The $\mathrm{p}K_\mathrm{a}$ of indicators can be determined via plotting a graph of $\log \left(\frac{\ce{In-}}{\ce{HIn}}\right)$ vs $\mathrm{pH}$ with the help of a spectrophotometer, from the equation $\log \left(\frac{\ce{In-}}{\ce{HIn}}\right) = \mathrm{pH} - \mathrm{p}K_\mathrm{a}$.
From this, the graph should have a gradient of 1, and $x$-intercept = -($y$-intercept) = $\mathrm{p}K_\mathrm{a}$. However, my graph has a slope of ~0.8 which means my $x$-intercept ≠ -($y$-intercept) so which value do I use as my calculated $\mathrm{p}K_\mathrm{a}$?
From the papers I've read, most if not all of them use the intercept with the $\mathrm{pH}$ axis instead of the $\log \left(\frac{\ce{In-}}{\ce{HIn}}\right)$ axis in this situation, why is this?