ABSTRACT
All professional ethical codes stress the importance of honesty and personal integrity, the striving for objectivity, and the avoidance of any attempt to mislead by virtue of professional status. Professional associations for those working in quantitative disciplines, such as the American Psychological Association (APA), the Royal Statistical Society (RSS), the International Statistical Institute (ISI), and the American Statistical Association (ASA), additionally stress the ethical imperative to make use of appropriate and generally accepted technical standards when collecting and analyzing data (APA, 2002; ASA, 1999; ISI, 1985; RSS, 1993). The ASA, for example, makes some specific technical points, such as the recognition that any frequentist statistical test has a nonzero probability of producing a “significant” result when the effect being tested is not present and warning against deliberately selecting the one “significant” result from a large number of tests. The ASA (1999) is also clear that statisticians and those carrying out statistical analyses should “remain current in terms of statistical methodology: yesterday’s preferred methods may be barely acceptable today.”
