Write down the frequencies for each category of data or each bar on the graph. For example, we would write 14 for the 90-to-100 test score bar.
Find the midpoint value on the horizontal axis of the histogram. In the 90-to-100 bar, this equates to a midpoint of 94, which is the 5th value in a series of 11 numbers.
Multiply each frequency count by its respective bar's midpoint value. In the test score example, we would multiply 14 by 94 for a total of 1316.
Add together all of the frequency times midpoint value calculations. Divide this number by the sum of all of the frequency values alone. In the test score example, if ten students scored between 80 and 89, and 7 students scored between 70 and 79, we would add (14*94)+(10*84.5)+(7*74.5)=2682.5. We would then divide this number by (14+10+7=31), for an estimated mean of 86.5.
Visually check whether your calculated mean makes sense by finding the visual balance point of the bar chart. Look for the point on the chart where the area of the bars is approximately equal on both sides. This should be reasonably close to your estimated mean.