QBUS5001 Chap.7 Estimation and Confidence Intervals
Estimation and Confidence Intervals
This chapter builds confidence intervals as a point estimate plus or minus a critical value times standard error. It covers means, proportions and sample-size planning, with attention to Normal versus Student t choices. Interpretation stays with the population parameter and the long-run procedure.
Precision improves with sample size, but no narrow interval can compensate for biased collection or dependence ignored by the standard error. The margin of error separates the observed estimate from each endpoint and combines a critical value with sampling variability. Higher confidence uses a larger critical value and therefore a wider interval when the data are fixed.
Larger samples reduce standard error through a square-root relationship, so halving the margin of error usually requires far more than doubling the sample. Mean intervals choose between a known-spread Normal structure and an estimated-spread Student t structure. Proportion intervals use a different standard error based on the estimated share and require suitable success and failure counts.
Sample-size planning works backward from a desired margin of error and must round upward. An interval describes a long-run method that captures the fixed parameter at the advertised rate, not a changing posterior probability about this one completed interval.
What this chapter covers
- 01
Point estimates and margins of error
- 02
Confidence as long-run procedure performance
- 03
Mean intervals with known or estimated variability
- 04
Proportion intervals and count conditions
- 05
Confidence level, sample size and interval width
- 06
Sample-size determination for a target margin
Worked example · free
Construct an interval for a population mean
- +1Use sample mean 38.2 as the point estimate.
- +1Estimate standard error as 8 divided by square root 64 equals 1 minute.
- +1Use the approximate 95% critical value 1.96 and margin 1.96 minutes.
- +1Calculate endpoints 36.24 and 40.16 minutes.
- +1Interpret the interval for the population mean represented by the sampling process.
Key terms
- Point Estimate
- A single sample-based value used to estimate a population parameter.
- Margin of Error
- The critical value multiplied by standard error in a confidence interval.
- Confidence Level
- The long-run proportion of intervals from a procedure that capture the fixed parameter.
- Critical Value
- A distribution cutoff chosen to produce the desired confidence level or test size.
- Sample Size
- The number of sampled units, which affects standard error through a square-root relationship.
Estimation and Confidence Intervals FAQ
What does 95% confidence mean?
It describes the procedure: across many comparable random samples, about 95% of intervals built this way would capture the fixed population parameter. It does not mean this already calculated interval has a 95% probability of containing a moving parameter.
Why does higher confidence make an interval wider?
Higher confidence requires capturing the fixed parameter across a larger share of repeated samples, so the procedure uses a more distant critical value. With the same estimate and standard error, that larger multiplier increases the margin of error.
How do I reduce the margin of error?
Increase sample size, reduce measurement variability through better design or accept a lower confidence level if appropriate. Because standard error changes with the square root of sample size, halving a margin usually requires roughly four times as many observations.
Exam move
Memorise the structure rather than isolated formulas: estimate, critical value, standard error, endpoints and contextual interpretation. Compare a mean interval with a proportion interval to see where the standard error changes. Recalculate one interval at a higher confidence level and larger sample size, predicting the direction before doing arithmetic.
Round designed sample sizes upward and discuss the collection mechanism separately. Build intervals in five labelled lines: point estimate, standard error, critical value, margin of error and endpoints. This layout makes it easy to locate an arithmetic or model-choice error. Before recomputing, predict how the width changes when confidence, variability or sample size changes.
Contrast a known population spread with an estimated sample spread so the Normal versus Student t decision is tied to information, not habit. For proportions, check that the estimated counts supporting the approximation are adequate. Rewrite every interpretation until it names the population parameter and the repeated-sampling procedure. Then add one design caveat about random selection or dependence.
That final caveat prevents a precise numerical interval from being presented as stronger evidence than the collection process can support. Audit the endpoints against the point estimate. For a symmetric textbook interval, the estimate should sit midway between them and each endpoint should be one margin of error away.
Then compare the interval with any operational threshold without claiming that crossing the threshold proves equality. The interval describes estimation uncertainty and should be read alongside practical decision tolerances.
Working through Estimation and Confidence Intervals in QBUS5001? Sia is AskSia’s AI Statistics tutor — ask any QBUS5001 Estimation and Confidence Intervals question and get a clear, step-by-step explanation grounded in how QBUS5001 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.