If a likelihood ratio test for $H_0: \theta = \theta_0$ vs $H_a: \theta \neq \theta_0$ yields a statistic $\Lambda = -2 \ln(L(\theta_0)/L(\hat{\theta})) = 3.841$, and $\theta$ is a single parameter, what is the approximate p-value?