Normal Distribution
nothing is called 95% until it has been counted
The collection

Every essay — page 4

Essays 73 to 96 of 613, in the same order.

What conditioning on a variable does

Putting a covariate into the regression is one arithmetic operation, and it is the right thing to do in one of the three worlds it could have come from. Here the three worlds are built to share a covariance matrix entry for entry, so the estimate they disagree about by 0.348 is a number no sample of any size can settle, and the rule that controls for everything measured is measured: it leaves a larger bias than controlling for nothing on 65.5% of four thousand structures.

The same covariate, three ways round. Three worlds over a treatment, an outcome and a covariate, joined by the same three edges at the same three strengths — 0.90, 0.50 and 0.70 — differing only in which way the two edges touching the covariate point. In the first the covariate causes both and adjusting for it recovers the effect of 0.50 exactly. In the second the treatment causes the covariate, the effect is 1.13, and adjusting returns 0.50 — the direct edge alone, with the part that travels through the covariate deleted. In the third the treatment and the outcome both cause the covariate, the effect is 0.50, and adjusting returns -0.087. The regression that produces those three numbers is one formula, and nothing in the data says which panel it is being run in.

One arithmetic, three decisions

A covariate beside a treatment and an outcome can be a common cause of both, a step on the path between them, or an effect of both. The regression that includes it is the same arithmetic in all three, and it is right in one — returning 0.5000, deleting 0.6300 of the effect, and turning 0.5000 into −0.0872.

6 figures · Conditioning, part 1 One distribution, two effects. Three causal structures fitted to one covariance matrix over a treatment, a covariate and an outcome. Each reproduces it exactly — the largest entry-wise disagreement across all three is 4.4e-16 — so no sample of any size distinguishes them. The regression of the outcome on the treatment and the covariate returns 0.500 in all three, to within 4.4e-16, because that coefficient is a function of the covariance and of nothing else. The effect the three worlds hold is 0.500, 0.848 and 0.848: adjusting is exactly right in the first and off by −0.348 in the other two. The arithmetic cannot see the difference and the difference is the whole question.

The two worlds that look the same

Three causal structures were fitted to one covariance matrix and agree with it to 4.4·10⁻¹⁶. The regression returns 0.5000 under all three; the effect they hold is 0.5000, 0.8481 and 0.8481. What separates structures is a missing edge, and the signature of one is a correlation of exactly zero.

6 figures · Conditioning, part 2 Two structures in three are made worse. What adjusting for every covariate measured does to the bias in the treatment's estimated effect, against adjusting for none, over 4000 randomly drawn structures of 6 covariates each. Each covariate is independently a common cause with probability 0.25, a cause of the treatment only, a cause of the outcome only, a cause of neither, a step on the causal path, or a common effect. The rule leaves a larger bias on 65.5% of structures, a smaller one on 33.8%, and the same on 0.7%. The share is a property of that population of structures rather than of adjustment, which is why the weights are stated; what does not depend on them is that the rule has no direction — it is not a conservative default that occasionally overcorrects, it is a rule whose error is whatever the structure happens to be.

Adjusting for everything

"Control for every covariate that was measured" leaves a larger bias than controlling for nothing on 65.5% of four thousand randomly drawn structures and a smaller one on 33.8%. Its squared error is 4.110 times that of using no covariate at all, and half of it sits in its worst tenth of structures.

6 figures · Conditioning, part 3 A covariate that is prior to everything and still ruins it. A covariate measured before the treatment, caused by neither the treatment nor the outcome, and not a common cause of them. Two unmeasured variables sit behind it: one reaches the treatment, the other reaches the outcome, and both reach the covariate. Every rule of thumb for including a baseline variable is satisfied, and the regression that leaves the covariate out estimates the treatment's effect of 0.50 without bias, while the regression that includes it is off by −0.2000 — because the covariate is a common effect of the two unmeasured causes, and conditioning on a common effect makes its causes dependent. The path it opens runs from the treatment back through the first unmeasured cause, through the covariate, and out through the second to the outcome.

A collider before the treatment

A covariate measured before the treatment, on no causal path, and not a common cause of anything, still biases the estimate by exactly −0.2000 against an effect of 0.5 — while the regression that leaves it out is exact. The bias saturates at 0.3536, and the two paths that make it a collider do not appear in that bound.

6 figures · Conditioning, part 4 The line is the sample, and the sample is the finding. 900 draws of two independent standard normal causes, with the 453 of them past a threshold of 0.00 marked and the 447 that fall short left pale. In the population the two are independent by construction. Inside the selected sample the correlation is -0.4669 in closed form and -0.5050 counted on these 453 rows, and the least-squares line through them has a slope of -0.545. The mechanism is visible in the picture rather than argued: the threshold removes one corner of the cloud, and a cloud with a corner missing is a cloud whose two coordinates carry information about each other.

The sample is a condition

Two independent standard normals, selected on their sum exceeding its median, read a correlation of exactly −1/(π − 1) = −0.4669 inside the sample. Nothing is measured badly and nothing is missing — and both halves of that split read it, in the same direction, while the population containing both reads zero.

6 figures · Conditioning, part 5 The covariate the treatment caused, and what it hides. A covariate on the causal path: the treatment causes it and it causes the outcome, so the treatment's total effect of 1.130 runs partly through it. Adjusting for it returns the direct edge alone, 0.500, which is what somebody wanting the total effect should not have asked for. The dashed variable is the second problem: an unmeasured cause of both the covariate and the outcome. It does not touch the treatment, so the unadjusted regression still recovers 1.130 exactly. It does touch the covariate, so once the covariate is conditioned on the treatment and the outcome are linked through it, and the adjusted coefficient lands on 0.050 — neither the total effect nor the direct one.

The variable the treatment caused

Adjusting for a covariate the treatment caused stops estimating the total effect and starts estimating the direct one. When that covariate shares an unmeasured cause with the outcome it estimates neither: the total effect is 1.1300, the direct effect is 0.5000, and the regression returns 0.0500.

6 figures · Conditioning, part 6 A proxy removes less than its reliability, always. The share of the confounding bias removed by adjusting for a proxy, against how well the proxy measures the confounder. The diagonal is the answer a reader would guess — a covariate that is 80% signal removes 80% of the problem. The curve is what the arithmetic gives: the reliability, times one minus the squared correlation between the treatment and the confounder, divided by one minus the product of those two. That squared correlation is 0.4475. A reliability of 0.8 removes 68.85% and one of 0.6 removes 45.32%. The two agree only at the ends, and the gap is widest where most applied covariates sit.

Adjusting for a shadow

A covariate that is 80% signal removes 68.85% of the confounding, not 80% — the share is λ(1 − ρ²)/(1 − λρ²) and it is below the reliability everywhere. The residual bias is 0.1084 against an effect of 0.5, and at 25,600 rows it is 17.6 standard errors wide.

4 figures · Conditioning, part 7 Three adjustment rules on four thousand structures, as every covariate gets noisier. Four thousand random structures of six covariates, the treatment's effect fixed at 0.5, every covariate observed with the same reliability. Mean absolute bias of adjusting for nothing stays 0.190; adjusting for everything measured falls from 0.377 with exact covariates to 0.279 at reliability 0.8 and 0.203 at 0.5; adjusting for the common causes rises from 0 to 0.056 and 0.118. Adjusting for everything beats adjusting for nothing on 33.8% at 1, 38.3% at 0.9, 42.0% at 0.8, 45.3% at 0.7, 48.3% at 0.6, 50.8% at 0.5. The common causes never do worse than nothing at any common reliability.

A set of shadows

Measure every covariate of four thousand random structures with reliability 0.8 and adjusting for the common causes still never does worse than adjusting for nothing — the bias left is each confounder's part times one minus its proxy's reliability, scaled alike. Adjusting for everything measured gets better, beating nothing on 42.0% of structures against 33.8% with exact covariates, because noise blunts the bad adjustments too. Spread the reliabilities to 0.75 ± 0.25 and the right set does harm on 7.15% of structures, every one of them with confounders pushing opposite ways.

4 figures · Conditioning, part 8

Corrections, and what each controls

Bonferroni bounds the chance of any false positive; Benjamini–Hochberg bounds the share of the findings that are false. Both get called correcting for multiple comparisons and they are different promises, so every procedure here is made to report both rates and the power each one costs.

The familywise error rate with no correction, α = 0.05. Two routes: the curve is 1 − (1 − α)^m and the points are counted over 6,000 simulated families of true nulls. With twenty tests the chance of at least one false positive is 64.1%.

What the correction corrects

Twenty tests of true nulls produce at least one false positive 64% of the time, and the closed form and the count agree. Bonferroni holds it at 5% and Holm holds it at 5% while finding more. Nobody should still be using Bonferroni.

6 figures · Multiplicity, part 1 Twenty tests, 10 of them real — what each procedure holds. no correction: familywise 40.8%, false discovery 5.3%, power 85%. Bonferroni: familywise 2.8%, false discovery 0.5%, power 49%. Holm: familywise 3.6%, false discovery 0.6%, power 53%. Benjamini–Hochberg: familywise 20.0%, false discovery 2.6%, power 75%.

Two different promises

Bonferroni bounds the chance of any false positive. Benjamini–Hochberg bounds the share of the findings that are false. Both are called correcting for multiple comparisons, and one of them lets the familywise rate reach 20%.

6 figures · Multiplicity, part 2 Power to find a real effect of 3 standard errors, 10 of 20 real. no correction finds 85.1%, Bonferroni finds 49.1%, Holm finds 52.5%, Benjamini–Hochberg finds 74.9%. The uncorrected procedure finds the most and controls nothing.

The price of control

Every correction is paid for in power, and the exchange rate can be measured. Holm buys familywise control for 33 percentage points of power; Benjamini–Hochberg buys a weaker guarantee for 10. Neither is free and neither is a matter of taste.

6 figures · Multiplicity, part 3 The false discovery rate of twenty correlated tests, against the correlation. BH, every null true: 5.08% at 0, 4.86% at 0.3, 3.70% at 0.6, 2.34% at 0.9. BH, 10 of 20 real: 2.55% at 0, 2.53% at 0.3, 2.26% at 0.6, 1.66% at 0.9. BY, every null true: 1.46% at 0, 1.31% at 0.3, 1.03% at 0.6, 0.69% at 0.9. BY, 10 of 20 real: 0.72% at 0, 0.75% at 0.3, 0.64% at 0.6, 0.50% at 0.9. 20,000 families at each correlation.

False discoveries that arrive together

Correlate twenty tests and Benjamini–Hochberg still holds its false discovery rate — 1.66% at a correlation of 0.9 with ten real effects, against 2.55% when the tests are independent. What changes is how the errors come. A family of true nulls reports anything 2.34% of the time instead of 5.08%, and when it does, it reports 16.56 false findings out of twenty.

8 figures · Multiplicity, part 9 Storey's estimate of the share of true nulls over twenty thousand families, independent and correlated at 0.6. The true share is 0.5. Independent tests: mean 0.610, spread 0.160, below half the truth in 0.92% of families. Correlated at 0.6: mean 0.609, spread 0.240, below half the truth in 9.33%.

Estimating how many nulls are true

Benjamini–Hochberg at 5% delivers 2.55% when half of twenty nulls are false, because it cannot tell how many are. Storey's estimate of that share, read off the p-values above one half, spends the rest and finds 81.93% of the real effects instead of 74.70% on independent tests. Correlated at 0.9, the same procedure reports a finding in 19.29% of families in which every null is true.

7 figures · Multiplicity, part 10 Twenty hypotheses tested in a declared order, the ten real effects listed first. Effects of three standard errors, ten real, familywise 5%. fixed sequence: 85.3% at position 1, 45.0% at 5, 20.4% at 10; overall power 46.10%; fallback: 49.1% at position 1, 56.4% at 5, 57.9% at 10; overall power 55.87%; Holm: 52.5% at position 1, 52.2% at 5, 52.5% at 10; overall power 52.53%.

An order that spends the error rate

Test twenty hypotheses in a declared order, each at the full 5% and each only if every one before it was rejected, and the first is found 85.3% of the time where Holm finds it 52.5%. The tenth is found 20.4% of the time, the product of the powers before it. Move one true null to the head of the list and every real effect behind it is found no more than 4.3% of the time.

6 figures · Multiplicity, part 11 Every finding Benjamini–Hochberg made in thirty families, with its interval, at real effects of 2. 85 findings, sorted by their estimate. 12 of their ordinary 95% intervals miss the true effect, every one of them on the far side; 6 of the wider false-coverage-rate intervals miss.

Intervals for the findings

Benjamini–Hochberg's findings usually go out each with its ordinary 95% interval. With ten real effects of two standard errors among twenty tests, 11.59% of those intervals miss their effect, every miss on the far side, and the interval around the most prominent finding covers 72.36% of the time — 2.38% when the effects are one standard error. Intervals widened for the number of findings hold the share that miss under 5%.

5 figures · Multiplicity, part 12 The interval for a finding's effect given that it cleared its threshold, against where its estimate landed. Threshold c = 2.8. At z just past it the selective 95% interval runs from −0.58 to 0.95; at z = 3.8, from 0.18 to 5.73; at z = 6.3, from 4.31 to 8.26, against the ordinary z ± 1.96.

An interval that knows its finding was found

An interval for one Benjamini–Hochberg finding can be read from the distribution of its estimate given that it cleared the threshold it had to clear. It covers the finding's effect 94.8% of the time with real effects of two standard errors, against the ordinary interval's 87.0%, and it covers the most prominent finding 91.3% of the time where the ordinary one covers 72.4%. The price is width — a median 5.06 standard errors against 3.92 — and something larger: for three findings in four at that effect size the interval reaches back past zero, so a finding that cleared the line is, once the line is accounted for, not distinguishable from nothing.

4 figures · Multiplicity, part 13

When the data stops early

A subject still event-free when a study ends is not missing and not observed — it is known to exceed something, which is a third state most tools have no slot for. The estimator that gives it that slot recovers the curve, and everything read off the curve afterwards has a condition attached: the interval printed around its end stops covering where the curve is read, a dropout that carries information leaves a record identical to one that does not, one minus the curve is not a risk once something else can end observation first, and a hazard ratio is an average whose weights the length of follow-up chose.

The same study read three ways. At time 2 the truth is 0.497. Kaplan–Meier gives 0.532; dropping the censored subjects gives 0.180; treating the censoring time as the event time gives 0.392. Both naive readings understate survival, because the subjects they mishandle are the ones doing well.

The data that stops early

A subject still event-free when a study ends is not missing and not observed. It is known to exceed something, which is a third state most tools have no slot for — and the two obvious ways of forcing it into one are wrong by 31 and 13 percentage points.

6 figures · Censoring, part 1 Kaplan–Meier from 120 subjects, 59 of them censored. The step curve is the estimate, the smooth curve is the truth it is trying to recover. 59 of 120 subjects were still event-free when observation stopped; they are not dropped, and they are not counted as events — they leave the risk set at the time they were last seen.

The curve that survives censoring

Kaplan–Meier recovers the true survival curve to within a fraction of a point at every censoring level from 37% to 71%, where dropping the censored subjects is off by 28 and then by 43. The estimator is a running product and the reason it works is in its denominator.

7 figures · Censoring, part 2 One cohort of 40, and two intervals around the end of its curve. A single simulated study of 40 subjects with exponential survival at rate 0.35, dropout at rate 0.15 and follow-up to 6 — the first seed from 8811 upward whose plain band reaches below −0.05, chosen to show the failure rather than its frequency. The step curve is Kaplan–Meier and the smooth curve the truth. The plain band, the estimate plus and minus 1.96 Greenwood standard errors, first dips below zero at t = 3.78 and reaches −0.052; early on it also rises to 1.023, above one. At t = 5 the estimate is 0.069 with 1 subject still under observation, the plain interval runs from −0.052 to 0.191 and the log-log interval from 0.006 to 0.251, against a truth of 0.174. The log-log band is built on a scale that cannot leave [0, 1], and it bends away from the edge rather than through it.

The interval at the end of the curve

The interval most software prints around a survival curve covers 89.7% at five years, where 3.3 of forty subjects are still being watched and where the curve is actually read. The same variance carried on a log–log scale covers 94.8% there — and the failure was never the width.

6 figures · Censoring, part 3 Two worlds, one Kaplan–Meier curve, two truths. World A gives each subject a frailty with mean one and variance 1, and multiplies both its event hazard (0.35) and its dropout hazard (0.5) by it, so the subjects likeliest to leave are the ones likeliest to fail. World B has independent event and dropout times whose hazards are world A's crude hazards. Kaplan–Meier over 1000 studies of 400 gives the same curve from both — 0.7750 and 0.7763 at t = 1; 0.6627 and 0.6641 at t = 2; 0.5924 and 0.5932 at t = 3; 0.5047 and 0.5042 at t = 5 — and that curve is world B's truth, 0.5052 at t = 5. World A's truth is 0.3636 there. The dashed lines are the two bounds that assume nothing, from every dropout failing on leaving (0.1905 at t = 5) to none ever failing (0.6667).

A dropout the data cannot see

Two worlds leave the same record to the last detail a study can write down — the same times, the same share ending in the event, the same share leaving first — and a log-rank test between them rejects at its own 5% level at every sample size from a hundred to sixteen hundred. Kaplan–Meier converges on 0.5052 at t = 5 from both. The truth is 0.5052 in one and 0.3636 in the other, and what is left to argue about is where between two bounds to stand.

6 figures · Censoring, part 4 The risk of one cause, estimated two ways. Two causes of an ending event with constant hazards 0.2 (the one of interest) and 0.3 (the competitor), random dropout at 0.1 and follow-up to 6. The lower line is the cumulative incidence, (0.2/0.5)(1 − e^(−0.5t)), the chance of actually having had this event by t; the dots on it are the Aalen–Johansen estimate over 2000 studies of 300, 0.3670 at t = 5 against 0.3672. The upper line is 1 − e^(−0.2t), and the dots on it are one minus Kaplan–Meier with the competing event treated as censoring: 0.6318 at t = 5 against 0.6321. The second is larger by a factor of 1.722 at t = 5, and it is not an error of estimation. It estimates, correctly, the risk in a population where the competing cause does not exist.

One minus Kaplan–Meier is not a risk

With two ways for observation to end, one minus Kaplan–Meier for one cause reads 0.6318 at t = 5 where the chance of actually having had that event is 0.3670. Added across the two causes, the complements reach 1.4088 — more than the whole cohort. Nothing is estimated badly: the complement estimates, correctly, the risk in a world where the other cause does not exist.

6 figures · Censoring, part 5 One treatment, a different hazard ratio at every follow-up. The hazard ratio a Cox model converges to, found as the root of its expected score by numerical integration, as the trial runs longer; dropout at 0.1 throughout. The proportional treatment reads 0.5 at every τ. The waning treatment reads 0.5000 at τ = 1, 0.6362 at τ = 3 and 0.7890 at τ = 8 — the same two arms, the same effect in the same first year, and a number that drifts towards one as later, effect-free events are added to the average. The dots are the mean of 400 Cox fits with 400 subjects an arm: 0.5014 at τ = 1, 0.7020 at τ = 2, 0.7635 at τ = 3, 0.8034 at τ = 5, 0.8208 at τ = 8. The crossing treatment reads 0.3429 at τ = 1, exactly 1 at τ = 3 by construction, and 1.0611 at τ = 8: beneficial, null or harmful according to when the trial stopped.

The hazard ratio the follow-up chose

A treatment that halves the hazard for one year and then does nothing has a Cox hazard ratio of 0.5000 if the trial stops at one year, 0.7617 at three and 0.8194 at eight. Nothing about the treatment differs between those numbers. When hazards are not proportional the hazard ratio is an average, and the length of follow-up and the dropout rate choose its weights.

7 figures · Censoring, part 6 The difference in restricted mean survival at every horizon, in three worlds. Treatment minus control, in closed form, with dropout irrelevant to the truth. The proportional treatment's difference grows to 0.4766 at τ = 3 and the waning treatment's to 0.2675. The crossing treatment's rises to 0.1776 at τ = 2, near where the two survival curves cross, and falls back to 0.1366 at τ = 3. The ticks along the bottom are the eleven horizons, from 0.5 to 3 in quarters, at which a trial below reads its differences.

A horizon chosen after looking

A difference in restricted mean survival read at whichever of eleven horizons looks most convincing rejects 11.24% of trials in which the treatment does nothing, against 4.70% at a horizon fixed in advance. The correlation of the differences across horizons is closed, and the Gaussian process it defines prices the choice at a critical value of 2.317 — which brings the counted size back to 4.99% and keeps 96.92% of the power that a horizon nobody could have known to fix would have had.

6 figures · Censoring, part 7 Power of four tests of two survival curves, in three shapes of treatment effect, at 50, 100 and 200 a side. 6,000 trials at each setting, dropout 0.1 a year, follow-up to three years. hazard halved throughout, 50 a side: log-rank 64.9%, best of eleven horizons 51.7%, restricted mean at 3 61.1%, one test for all twelve 56.6%; hazard halved throughout, 100 a side: log-rank 91.8%, best of eleven horizons 81.0%, restricted mean at 3 88.4%, one test for all twelve 86.0%; hazard halved throughout, 200 a side: log-rank 99.7%, best of eleven horizons 98.4%, restricted mean at 3 99.4%, one test for all twelve 99.3%; halved for a year, then nothing, 50 a side: log-rank 18.5%, best of eleven horizons 32.1%, restricted mean at 3 26.6%, one test for all twelve 30.8%; halved for a year, then nothing, 100 a side: log-rank 29.3%, best of eleven horizons 51.7%, restricted mean at 3 42.5%, one test for all twelve 50.2%; halved for a year, then nothing, 200 a side: log-rank 51.4%, best of eleven horizons 81.4%, restricted mean at 3 70.3%, one test for all twelve 80.2%; curves that cross, 50 a side: log-rank 5.1%, best of eleven horizons 45.7%, restricted mean at 3 10.4%, one test for all twelve 44.0%; curves that cross, 100 a side: log-rank 5.2%, best of eleven horizons 76.7%, restricted mean at 3 16.4%, one test for all twelve 75.5%; curves that cross, 200 a side: log-rank 5.2%, best of eleven horizons 96.7%, restricted mean at 3 26.5%, one test for all twelve 96.6%.

One test for a curve that might cross

The log-rank statistic and the differences in restricted mean survival at eleven horizons are one Gaussian vector under no effect, and its correlations are closed — from −0.461 at half a year to −0.949 at three, where the log-rank statistic is nearly the restricted mean itself. So one critical value, 2.374, prices the largest of all twelve. At 100 a side it keeps 93.6% of the log-rank test's power when the hazard is halved throughout and 98.5% of the best horizon's when the curves cross, where the log-rank test has none. Running both tests and keeping either rejection rejects 8.67% of trials with no effect.

4 figures · Censoring, part 8

FieldsThreadsSeriesConceptsFigure librarySearch