Robert FolgerA. B. Freeman School of Business, Tulane UniversityPresents a justification for following statements and discusses their implications: It isduplicitous (misleading) to use significance tests for making binary (either/or) decisions regardingthe validity of a theory; binary choice between calling results significant or not significant shouldnot govern confidence placed i n a theory, because such confidence cannot be gained in either/or fashion characterizing deductive certainty. The implications include grounds for describing waysthat effect size estimates become useful in makingjudgments about value of theories.Charles Sanders Peirce, who some say was America's greatestphilosopher of science, wrote that science can only rest on ex-perience; he then added that experience can never result inabsolute certainty, exactitude, or universality (Buchler, 1955,p. 47). Hypothesis-testing procedures make explicit science'slack of certainty. In particular, inferential statistics provide esti-mates of whether a finding reflects merely random variation, anoutcome owing to chance rather than to causal mechanismshypothesized by theories. I contend that such procedures canbe used in a manner that would deserve label of duplicity(i.e., an attempt to dupe unsuspecting) it were actuallyintentional. Inferential statistics can be especially misleadingwhen carelessly applied logic suggests an unwarranted degreeof deductive certainty.Chow's (1988, Tabl e 2) syllogistic description of hypothesistesting (see my Table 1) illustrates how unwarranted conclu-sions can be reached by seemingly steps. Syllogisms areeither valid or invalid, hence logical validity is an all-or-noneproperty of an (Chow, 1988, p. 109). Chow therebymirrors common practice by emphasizing binary choice be-tween rejecting or not rejecting a theory. Researchers followingcommon practice can dupe themselves and others, however, be-cause neither a theory's validity nor its invalidity is a matter ofdeductiv e certainty.Chow's (1988) idea that the role of a statistical analysis. . .is to supply investigator with minor premise for syl-logistic (p. 108) also raises issues about use ofeffect sizes:All that is required of a statistical analysis is a binary decision.. . .because validity of syllogistic argument requires only thatinformation. Even a quantitatively more informative index isavailable (e.g., effect size, amount of variance accounted for, orthe power of test), it will still be used in a binary manner.. . .Nothing is gained by using an effect-size estimate.. . .(p. 108)Chow claimed that significance tests are superior to effect sizesI thank Bill Dunlap and Irving LaValle for comments.Correspondence concerning this article should be addressed to Rob-ert Folger, A. B. Freeman School of Business, Tulane University, NewOrleans, Louisiana 70118.on basis of logic of syllogistic argument. I interpret thesame syllogism differently, making role of effect sizes oncemore an open issue.No Proof of Validity: Problems inAffirming ConsequencesTh e heading of third colum n i Tabl 1 (Affirming conse-quent) is also e nam of a fallacy, which should imme-diately raise suspicions about imposing either/or (binary) deci-sions when judging theories. The fallacy of affirming conse-quent involves reasoning backward from validity ofconclusions to e presumed validity of their premises. Suppossomeone claims that if we follow principles of supply-sideeconomics, economy will prosper becaus th election ofa supply-side President caused economic prosperity. Amongother interpretive problems (e.g., what does it mean to imple-ment supply-side policies; what measure reveals degree towhich President had in fact done so; and how should eco-nomic prosperity b e measured?), ther is a clea r dange of spuri-ous correlation: How can one be sure that prosperity was notcaused by something else? It is also easy to see danger ofaffirmin g consequent from example s of fallaciou reasoninsuch a s following: If Hitler i still alive, sun hot.The sun is hot. Therefore Hitler is still alive.Chow argued for all-or-nothing, dichototnous choices, evenwhil e using wor d probably twic i n thir colum f Ta-ble 1, yet reference to probability rather than to certaintyis demanded by fallacious nature of consequence-affirmingarguments: Because it is logically false to conclude with deduc-tive certainty that true premises are entailed by true conse-quences, then at best any evidence for consequences' valid-ity provides only arguments consistent with validity of thepremises. It is more accurate to say that experimental resultsare consistent with a hypothesis than to say that they con-firm it.