Means

Sample median.

Margin of error.

Population mean.

Sample proportion.

Substitute outlier values with mean values to minimize their effect on interval computation.

Increase your confidence level to account for extra variability introduced by outliers.

Apply robust statistical techniques like trimming or Winsorizing before computing the interval.

Compute the interval as usual and then adjust its width subjectively based on outlier magnitude.

2.33

1.96

2.58

2.81

Increased probability of non-significant results, decreasing the effect size

A narrowed confidence interval, leading to a higher chance of Type I error

Increased likelihood of correctly rejecting the null hypothesis despite violations of assumptions

The widened confidence interval, increasing the risk of falsely stating precision

Confidence level

Sample mean

Sample size

Population standard deviation

Binomial distribution

t-distribution

Chi-squared distribution

Normal distribution

Exactly 95% of all individuals in our sample will have values within this interval.

We expect that if we took many samples, about 95% of them would produce intervals containing the true population mean.

The true population means are found within this range with absolute certainty.

There is only a 5% chance that our single calculated interval does not contain the true population mean.

Sample proportion.

Population standard deviation.

Sample size.

Level of significance.

Block design

Matched pair design

Completely randomized design

Replicated Block Design

The range within which the population mean falls.

The range within which the sample size falls.

The range within which the standard deviation falls.

The range within which the sample mean falls.