The two-sided alternative hypothesis is appropriate in this case, the reason being we are asked "does the data indicate that the average body temperature for healthy humans is different from 98.6◦........?".
The test statistic is:
Using an inverse normal table, and halving
for a two-tailed test, we look up 
and find the critical value to be Z = 2.5758.
Comparing the test statistic Z = -5.47 with the rejection region Z < -2.5758 and Z > 2.5758. we find the test statistic lies in the rejection region. Therefore the evidence does not indicate that the average body temperature for healthy humans is different from 98.6◦.
The test statistic is:
Using an inverse normal table, and halving
Comparing the test statistic Z = -5.47 with the rejection region Z < -2.5758 and Z > 2.5758. we find the test statistic lies in the rejection region. Therefore the evidence does not indicate that the average body temperature for healthy humans is different from 98.6◦.