Rate coefficients for collisional processes such as rotational and vibrational excitation are essential inputs in many astrophysical models .
When rate coefficients are unknown , they are often estimated using known values from other systems .
The most common example is to use He-collider rate coefficients to estimate values for other colliders , typically H _ { 2 } , using scaling arguments based on the reduced mass of the collision system .
This procedure is often justified by the assumption that the inelastic cross section is independent of the collider .
Here we explore the validity of this approach focusing on rotational inelastic transitions for collisions of H , para-H _ { 2 } , ^ { 3 } He , and ^ { 4 } He with CO in its vibrational ground state .
We compare rate coefficients obtained via explicit calculations to those deduced by standard reduced-mass scaling .
Not surprisingly , inelastic cross sections and rate coefficients are found to depend sensitively on both the reduced mass and the interaction potential energy surface .
We demonstrate that standard reduced-mass scaling is not valid on physical and mathematical grounds , and as a consequence , the common approach of multiplying a rate coefficient for a molecule-He collision system by the constant factor of \sim 1.4 to estimate the rate coefficient for para-H _ { 2 } collisions is deemed unreliable .
Furthermore , we test an alternative analytic scaling approach based on the strength of the interaction potential and the reduced mass of the collision systems .
Any scaling approach , however , may be problematic when low-energy resonances are present ; explicit calculations or measurements of rate coefficients are to be preferred .