An experiment is conducted to determine the unknown rotational inertia \(I\) of an irregularly shaped rotor mounted on a fixed horizontal axle.

A light string is wrapped around a cylindrical hub of radius \(r\) fixed coaxially to the rotor, and a block of mass \(m\) is suspended from the free end. The block is released from rest and descends with a constant downward acceleration \(a\), and the student calculates \(I\) from the measured acceleration using the frictionless model \(I = m r^2 \left(\dfrac{g}{a} - 1\right)\).

If the axle bearings exert an unmodeled, constant frictional torque \(\tau_f\) that opposes rotation, how does the student's calculated rotational inertia compare to the true rotational inertia of the rotor, and what is the physical justification?