手を克服するためのまなび直しに寄与していることがうかがえます。

The student response to the content of the recommended lectures was largely positive and indicated effectiveness in overcoming academic weaknesses, with 79.02% responding “It was useful in reviewing parts I had difficulty with”. However, we find increasing the accuracy of the recommendation needed, as 20.98% responded “I already understood the content”.



The student population was grouped according to the survey responses noted above. The graph below indicates the change in average study time. (The upper half of the graph indicates the average value of number of questions answered in worksheets, and amount of lecture viewing time. The lower half normalizes the value per survey period for comparison). “Before Recommended Lecture” indicates the period before sending recommended lectures. “After Recommended Lecture” indicates the period following the first recommended lecture. The total collected study time includes both recommended lecture, and non-recommended lecture study materials.


For worksheets, the summer vacation homework became a source of increased study time for groups such as “I now recognize the risk I am under”. At the same time, the “I am more motivated” group shows trends of engaging with study in early parts of summer vacation, and their study time is noticeably increased in comparison to the Before Recommended Lecture period.


Visualization of the Usage Rate of Recommended Lectures, Study Behavior Pattern Analysis over discrete periods of time


the figure below indicates the ratio of recommended lectures to the total learning content. The vertical axis indicates the ratio of recommended lectures in the worksheet questions answered. The horizontal axis is the ratio of recommended lectures in the total video lecture view time. The diameter of the circles indicate the all worksheet questions answered for all lectures, including the recommended lectures.



As a general trend we find that a portion of students with less study time, indicated by smaller circles, are only studying the recommended lectures. While the overall study time is depressed following summer vacation, a certain number of students from the “It is now more clear what I need to do” group and “I am more motivated” groups are continuing to study along the recommended lectures.


Students with comparatively higher study times, indicated by the larger circles, are using recommended lectures as 30% to 50% of their study time. We find a trend where students of the “It is now more clear what I need to do” group more engage with the worksheets of the recommended lectures, indicated by a distribution along the vertical axis. Students of the “I am more motivated” group are more engaged with viewing the recommended lecture video, indicated by a distribution along the horizontal axis.


Our hypothesis is that students of the “It is now more clear what I need to do” group works with the worksheet questions to verify their own understanding of the subject, and the “I am more motivated” students are viewing the lecture to review and relearn material. We will be continuing our work to validate this hypothesis.


This validation experiment is scheduled to continue until the end of the year (2016 fiscal year). We will continue to analyze and validate the effect of Periodic distribution of the recommended lectures on changes to student motivation and student academic ability improvement.



Please contact from inquiry site under “Contact Us” on the demonstration experiment ※.