# Can % error in computed value be less than % error in one of its parameters

I am using an equation to compute energy as follows

$E= C_1\times t + C_2 \times a$

Here $C_1$ and $C_2$ are constants. $t$ shows the time-taken, $a$ shows the dynamic-activity.

Using a technique, I am estimating $t$ and $a$ and using them to estimate $E$. After estimating $t$ and $E$, I compare them with their actual values, obtained through actual measurement. I see % error in $t$ as, say 10% and that in $E$ as, say 7%. Is it mathematically correct and acceptable. There is also some error in estimation of $a$. Does the answer depend on $C_1$ and $C_2$ also? Thanks in advance for your help.

-

If the $C_1t$ term makes a relatively small contribution to $E$ compared $C_2a$, then any given absolute error in $t$ is going to be a larger fraction of $t$ than it is of $E$. So it will look like a larger relative error in $t$ than the resulting relative error in $E$.