Arhaan Anand

Peer-reviewed · Published October 2026

Tutor Quality Is Not Tutor-Student Fit: Evidence from Online Secondary-School Tutoring

Karan Gupta and Arhaan Anand · Journal of Education and Learning, Vol. 16, No. 2 (2027), pp. 36 to 51 · doi:10.5539/jel.v16n2p36

Tutoring platforms screen tutors on quality, yet families often say a tutoring relationship failed because the tutor did not suit the student. This study tests whether tutor quality and tutor-student fit are really different things, using administrative records for 220 tutoring relationships on an online IGCSE platform, 178 student questionnaires, 163 parent questionnaires and 134 matched student-parent pairs.

They are. Quality and fit were essentially uncorrelated and loaded on separate factors, and each was independently associated with student satisfaction, with fit carrying the larger weight. About a third of the variation in perceived quality lay between tutors, while fit showed essentially none: the same tutor suited some students and not others. On the parent side, perceived value and platform responsiveness were each linked to satisfaction, while the session fee added nothing once value was accounted for.

The practical conclusion: selecting better tutors and matching them better are different problems. The paper proposes a short pair-level check after the first or second lesson, asking about pace and suitability rather than the tutor's competence, to catch a mismatch before it ends the relationship.

I provided the study setting and administrative records and contributed to the design of the questionnaires. Data were collected in May and June 2026 under formal ethics approval.

LSE Students' Union Essay Competition · July 2026

Technology and Worker Wellbeing

Arhaan Anand · Essay, 5 pages

Remote and hybrid work has made employees easy to monitor, and artificial intelligence now makes monitoring cheaper and possible at far greater scale. The essay asks whether the information firms collect improves decisions enough to justify its financial, behavioural and organisational costs.

Using principal-agent theory and a simple profit model, Π = F(e,q) − C(m) − B(m), it argues that monitoring is neither harmful in itself nor a guarantee of efficiency. The real problem is measurement: time online, keystrokes and emails sent are easy to track but poor proxies for productive value, and under Goodhart's Law they become targets that distort behaviour. The essay draws on a 2022 meta-analysis of 70 samples, which linked electronic monitoring to slightly lower job satisfaction and slightly higher stress, with no relationship to performance.

It then tests the argument against its strongest counterexample. In banking, aviation and other regulated or safety-critical work, the process is part of the outcome, so close monitoring is justified. The conclusion is that monitoring should be designed around the job, with outcomes for knowledge work and process measures where safety or compliance depend on them.

"Sometimes, the most efficient monitoring system is the one that knows what not to measure."