I am not certain what you want to do.
For the boxplot, try this:
D = load('structure.mat');
S = D.S;
for k1 = 1:numel(S)
for k2 = 1:numel(S(1).s)
t(k1,k2) = S(k1).s(k2).time;
end
end
figure
boxplot(t.')
grid
xlabel('S')
ylabel('time')
figure
boxplot(t)
xlabel('time')
ylabel('S')
Fh = gcf;
pos = Fh.OuterPosition;
Fh.OuterPosition = pos + [0 -500 750 500];
The loop was necessary to recover the information from your structure.
I have no idea what you are doing, so I likely cannot help on any further statistical analyses.
