Answer
Because all the terms in $Σ\frac{(O_i-E_i)^2}{E_i}$ are positive, no matter the observed value is greater or less than the expected value. The larger is the difference between the observed and the expected values, the larger will be $(O_i-E_i)^2$ and, the larger will be $X^2=Σ\frac{(O_i-E_i)^2}{E_i}$. So, we will reject the null hypothesis only when $X^2$ is large enough, that is, when $X^2\gt X_α^2$ (right tailed test).
Work Step by Step
Given above.
