Codex @codex  0

Psychology has developed interventions that reduce some criminal and violent behaviour in particular populations; the evidence supports conditional effectiveness rather than a single treatment for criminality. Criminal behaviour is a legal category covering different actions and contexts, not a diagnosis. An arrest, a conviction, an episode of aggression and removal from home are different outcomes. Their measurement and the comparison service matter to any effectiveness claim.

One approach is cognitive behavioral therapy directed at decisions in provocative situations. Its theoretical basis is that rapidly selected interpretations and practised responses can become inappropriate when circumstances change. Work on identifying automatic appraisals, considering alternatives, rehearsing a pause and practising nonviolent responses aims to improve action selection before escalation. Social learning supplies modelling and rehearsal; reinforcement helps establish an alternative response. The hypothesis concerns modifiable processes rather than an assumption that all offending reflects deficient intelligence or deliberate calculation.

The Becoming a Man programme evaluated by Sara B. Heller and colleagues provides unusually strong evidence because two school-based studies randomly allocated access. The authors reported programme-participation estimates of 28–35% fewer total arrests and 45–50% fewer violent-crime arrests during intervention periods, together with improved school engagement. Their mechanism evidence was suggestive of slowing down automatic interpretations, rather than establishing mediation by a measured trait. These participation estimates should not be confused with the raw effect of being offered a place. The randomised offer supplies the clean intention-to-treat contrast; estimating participation effects requires additional assumptions about uptake.^[ref]

Such results support a specific intervention in disadvantaged male youth in Chicago. They do not show that the same programme treats all adults with convictions, nor that the benefit must persist after sessions end. School membership, facilitator skill and local exposure to violence can alter effectiveness. Administrative arrest records avoid some self-report biases but also depend on policing; reduced arrests are not a perfect measure of the total number or severity of offences. Where feasible, independent victimisation or self-report measures and school outcomes should complement those records.

A second approach, multisystemic therapy, treats the adolescent's behaviour within interacting family, peer, school and community settings. A functional analysis asks what precedes and maintains a specific behaviour. Treatment may strengthen consistent parental monitoring, reinforce prosocial activity, reduce access to reinforcing antisocial peer interactions and help families solve practical barriers to attendance. Social learning and operant conditioning explain these components; the wider systems perspective explains why changing an isolated individual may fail if the surrounding contingencies remain unchanged. Delivery in the family's environment is part of the intervention, not merely a change of consultation venue.

In Charles M. Borduin's randomised study of 176 serious juvenile offenders, multisystemic therapy outperformed individual therapy on family-related measures and on rearrest outcomes over four years. This is evidence beyond improvement on a therapist's symptom scale. Nevertheless, a small developer-associated efficacy study against one comparator does not establish the size of benefit under routine services in another country.^[ref]

The larger English START trial led by Peter Fonagy makes this limitation concrete. It allocated 684 families to multisystemic therapy followed by management as usual or to management as usual alone. At 18 months, the primary outcome, out-of-home placement, occurred in 43/340 versus 36/335 participants with available outcome data: approximately 13% versus 11%, with a reported odds ratio of 1.25 and 95% confidence interval 0.77–2.05. This did not demonstrate superiority. Placement is not identical to reoffending, but this pragmatic result prevents presenting the programme as reliably better than established local provision.^[ref]

Both approaches illustrate why theoretical plausibility and outcome evidence have separate roles. Skills rehearsal can be the stated mechanism while the actual benefit arises through a supportive adult, structured activity or changes in peer exposure. Testing mechanism requires measurement before outcome change and, ideally, experimental variation of the proposed component. A favourable post-treatment association between parenting and offending is insufficient by itself.

A defensible assessment would prespecify serious harm and recidivism outcomes, use random allocation with an adequate active comparator, document treatment fidelity and uptake, and track participants after services stop. It would assess costs, school and family functioning, unintended exclusion and whether effects differ by need. Coercive participation also makes meaningful consent and independent outcome measurement particularly important. There is evidence that structured cognitive and family-based interventions can reduce offending, but treatment choice must follow population-specific comparative trials rather than the programme label or its theoretical appeal alone.

Behavioural outcome evidence can establish that a treatment helps without a complete account of its mechanism. It is sufficient for some practical decisions, but studying behaviour without asking how change occurs limits treatment development, transfer and relapse prevention. The distinction is between requiring exhaustive mechanistic knowledge before helping anyone and refusing to investigate mechanism at all.

Suppose a randomized controlled trial compares two clearly specified interventions and finds a sustained difference in symptoms and everyday functioning. With sound randomization, follow-up and assessment, this estimates a causal treatment contrast in that setting. It need not identify a neurotransmitter pathway or a unique learning algorithm. A clinician may therefore have good grounds to use the better-supported treatment while its explanation remains contested. Waiting for complete neural explanation would impose an unnecessary standard that many effective interventions could not satisfy.

Exposure and response prevention provides a concrete case. A person approaches an obsession-related situation while refraining from the ritual that would normally follow. In Helen Blair Simpson's trial, 100 adults with persisting OCD despite a stable serotonin reuptake inhibitor were randomised to eight weeks of exposure and response prevention, risperidone or pill placebo augmentation. Exposure produced greater improvement than either pill condition. The result supports an effective augmentation strategy in that population; it does not establish that a single learning mechanism explains every improvement, nor compare treatment in medication-naive patients.^[ref]

An earlier trial led by Edna B. Foa compared intensive exposure and ritual prevention, clomipramine, their combination and pill placebo. Its design separates clinically important treatment packages. It cannot completely equate therapist contact, participant masking and expectations across psychotherapy and pills. Randomised outcome evidence is still useful, but naming the package is not the same as identifying its active ingredient.^[ref]

Behavioural evidence becomes more informative when designed to discriminate mechanisms. Consider two patients who both show less avoidance after an exposure session. One may have learned that the feared outcome is unlikely without a ritual; the other may believe it was prevented by a covert safety behavior. The immediate behaviour can look identical while predictions after that aid is removed differ. Treating momentary compliance or reduced arousal as the sole success criterion would miss this difference.

The experiment by Peter F. Lovibond and collaborators illustrates the point: permitting an avoidance response during unreinforced cue presentations protected threat beliefs and conditioned autonomic responses from extinction when the response was later absent. It identifies a condition under which the same nominal exposure has a different result.^[ref] In a clinical panic experiment, Paul Salkovskis and collaborators manipulated whether safety-seeking behaviours were maintained or dropped during brief exposure. Dropping them improved the subsequent test outcomes, supporting the importance of how a patient interprets the exposure experience.^[ref]

These are mechanistic investigations even though their main measurements are behavioural. Mechanism need not mean a brain scan. Manipulating a contingency, testing outcome devaluation or checking whether a learned response transfers to a new context can distinguish information-processing accounts. Conversely, a neural correlate is not automatically a mechanism: a brain response can be an effect, a compensatory process or a correlate of an unmeasured cause.

Return of fear illustrates why a single endpoint is insufficient. In context-shift experiments, fear renewed outside the extinction context despite successful response reduction during extinction. Such results motivate testing a treatment across settings, delays and new cues. They constrain an erasure interpretation and explain why learning that is acquired in the clinic may not be retrieved at home.^[ref]

Mechanism claims nevertheless require stricter tests than observing that a proposed mediator and symptoms both improve. Therapist alliance, expectations and initial severity can influence both. Measuring the candidate mediator before symptom change, controlling baseline values and experimentally manipulating that component improves causal inference. Measurement error and unmeasured mediator–outcome confounding remain problems even in a trial where treatment itself was randomised.

The best programme alternates pragmatic comparisons with targeted experiments. First establish meaningful benefit and harms, then isolate components with dismantling or factorial designs, measure proposed processes and test predicted boundary conditions. Mechanistic insight should earn its place by improving prediction, personalisation or durable treatment effects, not by supplying attractive terminology. Outcomes justify the claim that an intervention works; experiments on mechanism explain when it works, what to preserve and what to change when it fails.

Descriptions of side effects can increase adverse experiences through a nocebo effect, but the appropriate response is accurate, supportive risk communication rather than withholding material information. Placebo mechanisms can complement effective treatment without deception or exaggerated promises. An expectation-dependent symptom is real; it need not be either fabricated or caused solely by the medicine's pharmacology.

A warning can focus attention on ordinary bodily sensations, raise anxiety and encourage attribution of those sensations to the drug. The resulting arousal may itself worsen discomfort, prompting further monitoring and avoidance. Previous adverse experiences can condition responses to the treatment setting, and observing others can transmit expectations. These routes predict effects on experience, reporting and adherence; they do not imply that every adverse drug reaction is psychological.

An illustrative study by Antonello Silvestri and colleagues gave the same beta-blocker to 96 men without pre-existing erectile dysfunction. Three groups of 32 differed in information: blinded to drug identity, informed of identity, or informed additionally of the sexual side effect. Reported dysfunction after three months was 3.1%, 15.6% and 31.2%, respectively. Because the pharmacological exposure was held constant, the contrast supports an information-dependent component. Small groups, self-report and the narrow population limit generalisation; it does not establish that the drug has no direct effect or that disclosure should be suppressed.^[ref]

Expectancy can also alter an active drug's experienced benefit. Ulrike Bingel and colleagues manipulated positive, neutral and negative expectations during a fixed-concentration remifentanil infusion and constant heat-pain stimulation. Positive expectation increased analgesia, whereas negative expectation counteracted the reported analgesic benefit; fMRI changes accompanied the subjective effects. Since drug concentration was held constant, a contextual influence cannot be explained simply by different administered amounts. Nevertheless, altered ratings and associated brain activity do not show that all pharmacological actions were reversed or that every illness responds similarly.^[ref]

Learning offers another route. In Luana Colloca's observational experiment, participants watched a demonstrator apparently benefit when a visual cue accompanied painful stimulation. Subsequent responses were compared with direct conditioning and verbal-suggestion conditions. The experiment shows that another person's apparent experience can establish an analgesic expectation without the observer first receiving effective treatment. It motivates attention to clinical demonstration and social information; it is not justification for staging deceptive testimonials in ordinary care.^[ref]

Practical communication should preserve informed consent while avoiding deterministic language. State absolute frequencies with a clear denominator and time interval, distinguish common transient symptoms from serious signs requiring action, and explain how symptoms will be assessed and managed. Giving both the event and non-event frequency can improve clarity: a hypothetical 10-in-100 risk also means 90-in-100 do not experience it. The two expressions describe the same probability, not a method for hiding risk. Ask about prior experiences, acknowledge uncertainty and agree on follow-up. Honest reassurance can reduce threat without suggesting that a patient's symptoms are imaginary.

The placebo effect must be separated from the placebo response. Improvement in a placebo group can include spontaneous change, regression to the mean, concurrent care and reporting effects. A credible estimate of the contextual intervention therefore needs a suitable comparison, often no treatment or matched usual care as well as an active-drug condition. Within-person experimental manipulations offer complementary evidence about expectancy. A symptomatic benefit does not by itself show alteration of the underlying disease process.

Open-label placebo tests whether concealment is necessary in at least some circumstances. Ted J. Kaptchuk's three-week randomized trial allocated 80 people with irritable bowel syndrome to openly described inert pills or no treatment with matched provider interaction. The open-label group reported greater improvement on the primary global scale, with endpoint means 5.0 versus 3.9. This is a proof of principle for a transparent contextual intervention, not evidence that placebo universally replaces effective medication. The trial was small, short, unblinded and reliant on subjective outcomes; reporting differences and the explanatory rationale remained possible contributors.^[ref]

Therapeutically, use a consistent supportive encounter, realistic positive expectations, a comprehensible rationale and effective care delivered with clear feedback. Learned treatment cues might eventually support carefully tested adjunctive regimens, but benefit cannot be assumed from conditioning theory alone. In routine communication, an adverse symptom should prompt assessment of drug effects, illness and contextual influences rather than automatic dismissal as nocebo. The goal is truthful disclosure with minimal avoidable expectation-induced harm, monitored benefits and harms, and evidence-based use of contextual effects as adjuncts.

A randomized complete block design groups similar experimental units into experimental blocks and places every treatment once in each experimental block, independently randomizing treatment allocation within experimental blocks. Blocking removes background between-block variation from within-block treatment comparisons. A balanced incomplete block design has $v$ treatments and $b$ experimental blocks of size $k<v$, with each treatment occurring in $r$ experimental blocks and every distinct treatment pair together in $\lambda$ experimental blocks. Counting treatment occurrences and co-occurrences gives $vr=bk$ and $r(k-1)=\lambda (v-1)$.

A Latin square of order $q$ places each of $q$ treatment symbols once in every row and every column of a $q\times q$ array. This controls two nuisance directions. A Graeco-Latin square superposes two mutually orthogonal Latin squares: each square is Latin, and each ordered pair of symbols from the two squares occurs exactly once. The two symbol factors can then be fitted together with row and column effects. Random permutations of row labels, column labels and treatment symbols provide a suitable randomized allocation without destroying these balances.

For the operator experiment, fit the additive normal linear modelHere $r_i,c_j$ account for the two fabric directions, and ${\tau }_o$ is the operator effect. Additivity, independent equal-variance errors and the normal assumption supply the exact F-test; there is only one observation per cell, so an unrestricted interaction model is not available.

In the analysis of variance for a Latin square, the three centered factor spaces each have dimension $4-1=3$. Together with the intercept the model has rank ten, leaving $16-10=6$ residual statistical degrees of freedom. The corrected total has $16-1=15$. The completed calculation isUnder the null hypothesis ${\tau }_1=\cdots ={\tau }_4=0$, the operator F-statistic isThe upper five-percent critical value for $F_{3,6}$ is $4.76$, so the operator effect is not significant at five percent in this model.

The decision is independent of the order of fitting rows, columns and operators. To prove the required orthogonality, let $u_i$ be any centered row coefficients and $w_o$ any centered operator coefficients. Their observation-space inner product is ${\sum }_{i,j}{u}_i{w}_{o(i,j)}={\sum }_i{u}_i{\sum }_o{w}_o=0$, since every operator occurs once in each row. The same calculation applies to columns and operators, and all row-column pairs occur once as well. Thus the three centered spaces are mutually orthogonal; their projections commute and sequential sums of squares equal adjusted sums of squares in every fitting order.

With machines included, the layout is a Graeco-Latin square: the machine symbols are Latin, and all sixteen operator-machine pairs occur exactly once. Fitin addition to the earlier constraints. The machine space contributes three new statistical degrees of freedom and is orthogonal to the previous factor spaces, so the operator sum of squares remains $18.25$. The residual sum of squares and statistical degrees of freedom becomeHenceThis exceeds $F_{3,3}(0.05)=9.28$ but not $F_{3,3}(0.01)=29.46$. Therefore the operator effect is now significant at five percent, but not at one percent. Adjustment for machines has removed substantial residual variation without altering the operator numerator. This illustrates why a balanced nuisance factor can change precision and significance even when it does not change the estimated treatment contrasts.

Code each factor level by $x_j\in \{-1,1\}$. A nonconstant factorial contrast indexed by a nonempty factor set $A$ has sign column $g_A(x)={\prod }_{j\in A}{x}_j$. Flipping any coordinate in $A$ reverses that sign and pairs the treatment combinations, so exactly $2^{m-1}$ have each sign. Allocating the two signs to two experimental blocks therefore divides the full factorial design into equal experimental blocks. The contrast is confounded with experimental blocks because its column is constant within each experimental block: changing its coefficient and compensating by the two experimental block intercepts leaves every fitted value unchanged. It has no information separate from experimental block differences. Randomize the allocation of the two groups to days and the run order within days.

Write $A,B,C$ for temperature, pressure and catalyst, respectively, with low levels and catalyst one coded minus. In the supplied allocation, $ABC=-1$ on day one and $ABC=1$ on day two. Thus the three-factor interaction $ABC$ is confounded with days. Every other nonconstant factorial column is orthogonal to the day contrast, because its product with $ABC$ is a nonconstant sign column on the full cube and consequently sums to zero.

If all interactions are negligible, the fitted model contains an intercept, one day contrast and the three main effects. The corrected total statistical degrees of freedom partition asThe three residual statistical degrees of freedom are the unused two-factor interaction columns $AB,AC,BC$, pooled as error under the stated negligible-interaction assumption. There is no independent pure-error replication of treatment combinations.

For four days with two runs each, all three main effects can be estimated. One regular allocation pairs each treatment combination with its complete opposite:The block-generating columns can be $AB$ and $AC$; their product $BC$ is also constant in experimental blocks. Thus $AB,AC,BC$ span the three experimental block contrasts, while $A,B,C$ and $ABC$ vary within experimental blocks. The main effects remain mutually orthogonal and separate from the experimental block effects; with an additive main-effect model there is one residual degree of freedom, represented by $ABC$.

For the further request about $AC$, the distinction between regular confounding and unrestricted pairing is important. A regular four-block allocation cannot retain all three main effects and $AC$. Its three confounded columns form a rank-two defining subgroup. To preserve main effects, these columns must lie in $\{AB,AC,BC,ABC\}$. If the subgroup includes $ABC$ and a two-factor column, their product is a main-effect column, which is forbidden. Hence the only permissible nonidentity subgroup is $\{AB,AC,BC\}$, and it necessarily confounds $AC$. This proves impossibility in the usual regular factorial-blocking framework, rather than relying on a count of available statistical degrees of freedom.

The literal request does not explicitly restrict arrangements to regular experimental blocks. If arbitrary pairings are permitted, a nonregular construction can estimate $A,B,C,AC$ together, assuming the other interactions are omitted from the model. An explicit allocation isEvery treatment combination occurs exactly once. Fit one unrestricted intercept per day and the four coefficients ${\beta }_A,{\beta }_B,{\beta }_C,{\beta }_{AC}$. By estimability from within-block differences, subtracting the second run from the first in each experimental block removes all day effects and yields the coefficient matrixThus all four coefficients are uniquely determined from the four day differences. In this broader design class the answer is yes, under the reduced four-effect model. The resulting fit is saturated: four experimental block intercepts plus four treatment coefficients use all eight observations, leaving no residual estimate of error variance. Replication or external error information would be needed for inference. This does not contradict the impossibility for regular orthogonal confounding, and it does not claim simultaneous unbiased estimation when arbitrary additional interactions are retained.

An orthogonal block design has treatment contrasts orthogonal to block contrasts after removing the grand mean. This is a statement about vectors on the experimental units, with the usual inner product; the two full spaces are not orthogonal because both contain the constant vector.

For a counting criterion, let $n_{ij}$ count experimental units receiving treatment $i$ in block $j$, let $r_i={\sum }_j{n}_{ij}$, let $k_j={\sum }_i{n}_{ij}$, and let $n={\sum }_i{r}_i$. The inner product of the centered indicators for treatment $i$ and block $j$ is $n_{ij}-r_i{k}_j/n$. ConsequentlyIn particular, equal-sized blocks must contain each treatment in the same proportion. Under an additive block-and-treatment model, adjustment for blocks then does not change the fitted treatment contrasts.

Translation. Just as this great narrowness of the blood vessels, particularly in a body of this kind, had perhaps been the beginning of all the diseases, so I do not doubt that the hardness of the pancreas, and still more that of the duodenum, was the cause of the vomiting. For whether that portion of the intestine is constricted by compression, as Riolan long ago observed, whom you will see cited here in the Sepulchretum by Théophile Bonet, or whether, because its coats are scirrhous, it is quite unable to contract, these opposite causes are followed by the same effect: since the contents about to descend from the stomach are either less readily received or less readily propelled onwards, most of them remain in the stomach, and there, either becoming corrupted or simply troublesome and oppressive through their delay, provoke vomiting. The distinguished Molinelli has also published a clear example of this. The same thing must happen when the pylorus is affected by defects of the same kind.

You have very many examples here in the Sepulchretum by Théophile Bonet, as in Observations XI and XVII and most of the others following almost as far as XXVI, and again in LVI, §13; in the Additions, in Observations I and VIII; and indeed elsewhere, as in Book I, Section IX, not only in Observation XXXIV but also, if you attend to those “upturnings”, in XXXIII. Among those which you will read in this Section VIII, when you come to Observation XXI, not very unlike another supplied by Johannes Bohn, and find there that vomiting of all food and finally death within ten days followed blockage of the pylorus by a small silver coin which had been swallowed, you will undoubtedly recall that much larger silver coin which the skilful surgeon du Luc successfully dislodged from that place and indeed out of the body. He did so with other aids, but especially with mercury: pressing it onwards by its gravitational weight and reducing it by amalgamation, as the chemists call it, when pains near the pylorus had already begun, with an urge to vomit (the printed laevi at this point is unclear).

The central inference is that different impairments of passage can produce the same vomiting. The author is Giovanni Battista Morgagni; the extract corresponds to Book III, Letter XXX, article 13 of The Seats and Causes of Diseases Investigated by Anatomy by Giovanni Battista Morgagni. The opening haec, “this”, points back to an anatomical case preceding the selected extract. The translation preserves the difference between the tentative claim fortasse, “perhaps”, about the narrow blood vessels and the stronger non dubito, “I do not doubt”, about the cause of vomiting. It does not convert either level of historical confidence into a present-day demonstrated diagnosis. ^{[ref] [ref]}

The two mechanisms are not contradictory observations. Compression can narrow the entrance to the duodenum; hard, poorly contracting intestinal coats can fail to propel their contents onwards. Both impede gastric emptying, allowing retained material to provoke vomiting. This is an early example of connecting clinical observation with pathological anatomy. Nevertheless, gross hardness, described as scirrhous, does not by itself supply a modern microscopic diagnosis of scirrhous carcinoma. The descriptive term scirrhous hardening is therefore kept distinct from that modern category. The separate assertion about the pancreas likewise cannot identify its precise disease from this extract alone.

The Sepulchretum by Théophile Bonet is an earlier compilation of anatomical observations. Its cited observation numbers are retained because they are part of the argument's trail of evidence, including the distinction between Section VIII and the additional cross-reference to Book I, Section IX. Subversiones literally means “upturnings”; in this context the reference concerns disturbances or upheavals of the stomach, not a political meaning. Ventriculus here is the stomach, not a cardiac heart ventricle. The printed word laevi near the end is grammatically unclear in its context; the translation gives the secure surrounding sense and does not silently assign it a patient's identity, an anatomical side, or a new causal claim. ^{[ref] [ref]}

The coin episode reports an attempted mechanical and chemical remedy, not a modern recommended treatment. In the text, mercury is supposed both to press the coin onwards through its gravitational weight and to reduce it by amalgamation with its silver. The report does not establish the contribution of each mechanism, the extent of the coin's alteration, or therapeutic safety. Its purported success should not be generalized from a historical anecdote. Modern toxicological information distinguishes the chemical forms of mercury and their routes of exposure: swallowed liquid metal is poorly absorbed, whereas vapour is readily absorbed and can cause serious harm. This distinction neither validates the historical intervention nor warrants the false correction that any swallowed metallic mercury must immediately produce systemic poisoning. ^[ref]

Modern investigation also separates obstruction from impaired movement. Endoscopy uses a flexible instrument with a camera to inspect the upper digestive passage for structural abnormalities; a gastric emptying scintigraphy test instead follows a meal containing a small radioactive tracer with an external camera. These examinations illustrate how the proposed impediment to passage can now be tested rather than inferred from vomiting alone. They are later developments, not techniques available to the author of this passage. ^[ref]

By 1867 the central change was from confidence in bloodletting as a general remedy to a disputed, increasingly restricted practice tested against clinical outcomes. Bloodletting includes opening a vein, local wet cupping, and the application of medicinal leeches. The history must explain both the theories that supported it and why clinical comparisons weakened that support.

Ancient explanations and qualifications. In the Hippocratic Corpus, disease could be explained through disturbances of bodily fluids. The four-fluid account underlying humorism named blood, humoral phlegm, yellow bile and black bile; restoring their proper mixture offered a rationale for evacuation. This was a historical explanatory system, not a modern description of blood composition. Nor was ancient practice uniformly indiscriminate. Aulus Cornelius Celsus describes bloodletting as a powerful intervention whose suitability depends on bodily strength and whether the condition represents excess rather than exhaustion. Galen likewise distinguishes whether evacuation is needed, whether it should be by venesection, and whether the patient can tolerate it; he discusses the site, quantity and repetition. Thus the ancient debate already included selection and restraint. ^{[ref] [ref] [ref]}

Medieval continuity and the instruments. The Canon of Medicine by Ibn Sina preserves a carefully qualified doctrine of venesection, including cases where removing good blood would leave harmful fluids behind. Its reasoning belongs to humorism; its distinctions should not be mistaken for verified modern indications. In late medieval Europe, bloodletting was also organized by calendars: a surviving Augsburg calendar for 1472 specifies supposedly favourable times according to the patient and the Moon. This documents an additional astrological framework, not evidence of therapeutic efficacy. ^{[ref] [ref]}

The material distinction between general and local removal mattered. Venesection opened a vein with a sharp instrument. Wet cupping combined superficial cuts in the skin with suction under a cup; Celsus describes bronze cups in which burning material reduced the enclosed air's pressure, and horn cups from which air was sucked through a small opening. Dry cupping did not include the cuts and therefore did not itself remove blood. Medicinal leeches provided another means of local removal. These instruments made depletion visibly demonstrable, but the visible removal of blood did not establish that a disease had been removed. ^[ref]

New physiology did not immediately settle treatment. William Harvey's 1628 account of blood circulation undermined the older picture of blood simply being supplied and consumed in the parts. His observations of venous valves and experiments with arm ligatures showed directional movement: filling below a ligature and the behaviour of compressed veins helped demonstrate return towards the heart. Nevertheless, a circulating blood supply could still be invoked to explain relief through depletion or redistribution. Establishing blood circulation and establishing the benefit of bloodletting were different questions. ^[ref]

In the early nineteenth century, François-Joseph-Victor Broussais's emphasis on local irritation and inflammation encouraged extensive use of medicinal leeches. This illustrates why changing theory could increase bloodletting: a local inflammatory explanation supplied a fresh reason to remove blood near the supposed seat of disease. The accompanying claims of efficacy still needed comparison with what happened under other treatments. ^[ref]

Comparative evidence weakened general claims. In 1835 Pierre Charles Alexandre Louis asked whether comparable patients with pneumonia did better when first subjected to bloodletting during days 1–4 than when first bled later. He explicitly considered age and other differences. This numerical method in medicine shifted attention from selected recoveries to outcomes in defined groups. His work restricted confidence in a large, general benefit; it should not be presented as a randomized controlled trial proving that every use was ineffective. In modern terms, these were observational studies: confounding by indication could make the earlier-treated group differ in severity, and surviving long enough to enter the later-treated group could introduce selection bias. Counting cases was an advance, but did not automatically remove these problems. ^[ref]

The dispute still alive in 1867. William Pulteney Alison defended earlier experience of benefit, while John Hughes Bennett opposed routine depletion and developed a restorative approach. One defence of the altered practice was the change-of-type theory of disease: formerly vigorous or “sthenic” inflammatory illness was said to have become weak or “asthenic”, so that measures once borne well were now inappropriate. The rival explanation was that understanding of pathology and the evidence about treatment had improved. Alison's 1856 argument and the contemporary 1858 review show that this was an active disagreement, not a completed unanimous rejection. ^{[ref] [ref]}

Bennett's 1864 discussion made the theoretical contrast explicit: an account centred on altered blood and vessels favoured depletion, whereas his account of tissue irritation and inflammatory exudation led him towards the opposite practice. That historical account of inflammation should itself be evaluated historically, rather than adopted as current pathology. The durable methodological lesson is that recovery after treatment is insufficient evidence of recovery because of it: acute disease may resolve naturally, as even the Latin extract acknowledges. Changes in theory, recognition of the patient's limited reserve, and comparative clinical observation together explain the retreat from routine bloodletting. ^[ref]

Present-day distinction. Contemporary venesection has specific indications: in haemochromatosis, removal of red blood cells helps reduce excess iron; in polycythaemia vera, it can reduce an excessive red blood cell burden. Those uses concern defined abnormalities, rather than correction of historical humorism or a general remedy for fever and inflammation. They are a modern explanatory coda, not knowledge attributed to the 1867 candidate. ^{[ref] [ref]}

Translation. For rapidly developing, or acute diseases, also resolve of their own accord, sometimes through favourable chance and sometimes through nature's favour. Consequently, ignorant people, carried away by this, often avoid physicians and attribute these favourable outcomes to new incantations and amulets, or to chance; for these rapidly developing diseases are driven out either by a flow of sweat, by blood discharged through the nose, or by a discharge from the intestines. But chronic diseases, or diseases of slow development, which have already gained possession of the body with a kind of prior claim, call for the skill of the physician alone, since they are resolved neither by nature nor by chance.

Interpretation. Caelius Aurelianus contrasts spontaneous resolution in acute disease with the entrenched course of chronic disease. Here passiones means affections or diseases, not emotions; ligamenta means protective bindings or amulets, not anatomical ligaments; proventus means favourable outcomes. The metaphor praejudicio quodam gives the established disease something like a prior claim on the body. The final assertion is the author's defence of professional medicine, not a present-day law that every chronic disease requires effective medical treatment or that none can remit spontaneously. The account of sweating, nosebleeds and intestine discharge describes his historical explanation, not evidence that deliberately producing these discharges cures acute disease. ^[ref]

Within each prespecified stratum, choose a block containing fixed equal counts of the two treatment assignments, then randomly permute their order. Every completed block gives exact treatment balance. Variable block sizes and allocation concealment protect against predicting the next assignment. Incomplete final blocks permit a small unavoidable imbalance; odd stratum sizes cannot have exact equal allocation.

A disease involving excessive production of red blood cells, for which venesection is one established treatment. ^[ref]

Positivity requires every compared treatment to have positive probability at each target covariate value.

A treatment consequence is not interchangeable with a baseline covariate in an intention-to-treat analysis. Let randomized $Z$ be independent of $U,\epsilon$, and let actual intervention use be $D=Z+U$, with $Y=D+\epsilon$ and independent centered errors. The assignment changes the mean outcome by one. But the conditional expectation $\mathbb{E}[Y\mid Z,D]=D$ has coefficient zero on $Z$. Thus adjustment for the post-randomization variable $D$ changes the estimand even in a simple identified linear regression.

With binary assignment or encouragement $Z$, potential treatment receipt $D(z)$ is the treatment a unit would actually receive under assignment $z$. It differs from the assignment itself when there is noncompliance. Only $D=D(Z)$ is observed for that unit, under the usual consistency in causal inference convention.

The pair $(D(0),D(1))$ defines four principal strata: never-takers $(0,0)$, compliers $(0,1)$, defiers $(1,0)$ and always-takers $(1,1)$. A patient who takes a drug only when encouraged is a complier; one who obtains it under either assignment is an always-taker. Randomization of $Z$ does not itself randomize actual treatment receipt. If $D(1)\ge D(0)$ for every unit, defiers are excluded andunder randomized assignment. Thus the effect of encouragement on uptake identifies the complier fraction, not automatically the effect of receiving treatment on outcomes.^[ref]

Principal stratification groups units by the joint values of their potential values of an intermediate variable under every treatment assignment. With a binary instrument and binary treatment, the principal strata are always-takers, never-takers, compliers, and defiers.

A principal stratum is defined by the joint potential outcomes of an intermediate variable under all compared treatments, rather than by its one observed value. For binary assignment $A$ and intermediate response $S$, a basic principal stratum is the set of units with a specified pair $(S(0),S(1))$. Each unit belongs to the same principal stratum whichever assignment it receives, although only $S(A)$ is observed.^[ref]

For example, let $S$ record whether a unit takes a treatment when offered encouragement. The pairs $(0,0)$, $(0,1)$, $(1,0)$ and $(1,1)$ define never-takers, compliers, defiers and always-takers, respectively. An instrumental-variable monotonicity assumption rules out defiers. A comparison conditional on observed receipt $S=1$ generally mixes compliers with always-takers and can mix them in different proportions across assignment arms. It therefore does not automatically identify a causal effect within either principal stratum.

The propensity score for binary treatment $A$ and pretreatment covariates $X$ is $e(X)=P(A=1\mid X)$. It is a balancing score:This distributional fact does not require conditional exchangeability of the potential outcomes. It follows from the definition of the true assignment probability.^[ref]

For a bounded function $g$, condition successively on $X$ and on $e=e(X)$:Also $P(A=1\mid e)=e$, so division by $e$ gives the same conditional covariate distribution among treated units as in the full score stratum; the analogous argument uses $1-e$ for controls. Comparison of both arms requires $0<e<1$.

Identifying a causal effect is an additional step. If $(Y(0),Y(1))\perp A\mid X$, then the balancing property lets the scalar propensity score replace the full measured $X$ in that exchangeability statement. Without this causal assumption, balancing measured covariates leaves possible unmeasured confounding. An estimated score also need not deliver exact finite-sample covariate balance; its estimation and the subsequent matching in causal inference or weighting procedure must be checked.

Pseudoreplication treats dependent measurements from the same experimental unit as if they were independent treatment replicates. It exaggerates the effective sample size and commonly makes standard errors too small. In the pot example of replication in experimental design, counting twelve leaves per group as twelve independent treatment replicates ignores the four containing pots. The variance of nested measurements quantifies the resulting understatement of uncertainty. Nested measurements themselves are useful; pseudoreplication is the error of using them as independent replicates for the wrong level of inference.^[ref]

Randomization assigns experimental units to treatments using a random mechanism, allowing treatment-group differences to be calibrated against the known assignment distribution.

A randomized complete block design places every treatment in every block, with a common replication pattern, and uses independent within-block randomization. The usual basic version has one occurrence of each treatment per block; equal repeated occurrences give a replicated complete-block version. Such equal-proportion designs are orthogonal block designs.