Tutoring, part 3 of 3
What moves a student's results
Raising a score and changing someone's options are different jobs.
Context
Over two years I taught students preparing for very different systems, from school entrance exams up to Oxbridge interviews. A large share of it was SAT preparation for international students: more than 800 hours of teaching with more than 30 students.
The problem
Different systems meant different definitions of “ready.” An SAT student needs speed and accuracy under a clock. An A Level student needs depth on a syllabus. An Oxbridge applicant needs to think out loud in front of a stranger and stay calm when they don’t know the answer.
And many of my students were applying into systems their families hadn’t been through. They had the ability. What they didn’t have was someone who knew how the thing actually worked.
What I did
- One structure for everyone. Every student worked through the same sequence. That made it obvious where each person was, and where they were stuck.
- Tracking mistakes, not scores. Scores tell you how a student did. Mistakes tell you why. I kept a record of every error and taught to the patterns behind them, not the individual questions.
- Owning the time between lessons. Most of the improvement happens when the tutor isn’t there. I set and checked work between sessions as carefully as I ran the sessions themselves.
- Mixing two teaching cultures. I combined the intensity of Korean exam preparation with the flexibility of UK and US teaching: high expectations, with room to adapt to each student.
- in the lessonTeach the lesson
- afterLog every mistake
- afterFind the pattern behind them
- between lessonsSet work on that pattern
- between lessonsCheck it before the next lesson
Then again, with the next pattern
About 80% of my SAT students finished at 1520 or higher. In the first year I prepared students for Oxbridge, three of the five received offers.
What I learned
Getting a test right myself and getting someone else there are different skills, and the second is the one I cared about more.
The other thing I noticed: what moved results was rarely the explanation itself. It was the structure around it, the sequence, the mistake log, the work between lessons. That observation is what made me think a platform could carry some of the teaching without me.