Consider a continuous random variable X with expectation 15 and variance 4. Determine the smallest interval [15 − c, 15 + c] which contains at least 90 % of the values of X.
Using Tschebyschev’s inequality (7.24)
P\left(\left|X-\mu \right|\lt c \right) \geq 1-\frac{Var\left(X\right) }{c^{2}} . (7.24)
P\left(\left|X-\mu \right|\lt c \right) \geq 0.9=1-\frac{Var\left(X\right) }{c^{2}} ,
we can determine c as follows:
1-\frac{Var\left(X\right) }{c^{2}}=0.9
c^{2}=\frac{Var\left(X\right) }{0.1}=\frac{4}{0.1}=40
c= \pm \sqrt{40} = \pm 6.325.
Thus, the interval is [15 − 6.325; 15 + 6.325] = [8.675; 21.325].