1. What is the big data around us?
"Big data" has penetrated into every aspect of our lives. For example, when we open the mobile phone Taobao, the interface presented to us is different. The goods it pushes to us are different, and these goods can often really grasp our needs and psychology. Why?
This is actually the conclusion of big data analysis.
Taobao, a platform, can easily get a lot of our information by analyzing all the data of everyone who has browsed and bought goods.
For example, our gender, age, family members, preferences, whether we are married or not, whether we have children, the gender of our children, and even whether you like to wear casual clothes, small fresh clothes or professional clothes. After collecting these data through your every operation, it analyzes and processes them, further infers the goods you may order, and then pushes them to you, so that you can spend less time searching and more money.
For example, if you buy some products for pregnant women, it may soon push some related baby products for you.
And the evaluation and feedback after our consumption let them constantly improve themselves, such as diamond stars from different sellers, or the removal of some unqualified sellers and so on. This is Taobao's adjustment to itself.
This mutually beneficial double-ring operation mode can be regarded as a benign interaction between sellers and buyers, which is unimaginable and difficult to realize in traditional stores.
2. What is the interactive way of classroom teaching?
Classroom teaching interaction refers to a way for teachers and students to exchange information in the classroom.
In the traditional classroom, the interaction between teachers and students is relatively simple, and the classroom is a one-way conduction process in which teachers are talking and students are listening.
Some people say that teachers are porters of knowledge, and there is little communication between teachers and students in class.
Another concept is that teachers ask students questions and students answer them. This is teacher-student interaction.
Obviously, this kind of understanding is superficial and will make the interaction between teachers and students become a mere formality. The fundamental purpose of teacher-student interaction is to guide and cultivate students' higher-order thinking.
Therefore, the real interaction between teachers and students should be defined as the collision of thinking and the source of wisdom sparks.
The teaching and learning methods of Khan Academy mentioned in recent years have attracted attention because they have solved the problem of classroom teaching interaction based on big data analysis.
Big data can realize classroom teaching interaction because it has three main characteristics: feedback, personalization and probability prediction.
Our traditional classroom teaching is a single cycle of learning, that is, teachers give and students accept. We examine the students and then evaluate them.
We will not and have no conditions to reflect on the appropriateness of our teaching contents or methods through students' achievements.
We can't get really useful feedback from students to change our teaching content and behavior.
Therefore, the traditional classroom teaching is a single cycle, and there is no benign interaction between teachers and students at all.
In addition, we consider the average level of students in the arrangement of teaching content, which may not exist in reality.
In other words, our teaching doesn't take care of "good" students, and ignores those "poor" students, even those we think are middle-level students, because they are an average imaginary group.
Therefore, there is no personalized design for students in our teaching, which is the choice that education popularization has to make.
Traditional teaching is a kind of teaching with no feedback or little feedback (no time or care, no skills of being in two places at once), no personalization, and no probability prediction.
The new interactive way of classroom teaching under big data can change this situation.
1. Reference case
Victor Mayer-schoenberg and Kenneth Cookeye wrote "Walking with Big Data-The Future of Learning and Education", which gives an example of Khan Academy.
In 2004, Khan was a fund analyst who just graduated from Harvard Business School for one year, tutoring his cousin in mathematics.
Because they lived in different cities, he tutored her on the Internet, which changed the world of education forever.
He wrote several programs to assist in teaching. These programs can generate math exercises and show whether the answers submitted by children are correct.
At the same time, it collects data. The program can track the number of exercises that each student answers correctly and incorrectly, and the time they spend on homework every day.
Khan Academy, which was established on this basis, is famous for collecting data of students' behaviors, obtaining useful information from it to change the design of teaching content and customize personalized learning programs for each student.
It can be said that data is the core of the operation of Khan Academy. With the support of big data and the rapid development of Internet technology, effective classroom teaching interaction has been formed between teachers and students thousands of miles apart.
It has changed our traditional understanding that face-to-face interaction can be realized.
In addition, there is an example about Andrew Ng of Stanford University and his machine learning course.
Professor Wu put the course online, and he used videos to track the students' interactions.
Where did he press the pause button, where did he press the repeat button, and where did he give up and continue his lecture? His purpose is not to urge students to study, but to reflect on what problems students are stuck in and what teaching content is difficult to understand, so as to adjust the curriculum.
For example, he found that students usually study online in order, but many students will go back to the math review class in Class 7.
So he found out that when solving problems in lesson 7, he needed to use a mathematical formula reviewed in lesson 3, but many students couldn't remember it, so he changed the teaching video of lesson 7 and a window would automatically pop up to help students review the mathematical formula.
On another occasion, he found that the normal learning order was disrupted when students were studying lessons 75 to 80, and students watched these lessons repeatedly in various orders.
Through repeated analysis, he found that students' behavior is to repeatedly understand concepts, so he made this part of the teaching content more detailed and more helpful to help students understand concepts.
evaluate
This is a typical example of classroom teaching interaction change to achieve teaching feedback under big data analysis.
I think our traditional teaching can't get these dynamic data only by judging students' homework and looking at their exam results every day, let alone valuable information that changes our teaching content and methods.
So our teaching may repeat the same content and actions for years or even decades. Because we don't know how students study.
2. Reference cases
Another example is the summer school program of Peninsula University, which uses the mathematics courses of Khan Academy to teach middle school students in poor communities in the San Francisco Bay Area.
At the beginning of the course, a seventh-grade girl was always at the bottom of the class. She was the slowest student during most of the summer vacation, but after the course, she was the second in the class.
Khan was curious about this, so he retrieved her complete study record and looked at the time of each exercise and problem solving. The chart created by the system described her study and found that he lingered at the bottom of the class for a long time until he suddenly rose in a straight line at an incident point, surpassing almost all the students.
This fully shows that even a "poor student" who seems to have no ability can become an excellent student when students study according to their most suitable rhythm and order.
evaluate
This is a typical example of personalized teaching under big data analysis.
If we put this girl in our traditional teaching class based on small data, and her test scores are not satisfactory several times, she may be classified as a "poor student" by us, so all kinds of counseling and counseling will completely hit her self-confidence, and the shadow of her grades will even affect her life.
Khan Academy's courses use data to monitor all her learning processes. Time is a continuous variable. Design exercises suitable for her according to her characteristics, step by step, and stimulate her maximum energy.
According to this personalized customization, she studies at her own pace without paying attention to others' learning progress and achievements. I dare not think about it. I wonder how many such talents have been killed by our education.
We should truly realize the changes in classroom teaching interaction brought by big data, which are often not even technical, but conceptual.
On the basis of feedback and personalization, the greater advantage of big data is reflected in probability prediction.
For example, we can predict the behavior that a single student needs to do in order to improve his academic performance. For example, choose the most effective teaching materials, teaching style, feedback mechanism and so on.
In fact, in the era of small data, we told parents some suggestions, such as that your child should strengthen the study of mathematics, and that your child is suitable for studying liberal arts. In fact, it is not a definite fact, but a probabilistic intervention.
Because it may be based on the teacher's so-called experience, this student chooses to study liberal arts and is more likely to be admitted to one in the future. The biggest difference between big data and the past is that we speak with higher accuracy by measuring and quantifying things. Its prediction accuracy is high.
For example, in the course selection of universities, you can predict which course you choose will have a higher pass rate and how your future career planning will be smoother according to your previous learning foundation and learning behavior.
This kind of probability prediction through big data seems to have no direct relationship with the changes in classroom teaching interaction.
But after careful analysis, it is not difficult to find that this prediction is actually a continuation of the interaction between teachers and students. Our influence on students is not limited to the classroom, but continues to the level of future choices, making interaction and communication by going up one flight of stairs.
1. Using data feedback information to adjust classroom teaching strategies
Take the college entrance examination preparation as an example:
The above figure tracks the scoring rate of all students in a senior high school for four years. We can see that the scoring rate of some knowledge points remains at a high level.
This shows that the school's consistent training strategies and daily teaching methods are correct, and only need to be maintained. Teachers and students don't need to be too anxious, because the results of big data feedback can predict the future teaching effect.
2. Pay attention to the individualized development of students.
Big data not only carries out full sample analysis on huge data, but also obtains universal laws. More importantly, it can reflect personality. It can record the changes of each student and facilitate teachers to adjust classroom teaching methods for each student.
The above picture shows a student in an exam given by the big data analysis system. As can be seen from the picture, mathematics and physics are the dominant subjects of this student, while English is the weakest subject of this student. Therefore, when making improvement strategies, we should listen to the suggestions of English teachers.
Big data can help teachers' classroom teaching behavior. Unlike traditional classrooms, which are aimed at so-called "ordinary" students, they can take care of every student.
For example, using information technology to monitor students' classroom tests and classroom exercises can review any student's process at any time and count the problems in each student's process, so that teachers can adjust their judgment on the classroom process at any time according to the actual situation rather than experience.
In a word, the change of classroom teaching interaction mode should not only be a technical change, but also a very mature media technology and network platform. The changes we need are organizational changes and ideological changes.
The popular micro-courses and massive open online courses are actually the tip of the iceberg where big data penetrates into the field of teaching interaction. Form is not important. What is important is the students' behavior reflected by the data hidden in these tables and the teaching information fed back to teachers, thus triggering their thinking and change, forming a two-way cycle and realizing real "interaction". This is the real value of big data.
Teachers under big data should have "data literacy". We need to track students' progress by reading data and explain what is the most effective learning for students through probability prediction.
I think this should mean that we need to establish a perfect system. In this system, there are experts in data processing, analysts who interpret and analyze data, and teachers who use data to improve teaching.
Only in this virtuous circle system can we truly realize classroom teaching interaction, present personalized teaching and educate every child.
I hope that our education and teaching can really change because of big data.