FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W279996790> ?p ?o ?g. }

• W279996790 abstract "COMMENT ON: OLDFORD, R.W., A PHYSICAL DEVICE FOR DEMONSTRATING CONFOUNDING, BLOCKING, AND THE ROLE OF RANDOMIZATION IN UNCOVERING A CAUSAL RELATIONSHIP, THE AMERICAN STATISTICIAN , 49:210-216, 1995 by Judea Pearl Oldford presents an ingenious teaching device which vividly displays the di culties as- sociated with attributing causal interpretation to regression equations, and thus enhances students' appreciation for several aspects of causal relations and experimental design. This letter proposes a more drastic reform of current teaching of causality, one based on replacing the vocabulary of regression analysis with extensive use of causal diagrams. Oldford advocates, and I agree, that in the classroom the teacher should de ne the parameter of the causal equation Y = + X + Z as the change expected in y when x is changed by a unit amount and everything else i.e., z is held xed. However, the routine translation of is changed and is held xed into the language of conditional expectations as in ...they students need to x z and examine the relationship between x and y conditioned on z may lead to disaster. To illustrate: Consider an extreme case where z and x are functionally related, say through z = ax + b , e.g., the amount of blood transfused z was determined as a linear function of the amount x of anesthetic used. This would render y and x conditionally independent given z and would bias regressional calculations of to zero or, at best, leave undetermined regardless of the actual value of in Eq. 1. Such bias is not limited to functional relationships between x and z ; in the general case, where a disturbance term uncorrelated with x is introduced into Eq. 1, the bias surfaces when z is correlated with both x and . Thus, teachers should discourage the routine translation of causal notions e.g., x z into regressional notions e.g., conditioned on z . The confusion between regression and causal analysis has long been a focus of discon- tention between statisticians, on the one hand, and econometricians and social scientists, on the other Freedman, 1987 and Wermuth, 1992. The source of this controversy is indeed the expression holding z xed, which economists and social scientists interpret as x- ing z by external intervention, and statisticians interpret as conditioning on z , namely, considering only samples in which z attains a certain value. While the distinction between xing and conditioning cannot be formulated in the standard language of statistical analysis, it is vividly displayed in the language of causal diagrams Spirtes etal., 1993; Pearl, 1993, a language that clari es and explicates the precise conditions under which the two interpretations are interchangeable Pearl, 1995. In The American Statistician , Vol. 50, No. 4, 387 388, November 1996." @default.
• W279996790 created "2016-06-24" @default.
• W279996790 creator A5022404172 @default.
• W279996790 date "2011-10-25" @default.
• W279996790 modified "2023-09-24" @default.
• W279996790 title "Comment on: Oldford, R.W., A Physical Device for Demonstrating Confounding Blocking, and the Role of Randomization in Uncovering a Causal Relationship." @default.
• W279996790 hasPublicationYear "2011" @default.
• W279996790 type Work @default.
• W279996790 sameAs 279996790 @default.
• W279996790 citedByCount "0" @default.
• W279996790 crossrefType "journal-article" @default.
• W279996790 hasAuthorship W279996790A5022404172 @default.
• W279996790 hasConcept C104317684 @default.
• W279996790 hasConcept C105795698 @default.
• W279996790 hasConcept C11671645 @default.
• W279996790 hasConcept C121332964 @default.
• W279996790 hasConcept C126322002 @default.
• W279996790 hasConcept C135763542 @default.
• W279996790 hasConcept C138885662 @default.
• W279996790 hasConcept C14036430 @default.
• W279996790 hasConcept C144745244 @default.
• W279996790 hasConcept C149782125 @default.
• W279996790 hasConcept C15744967 @default.
• W279996790 hasConcept C158600405 @default.
• W279996790 hasConcept C168563851 @default.
• W279996790 hasConcept C185592680 @default.
• W279996790 hasConcept C204243189 @default.
• W279996790 hasConcept C2779677306 @default.
• W279996790 hasConcept C2779901538 @default.
• W279996790 hasConcept C2993967602 @default.
• W279996790 hasConcept C33923547 @default.
• W279996790 hasConcept C41895202 @default.
• W279996790 hasConcept C527412718 @default.
• W279996790 hasConcept C55493867 @default.
• W279996790 hasConcept C62520636 @default.
• W279996790 hasConcept C64357122 @default.
• W279996790 hasConcept C71924100 @default.
• W279996790 hasConcept C77350462 @default.
• W279996790 hasConcept C78458016 @default.
• W279996790 hasConcept C83546350 @default.
• W279996790 hasConcept C86803240 @default.
• W279996790 hasConceptScore W279996790C104317684 @default.
• W279996790 hasConceptScore W279996790C105795698 @default.
• W279996790 hasConceptScore W279996790C11671645 @default.
• W279996790 hasConceptScore W279996790C121332964 @default.
• W279996790 hasConceptScore W279996790C126322002 @default.
• W279996790 hasConceptScore W279996790C135763542 @default.
• W279996790 hasConceptScore W279996790C138885662 @default.
• W279996790 hasConceptScore W279996790C14036430 @default.
• W279996790 hasConceptScore W279996790C144745244 @default.
• W279996790 hasConceptScore W279996790C149782125 @default.
• W279996790 hasConceptScore W279996790C15744967 @default.
• W279996790 hasConceptScore W279996790C158600405 @default.
• W279996790 hasConceptScore W279996790C168563851 @default.
• W279996790 hasConceptScore W279996790C185592680 @default.
• W279996790 hasConceptScore W279996790C204243189 @default.
• W279996790 hasConceptScore W279996790C2779677306 @default.
• W279996790 hasConceptScore W279996790C2779901538 @default.
• W279996790 hasConceptScore W279996790C2993967602 @default.
• W279996790 hasConceptScore W279996790C33923547 @default.
• W279996790 hasConceptScore W279996790C41895202 @default.
• W279996790 hasConceptScore W279996790C527412718 @default.
• W279996790 hasConceptScore W279996790C55493867 @default.
• W279996790 hasConceptScore W279996790C62520636 @default.
• W279996790 hasConceptScore W279996790C64357122 @default.
• W279996790 hasConceptScore W279996790C71924100 @default.
• W279996790 hasConceptScore W279996790C77350462 @default.
• W279996790 hasConceptScore W279996790C78458016 @default.
• W279996790 hasConceptScore W279996790C83546350 @default.
• W279996790 hasConceptScore W279996790C86803240 @default.
• W279996790 hasLocation W2799967901 @default.
• W279996790 hasOpenAccess W279996790 @default.
• W279996790 hasPrimaryLocation W2799967901 @default.
• W279996790 hasRelatedWork W1530127219 @default.
• W279996790 hasRelatedWork W1588913678 @default.
• W279996790 hasRelatedWork W1791321265 @default.
• W279996790 hasRelatedWork W1999804543 @default.
• W279996790 hasRelatedWork W2005626088 @default.
• W279996790 hasRelatedWork W2009780301 @default.
• W279996790 hasRelatedWork W2027501736 @default.
• W279996790 hasRelatedWork W2070806897 @default.
• W279996790 hasRelatedWork W2073234747 @default.
• W279996790 hasRelatedWork W2105121211 @default.
• W279996790 hasRelatedWork W2105787136 @default.
• W279996790 hasRelatedWork W2146268665 @default.
• W279996790 hasRelatedWork W2156480747 @default.
• W279996790 hasRelatedWork W2335097232 @default.
• W279996790 hasRelatedWork W2472534060 @default.
• W279996790 hasRelatedWork W2767381899 @default.
• W279996790 hasRelatedWork W2797296389 @default.
• W279996790 hasRelatedWork W3021304644 @default.
• W279996790 hasRelatedWork W132873813 @default.
• W279996790 hasRelatedWork W1638204891 @default.
• W279996790 isParatext "false" @default.
• W279996790 isRetracted "false" @default.
• W279996790 magId "279996790" @default.
• W279996790 workType "article" @default.