Effect sizes (ES) tell the magnitude of the difference between treatments and, ideally, should tell clinicians how likely their patients will benefit from the treatment. Currently used ES are expressed in statistical rather than in clinically useful terms and may not give clinicians the appropriate information. We restrict our discussion to studies with two groups: one with n patients receiving a new treatment and the other with m patients receiving the usual or no treatment. The standardized mean difference (e.g. Cohen's d) is a well-known index for continuous outcomes. There is some intuitive value to d, but measuring improvement in standard deviations (SD) is a statistical concept that may not help a clinician. How much improvement is a half SD? A more intuitive and simple-to-calculate ES is the probability that the response of a patient given the new treatment (X) is better than the one for a randomly chosen patient given the old or no treatment (Y) (i.e. P(X > Y), larger values meaning better outcomes). This probability has an immediate identity with the area under the curve (AUC) measure in procedures for receiver operator characteristic (ROC) curve comparing responses to two treatments. It also can be easily calculated from the Mann-Whitney U, Wilcoxon, or Kendall tau statistics. We describe the characteristics of an ideal ES. We propose P(X > Y) as an alternative index, summarize its correspondence with well-known non-parametric statistics, compare it to the standardized mean difference index, and illustrate with clinical data.

The current article presents a systematic approach to theory pruning (defined here as hypothesis specification and study design intended to bound and reduce theory). First, we argue that research that limits theory is underrepresented in the organizational sciences, erring overwhelmingly on the side of confirmatory null hypothesis testing. Second, we propose criteria for determining comparability, deciding when it is appropriate to test theories or parts of theories against one another. Third, we suggest hypotheses or questions for testing competing theories. Finally, we revisit the spirit of ‘‘strong inference.’’ We present reductionist strategies appropriate for the organizational sciences, which extend beyond traditional approaches of ‘‘critical’’ comparisons between whole theories. We conclude with a discussion of strong inference in organizational science and how theory pruning can help in that pursuit.

No abstract

Manufacturers of pharmaceuticals and biopharmaceuticals are facing increased regulatory pressure to understand how their manufacturing processes work and to be able to quantify the reliability and robustness of their manufacturing processes. In particular, the ICH Q8 guidance has introduced the concept of design space. The ICH Q8 defines design space as "the multidimensional combination and interaction of input variables (e.g., material attributes) and process parameters that have been demonstrated to provide assurance of quality." However, relatively little has been put forth to date on how to construct a design space from data composed of such variables. This study presents a Bayesian approach to design space based upon a type of credible region first appearing in Peterson's work.This study considers the issues of constructing a Bayesian design space, design space reliability, the inclusion of process noise variables, and utilization of prior information, as well as an outline for organizing information about a design space so that manufacturing engineers can make informed changes as may be needed within the design space.

HfO 2 films have been grown with two atomic layer deposition (ALD) chemistries: (a) tetrakis(ethylmethylamino)hafnium (TEMAHf)+O3 and (b) HfCl4+H2O. The resulting films were studied as a function of ALD cycle number on Si(100) surfaces prepared with chemical oxide, HF last, and NH3 annealing. TEMAHf+O3 growth is independent of surface preparation, while HfCl4+H2O shows a surface dependence. Rutherford backscattering shows that HfCl4+H2O coverage per cycle is l3% of a monolayer on chemical oxide while TEMAHf+O3 coverage per cycle is 23% of a monolayer independent of surface. Low energy ion scattering, x-ray reflectivity, and x-ray photoelectron spectroscopy were used to understand film continuity, density, and chemical bonding. TEMAHf+O3 ALD shows continuous films, density >9g∕cm3, and bulk Hf–O bonding after 15 cycles [physical thickness (Tphys)=1.2±0.2nm] even on H-terminated Si(100). Conversely, on H-terminated Si(100), HfCl4+H2O requires 50 cycles (Tphys∼3nm) for continuous films and bulk Hf–O bonding. TEMAHf+O3 ALD was implemented in HfO2∕TiN transistor gate stacks, over the range 1.2nm⩽Tphys⩽3.3nm. Electrical results are consistent with material analysis suggesting that at Tphys=1.2nm HfO2 properties begin to deviate from thick film properties. At Tphys=1.2nm, electrical thickness scaling slows, gate current density begins to deviate from scaling trendlines, and no hard dielectric breakdown occurs. Most importantly, n-channel transistors show improvement in peak and high field electron mobility as Tphys scales from 3.3 to 1.2nm. This improvement may be attributed to reduced charge trapping and Coulomb scattering in thinner films. Scaled HfO2 enables 1nm equivalent oxide thickness and 82% of universal SiO2 mobility.