A researcher wants to collect data to create a confidence interval. Which change can the researcher make to increase the level of confidence?

Using a different estimator may not necessarily increase the level of confidence. While some estimators can be more efficient or unbiased, simply changing the estimator does not directly influence the confidence level associated with the interval. The confidence level is primarily determined by the significance level and sample size rather than the choice of estimator.

Raising the significance level decreases the level of confidence in the interval. For instance, increasing α from 0.05 to 0.10 would reduce the confidence level from 95% to 90%, resulting in a narrower interval that offers less certainty about containing the population parameter.

Collecting another sample can improve the accuracy of the estimate and reduce variability but does not directly increase the confidence level unless it is accompanied by adjustments to the significance level or additional statistical methods. The confidence level is fundamentally tied to the chosen significance level.