You posted nothing about what you already understand or what you've already tried, so I don't know how to help you.
I can make some observations; they might spur your thought process to forge a way to either a solution or a specific question.
S1 = {5, 5, 5, 5, 5, 5, 5, 5} yields mean: 5 and average absolute deviation: 0
If you don't understand this, stop now.
S2 = {1, 5, 5, 5, 5, 5, 5, 9} yields mean: 5 and average absolute deviation: 1
What I did here is this:
5 + 5 + 5 + 5 + 5 + 5 + 5 + 5 = 40
To maintain the mean of 5, the data set must always sum to 40.
Change 5 + 5 into 1 + 9. That won't alter the sum.
Replacing two elements of 5 with elements 1 and 9 gives set S2: {1, 5, 5, 5, 5, 5, 5, 9}.
The mean is still 5, and now two elements of S2 deviate from the mean by 4 each.
(4 + 0 + 0 + 0 + 0 + 0 + 0 + 4)/8 = 8/8 = 1
In other words, there is 2*4 in absolute deviation, and 2*4/8 is the average absolute deviation: 1.
Where might this train-of-thought go next ?
I mean, what might happen, if we replaced more pairs of element 5 with 1s and 9s ?