The number is correct… but it is wrong

The figure may be accurate to the last decimal place, yet still lead us to a wrong conclusion. Numbers do not lose their innocence when they are forged; they may lose it when they are stripped of the context that gave them meaning and are made to answer a question for which they were not collected. I do not use “true” and “correct” in their mathematical senses, but I distinguish between the truth of information and the validity of the inference drawn from it: truth is an attribute of the information, while validity is an attribute of the relationship we construct between that information and the conclusion. The price of a barrel of oil may show a genuine increase, but inferring from it an improvement in the general budget is not universally valid. The impact depends on production volume, the breakeven price, expenditures and commitments, and the measurement period. The number did not lie, but we burdened it with testimony in a case where it had not heard the other witnesses. Here begins the problem of a mind that sees the value but not the system within which it operates. Statistician Francis Anscombe created four datasets with similar means, variances, and correlations, yet their plots revealed four clearly different patterns.[1] The calculations were not wrong; the error was in relying on them alone. This is why the Organisation for Economic Co-operation and Development does not equate statistical quality with accuracy alone, but includes relevance, credibility, timeliness, interpretability, and consistency.[2] An accurate number may be of little value if it is used to answer the wrong question. The rule does not stop at numbers: a genuine photograph may be cropped, a correct statement may be removed from its context, a real clinical presentation is not enough to diagnose a disease, and a legally correct text may not apply to the specific case. Tversky and Kahneman showed that presenting the same results in different frames changes people’s choices.[3] The information did not change; the window through which it was viewed changed, and so did the judgment. Thus, a part may be true, while the picture we construct from it may be false. This is why the Almighty says: “And if they had referred it to the Messenger and to those in authority among them, those who derive it from them would have known it.” Al-Tabari, in his exegesis, explained that referring the matter to the scholars means describing those who derive it as those who investigate it.[4] There is meaning in this: the arrival of information is not equivalent to possessing its significance. Information is a right for everyone, but converting it into a specialized judgment is a responsibility that requires tools; for news arrives all at once, but its meaning can only be extracted after asking about what surrounds it. Therefore, I propose a criterion for the validity of inference from information: the truth of the source, the precision of the meaning, the completeness of the context, the presence of influencing factors, the relevance of the information to the question, and the consistency of the conclusion with what is substantiated. If any of these is lacking, the judgment must be suspended or limited to the bounds of what we know. The most dangerous judgments are not those built on lies, but those erected on a single truth after the rest of the truth has been absent. A true piece of information does not become a correct judgment simply because we found it first.