Chebyshev's theorem says
[tex]P\left(|X-\mu|\le k\sigma\right)\ge1-\dfrac1{k^2}[/tex]
That is, the probability that a data point [tex]X[/tex] falls within [tex]k[/tex] standard deviations of the mean [tex]\mu[/tex] is at least [tex]1-\dfrac1{k^2}[/tex].
In this case, we have
[tex]P\left(|X-400|\le1.6\cdot40\right)\ge1-\dfrac1{1.6^2}\approx0.6094[/tex]
so that at least 60.94% of the values can be expected to lie within 1.6 standard deviations of $400.
