Abstract
A standardized index for effect intensity, the translocation relative to range (TRR), is discussed. TRR is defined as the difference between the expectations of an outcome under two conditions (the absolute increment) divided by the maximum possible amount for that difference. TRR measures the shift caused by a factor relative to the maximum possible magnitude of that shift. For binary outcomes, TRR simply equals the risk difference, also known as the inverse number needed to treat. TRR ranges from –1 to 1 but is – unlike a correlation coefficient – a measure for effect intensity, because it does not rely on variance parameters in a certain population as do effect size measures (e.g., correlations, Cohens d). However, the use of TRR is restricted on outcomes with fixed and meaningful endpoints given, for instance, for meaningful psychological questionnaires or Likert scales. The use of TRR vs. Cohens d is illustrated with three examples from Psychological Science 2006 (issues 5 through 8). It is argued that, whenever TRR applies, it should complement Cohens d to avoid the problems related to the latter. In any case, the absolute increment should complement d.
References
(2006). An examination of sources of peer-review process. Psychological Science, 17, 378–382.
(2006a). Arbitrary metrics in psychology. American Psychologist, 61, 27–41.
(2006b). Arbitrary metrics redux. American Psychologist, 61, 62–70.
(2003). Meta-analysis of raw mean differences. Psychological Methods, 8, 406–418.
(1988). Statistical power analysis for the behavioral sciences (2nd ed.). New York: Erlbaum.
(2006). The continued search for nonarbitrary metrics in psychology. American Psychologist, 61, 50–55.
(1999). Indices of contrast and quantitative significance for comparison of two groups. Statistics in Medicine, 18, 2557–2581.
(2006). Spatial distance and mental construal of social events. Psychological Science, 17, 278–282.
(1987). Interpretation and choice of effect measures in epidemiological analyses. American Journal of Epidemiology, 5, 761–768.
(1993). Basic problems in interaction assessment. Environmental Health Perspectives Supplements, 101 (Suppl. 4), 9–66.
(2004). Model-based estimation of relative risks and other epidemiologic measures in studies of common outcomes and in case-control studies. American Journal of Epidemiology, 160, 301–305.
(1986). The fallacy of employing standardized regression coefficients and correlations as measures of effect. American Journal of Epidemiology, 123, 203–208.
(2005). Causal inference based on counterfactuals. BMC Medical Research Methodology, 5, 28.
(1993). Estimation of standardized parameters from categorical regression models. Statistics in Medicine 1995, 14, 2131–2141.
(2006). Implications for identifying evidence-based treatments. American Psychologist, 61, 42–49.
(2005). An alternative to null-hypothesis significance tests. Psychological Science, 16, 345–353.
(1999). Measuring the potency of a risk factor for clinical or policy significance. Psychological Methods, 4, 257–271.
(2002). Number needed to treat: Easily understood and intuitively meaningful? Theoretical considerations and a randomized trial. Journal of Clinical Epidemiology, 55, 888–892.
(2004). Discovering knowledge in data. An introduction to data mining. New York: Wiley.
(1992). Factors and taxa, traits and types, differences of degree, and differences in kind. Journal of Personality, 60, 117–174.
(1978). Psychometric theory. New York: McGraw-Hill.
(1996). Measures of effect size. Behavior Research Methods, Instruments, and Computers, 28, 12–22.
(2000). Contrasts and effect sizes in behavioral research. A correlational approach. Cambridge: Cambridge University Press.
(2003). requivalent: A simple effect-size indicator. Psychological Methods, 8, 492–496.
(2000). Contrasts and correlations in effect-size estimation. Psychological Science, 11, 446–453.
(1972). The methodology of evaluation. In , Evaluating action programs: Reading in social action and education (pp. 123–136). Boston: Allyn and Bacon.
(2004). Combining estimates of effect size. In , The handbook of research synthesis (pp. 261–281). New York: Russell Sage Foundation Publications.
(2003). Temporal construal. Psychological Review, 110, 403–401.
(2006). Conditions under which function information attenuates name extension via shape. Psychological Science, 17, 367–371.
(1986). Binomial regression in GLIM: Estimating risk ratios and risk differences. American Journal of Epidemiology, 123, 174–184.
