Answer:
The sample statistics are attached.
First, we need to determine the hypothesis.
The null hypothesis should be:
.
The alternative hypothesis should be: 
The next step is about calculating the<em> t statistics </em>based on the samples. This <em>t test </em>will give the probability value or p-value, which determines if there's enough to reject the null hypothesis. The formula we need to use to find the t-value is: 
Replacing all the values into the formula, we have:

Therefore, the t-value is -3.2.
Now, we using a level of significance of 0.01, and a degree of freedom (df) of 34, we use the t-table to find the p-value for this results. (df = N -1; in this case, we take the smaller sample, which is 35, giving us 34 of df).
Therefore, according to the table attached, <em>the p-value is less than 0.01,</em> which is less than our level of significance. This result means that the null hypothesis is rejected. In other words, there's enough evidence to say that <em>the life expectancy of people in Africa is less than the life of expectancy of people in Asia.</em>
<em></em>