When trying to determine the ohmic resistance from measurements of current $I(U)$ at different voltages $U$ of a circuit with the ohmic resistance built in, which of the following methods evaluate the data more "accurate"?:
When a regression line is plotted with the form $y=mx$ approaching the ohmic law $I(U)=\frac{1}{R} U$ you can read out the slope $m$ and take its inverse to get the ohmic resistance $R$. At the same time you could get $R$ by taking the quotient of each individual voltage $U$ meassured and its corresponding $I(U)$ and then calculate the average.
Both ways make sense plausibly but which one is more accurate from a mathematical standpoint? And how do you determine the error $R$ in both cases?