Metode Statistika | Pendugaan Parameter (Part 2) | Penduga Selang

Interval Estimation and Confidence Intervals

The video explains why interval estimation is useful alongside point estimation when estimating a population parameter.

• A point estimate gives a single value, but it is unlikely to match the true population parameter exactly. By itself, it also does not indicate how close the estimate may be to that parameter.
• An interval estimate gives a range of plausible values for the parameter. For example, an estimated voter proportion of 55.5% might be accompanied by an illustrative interval from 52% to 56%. The lower and upper endpoints are the interval’s limits.
• An interval estimate is called a confidence interval because its interpretation relies on probability. A 95% confidence level means that if we repeatedly took samples and constructed an interval from each one, roughly 95 out of 100 intervals would cover the true parameter; about five would not. It does not mean that a particular interval has a 95% chance of containing the parameter.
• Constructing a confidence interval requires information about the sampling distribution of the point estimate. For example, an interval for a population proportion requires the sampling distribution of the sample proportion, while an interval for a population mean requires that of the sample mean.

The general form is:

Point estimate ± margin of error

• Lower limit = point estimate − margin of error
• Upper limit = point estimate + margin of error

The margin of error indicates the estimate’s precision and is described as a critical (z)-value multiplied by the estimate’s standard error.

The presenter says that practical examples of constructing confidence intervals will be covered in a subsequent video.

Speakers and Sources

• Speaker: One presenter, whose name is not given in the subtitles.
• Other sources: No other speakers or sources are identified. The video uses an illustrative election-vote example and refers to KPU results.