#log-likelihood-ratio

A sequential probability ratio test decides between two hypotheses about i.i.d. measurements: $m \sim \mathcal{N}(+1, \sigma^2)$ or $m \sim \mathcal{N}(-1, \sigma^2)$, with $\sigma = 0.5$. Each measurement adds its log-likelihood ratio $\log\frac{p(m \mid +1)}{p(m \mid -1)}$ to the accumulated evidence. How much evidence does one measurement $m = 0.3$ add?