A Problem with Point Estimation: The value of the statistic alone provides us with no information about the accuracy of the estimate.
eg. A coin flipping experiment. Consider your three different results,
Each of these is a point estimate for p = probability of head.
**Because any statistic is a random variable, it has variability inherent in its value. We need to take that variability into account.
Attempt |
p-hat |
Trial 1 |
0.45 |
Trial 2 |
0.5 |
Trial 3 |
0.482 |