Is acceleration due to gravity constant? I was taught in school that acceleration due to gravity is constant. But recently, when I checked Physics textbook, I noted that
$$F = \dfrac{G  m_1   m_2}{r^2}. $$
So, as the body falls down, $r$ must be changing, so should acceleration due to gravity.
 A: This is a first introduction to the issue of the relative changes in physics.
Consider the motion of objects near the Earth's surface. Call the nominal radius of the Earth $R \approx 6400\text{ km}$, and the height of the object $h$.
Now the acceleration due to gravity at $h$ is 
$$ g = \frac{F_g}{m} = G\frac{M_e m}{(R + h)^2 m} = G\frac{M_e}{(R + h)^2} $$and lets manipulate this a little
$$ g = G\frac{M_e}{(R + h)^2} = G\frac{M_e}{R^2(1 + h/R)^2} \approx G\frac{M_e}{R^2}\left(1 -2\frac{h}{R} \right).$$ 
The last approximation there is dropping higher order terms in $\frac{h}{R}$ which will shortly be seen to be justified.
So, ask yourself how big is $\frac{h}{R}$ for the situations you encounter in your life. A few meters or a few tens of meters at most, right? So $\frac{h}{R}$ is of order $10^{-5}$ or smaller over human scales or $10^{-3}$ even over the whole height range that we use including airplane elevations.
So, for almost all calculation that you want to make the variation of $g$ negligible.

Physicists get a lot of millage out of these kinds of considerations to the point that you there is a fair amount of shorthand devoted to discussing fractional changes. People say things like "Yeah, but it's down by two orders of magnitude, so we can neglect it".
