Hypothesis Confidence Interval
A capped interval around a point estimate contracts onto its estimate as precision improves, and the fixed true-value rule it straddles stays inside the caps when the interval settles.
CSS
Hypothesis
15 animationsHypothesis Effect Size
The gap between two group means is held fixed while their spread widens and tightens, and the unit comb beneath the bracket shows how many standard deviations fit inside that gap.
CSSHypothesis Equivalence Test
A capped interval widens past both equivalence bounds and draws back until it lies wholly inside the shaded band, lighting the verdict mark for no meaningful difference.
CSSHypothesis Error Trade-Off
A sliding critical threshold between the null and alternative curves swaps the two shaded error regions, the alpha tail growing exactly as the beta region shrinks and back again.
CSSHypothesis False Positive Risk
A scan runs down a hundred-cell grid of tests and lights every significant result: eight from the banded row that holds real effects and five from the ninety nulls below it.
CSSHypothesis Interval Coverage
Eighteen capped intervals from repeated samples are wiped into view down a column against one true-value rule, and the single interval that fails to cover it is the one left lit.
CSSHypothesis Multiple Comparisons
Twenty test bars stand against one criterion rule; dropping the rule to the uncorrected level lights four more findings, and raising it back to the corrected level leaves only the one that was never borderline.
CSSHypothesis Null and Alternative
Two distribution curves share one baseline; the alternative slides away from the null while the dimension rule between their means stretches to measure the effect, then settles back into overlap.
CSSHypothesis One-Tailed and Two-Tailed
A standardised axis holds the observed statistic still while the rejection region switches from a single right-hand band to a symmetric pair, pushing the critical value past the statistic and flipping the verdict.
CSSHypothesis P-Value Tail
A null distribution carries a vertical observed-statistic marker; as the marker travels out along the curve the shaded tail area beyond it is revealed by a sliding mask and shrinks to a sliver.
CSSHypothesis Permutation Test
Group labels swap sides of the divider and the shuffled null histogram rebuilds itself from the baseline, leaving the observed statistic standing alone out past its tail.
CSSHypothesis Power Curve
A power curve lights up across the sample-size axis as the guide rule advances, stopping where the curve meets the eighty-percent target and marking the sample size that reaches it.
CSSHypothesis Sample Size
A comb of specimen ticks fills in discrete enrolment batches from the axis and stops exactly on the marker for the sample size the study was sized for.
CSSHypothesis Test Statistic
A raw difference measured against a one-standard-error bracket is traced down onto the standardised reference axis, where it lands beyond the critical mark.
CSSHypothesis Type I and Type II Errors
A two-by-two grid of truth against decision holds the correct diagonal recessive while the false-positive and false-negative cells take the accent in turn.
CSS