I don't completely understand how to approach these questions. I suppose the notation $\oint_C$ is something I'm not used to.
So far, I have $\oint_C\frac{dz}{z-2} = \log(z-2)$. From here, I suppose I could utilize $|z-2| = 4$ and $z-2 = re^{i\theta}$ somehow, but I'm not completely sure.
Is it the matter of that that $r = 4$, and $\theta = 2\pi$ and $\theta = 0$ (considering we're going over the entire region? Then, we have $\log(4e^{i2\pi}) - \log(4e^{i0}) = 2\pi i+\log(4) - (i0 + \log(4)) = 2\pi i $
I think I may be right. I know the answer is correct, but I don't know if my approach was necessarily correct.
By definition,
$$\oint_C f(z) dz = \int_a^b f(\gamma(t)) \gamma'(t) dt$$
where $\gamma : [a, b] \to \mathbb{C}$ is a parameterization of $C$ (provided $C$ is nice, as occurs in this case). This is simply a matter of definition.
Now in order to compute the integral, finding an antiderivative won't really work as you've tried. The problem is that $\log{(z - 2)}$ cannot be made holomorphic (differentiable) away from $2$ - you'll still have to have a branch cut, so it's not even an antiderivative anymore. Note that $e^{2\pi i} = e^0$ in the complex plane, so if we have
$$\ln{e^{2\pi i}} = 2\pi i \text{ while } \ln{1} = 0$$ then $\ln$ isn't even a single-valued function anymore. This is why we'd have to take a branch cut or view the logarithm as multi-valued.
The general technique for solving these problems is to write down a parameterization and make it an integral over some real interval. A circle can be parameterized very nicely; let's take
$$\gamma : [0, 2\pi] \to\mathbb{C}$$
defined by $$\gamma(t) = 2 + 4e^{it}$$
Then
\begin{align*} \oint_C f(z) dz &= \int_0^{2\pi} \frac{1}{(2 + 4e^{it}) - 2} 4ie^{it} dt \\ &= \int_0^{2\pi} \frac{4ie^{it}}{4e^{it}} dt \\ &= i \int_0^{2\pi} dt = 2\pi i \end{align*}