A parameter θ\thetaθ has true value θ=50\theta = 50θ=50. The graph shows the approximate sampling distributions of two competing estimators, A and B, obtained by repeating the sampling process many times. Based on the graph, which statement is correct?
Two target diagrams, one for each estimator. Each dot is one of eight equally likely possible values in that estimator’s complete sampling distribution at the shown θ\thetaθ; the offsets from θ\thetaθ stay the same for every possible θ\thetaθ. The crosshair marks the true parameter value. The left panel belongs to Estimator 1 and the right panel to Estimator 2.
Approximate sampling distributions of two competing estimators of the same parameter
Estimator A appears unbiased and B biased: A is centred near 50, B to the right of 50
Neither is unbiased, because neither distribution is a spike exactly at 50
Both are unbiased, because both curves look roughly symmetric and bell shaped
Estimator B is unbiased, because its curve reaches nearly the same peak height as A
Unlock full course notes by signing up
Unbiased Estimators: difficulty 3
Let X1,…,X8X_1,\ldots,X_8X1,…,X8 be an independent random sample from a population with variance σ2=64\sigma^2=64σ2=64. A student estimates that variance using σ^2=18∑i=18(Xi−Xˉ)2\hat\sigma^2=\dfrac18\sum_{i=1}^{8}(X_i-\bar X)^2σ^2=81∑i=18(Xi−Xˉ)2 rather than the usual n−1n-1n−1 divisor. Find the bias of σ^2\hat\sigma^2σ^2, defined as E(σ^2)−σ2E(\hat\sigma^2)-\sigma^2E(σ^2)−σ2.
Give your answer in same units as the variance.