The flat-maximum effect and generic linear scoring models: a test

Received on 1 July 1991. Predicting human behaviour patterns with linear correlationmodels has absorbed researchers for the past five decades. Althoughmost observers generally concede that humans are inferior tosuch models in combining information, linear scoring modelsare unfortunately, plagued by the flat-maximum effect or the‘curse of insensitivity’. As Lovie & Lovie(1986)observe: ‘The predictive ability of linear models is insensitiveto large variations in the size of regression weights and tothe number of predictors.’ In essence, seemingly differentscoringmodels tend to produce indistinguishable predictive outcomes. Since its demonstration by Dawes & Corrigan (1974), observershave cast the flat maximum in a decidedly negative light. Incontrast, Lovie & Lovie (1986) present a provocatively contrarianview of the flat maximum‘s positive potential. In thissame vein, we examine the predictive power of a generic credit-scoringmodel versus individual empirically derived systems. If, asWainer (1976) noted in regard to the flat maximum, ’itdon‘t make no nevermind’, generic credit-scoringmodels could provide cheaper alternatives to individual empiricallyderived models. During the period 1984–8, a series of linear credit-scoringmodels were developed for ten Southeastern U.S. credit unions.For each credit union, stepwise multiple regression was employedto select a subset of explanatory variables to be used in adiscriminant analysis. A generic credit-scoring equation wasdeveloped from the resulting discriminant analyses using weightedaverage coefficients from five systems. The predictive powerof the generic model was compared to the predictive power ofholdout sample of the five remaining credit-scoring models. In all cases, the generic model's performance was very closeto that of the empirically derived models. Thus, our findingssupport Lovie & Lovie's (1986) challenge to the conventionalwisdom that the flat maximum casts a pall on the successfulmodelling of judgement processes. Indeed, the flat maximum impliesa positive role for simpler, and hence cheaper, generic models.Although further research is needed, it should be possible todevelop hybrid models with generic cores that perform as wellas empirically derived linear models.     

U impurities in Au

The properties of U impurities in Au have been studied by magnetic susceptibility, electrical resistivity, and electron spectroscopic measurements. The results indicate that U forms a well-defined local magnetic moment of 3.6 _B , giving rise to a Kondo resistance anomaly at low temperature. This is in agreement with spectroscopic data, which show localized impurity states well removed from the Fermi level. Comparison to calculated atomic multiplets shows the need for inclusion of crystal or ligand field effects for the multiplets. The intra-5f Coulomb correlation energy is found to be 2.6 eV.     

Synthesis,Crystal Structure,and Optical Properties of (CH3NH3)2CoX4 (X = Cl,Br, I,Cl0.5Br0.5, Cl0.5I0.5, Br0.5I0.5)

The reaction of methylammonium halides and cobalt halides yielded the organic‐inorganic hybrid compounds of general formula (CH₃NH₃)₂CoX₄. By varying the different halides, we were able to synthesize the whole row from Cl to I as well as some mixed halides compounds and to determinate the crystal structures. (CH₃NH₃)₂CoX₄ (X = Cl, Br, Cl_0.5Br_0.5, Br_0.5I_0.5) crystallize isotypic to (CH₃NH₃)₂HgCl₄ in space group P2₁/c with Z = 4 [X = Cl: a = 7.6483(9), b = 12.6885(18), c = 10.8752(12) Å, β = 96.639(9)°; X = Cl_0.5Br_0.5: a = 7.8271(9), b = 12.9543(9), c = 11.1007(11) Å, β = 96.320(8)°; X = Br: a = 7.9782(2), b = 13.1673(2), c = 11.2602(2) Å, β = 96.3260(10)° and X = Br_0.5I_0.5: a = 8.2435(12), b = 13.645(2), c = 11.5856(18) Å, β = 95.54(2)°]. The mixed halides show a statistic distribution in both cases. In (CH₃NH₃)₂CoCl₂I₂ an ordered variant is realized representing a new structure type [C2/m, Z = 4, a = 18.808(4), b = 7.3604(7), c = 10.4109(17) Å, β = 120.364(13)°]. (CH₃NH₃)₂CoI₄ crystallizes again isotypic to the respective mercury compound [(CH₃NH₃)₂HgCl₄] [Pbca, Z = 8, a = 10.9265(5), b = 12.1552(5), c = 20.9588(9) Å]. All structures are build up by inorganic tetrahedral [CoX₄]^2– anions and organic (CH₃NH₄)⁺ cations. Additionally the Raman spectra as well as the optical reflectance spectra are discussed.     

Review of Chemometrics Applied to Spectroscopy: 1985-95, Part 2

INTRODUCTION

Before beginning Part 2 of this review, a caveat noted by Deming and Palasota is brought to the reader's attention: [1] “Press et al. [2] have emphasized that data ‘consist of numbers, of course. But these numbers are fed into the computer, not produced by it. These are numbers to be treated with considerable respect, never to be tampered with, nor subjected to a numerical process whose character you do not completely understand. You are well advised to acquire reverence for data that is rather diferent fiom the “sporty” attitude which is sometimes allowable, or even commendable, in other numerical tasks.’ Yet by and large within chemometrics, preprocessing often seems to be carried out with little understanding of its fundamental efect on the structure of the data.”     

Photoinduced surface potential change of bacteriorhodopsin mutant D96N measured by scanning surface potential microscopy

We report the direct measurement of photoinduced surface potential differences of wild-type (WT) and mutant D96N bacteriorhodopsin (BR) membranes at pH 7 and 10.5. Atomic force microscopy (AFM) and scanning surface potential microscopy (SSPM) were used to measure the BR membrane with the extracellular side facing up. We present AFM and SSPM images of WT and mutant D96N in which the light-dark transition occurred in the mid-scan of a single BR membrane. Photosteady-state populations of the M state were generated to facilitate measurement in each sample. The photoinduced surface potential of D96N is 63 mV (peak to valley) at pH 10.5 and is 48 mV at pH 7. The photoinduced surface potential of WT is 37 mV at pH 10.5 and approximately 0 at pH 7. Signal magnitudes are proportional to the amount of M produced at each pH. The results indicated that the surface potentials were generated by photoformation of surface charges on the extracellular side of the membrane. Higher surface potential correlated with a longer lifetime of the charges. A mechanistic basis for these signals is proposed, and it is concluded that they represent a steady-state measurement of the B2 photovoltage.     

High-Modulus Wholly Aromatic Fibers. II. Partially Ordered Polyamide-Hydrazides

Partially ordered polyarnide-hydrazides were produced by the poly condensation of diacid chlorides with aminobenz-hydrazides, the order that results being a consequence of the considerably more rapid reaction of a diacid chloride with the hydrazide group in competition with the aromatic amine group of the aminobenzhydrazide. Fibers were produced from a series of such polymers containing from 50 mole % meta-oriented phenylene rings to 100 mole % para-oriented ones. Fiber from the wholly para-oriented type of polymer exhibited very high strength and modulus: 12.5 and 468 g/den, respectively, at 4.3% elongation-to-break. Although the crystallinity and density observed for hot-drawn fibers of partially ordered completely para-oriented polyarnide-hydrazides were comparable to the crystallinity and density of fibers of the isomeric wholly ordered polymer, the partially ordered polymers were more readily spun to the ultra-high strength and high modulus type fibers, probably because their greater solubility made them easier to spin.     

High-Modulus Wholly Aromatic Fibers. I. Wholly Ordered Polyamide-Hydrazides and Poly-1,3,4-oxadiazole—Amides

Several completely ordered polyamide-hydrazide copolymers were prepared via low temperature poly condensation of aromatic diacid chlorides with symmetrical aromatic diamines containing preformed dihydrazide linkages. Highly crystalline, hot-drawn fibers of the polyamide-hydrazide containing only para-oriented phenylene units showed unusually high strength and exceptionally high initial modulus: 10.8 and 508 g/den, respectively, at 2.9% elongation-to-break. The as-spun fibers also exhibited rather high tensile strength and unusually high initial modulus: 8.2 and 291 g/den, respectively, at 9.4% elongation-to-break. The hot-drawn fiber retained considerable strength at elevated temperatures, exhibiting a tenacity of 1.4 g/den and an initial modulus of 169 g/den at 350°C. Heat-aging of the as-spun fiber at 185°C in air showed that 66% of the original tenacity, 41% of the elongation, and 86% of the modulus were retained even after 336 hr. Substitution of as little as 25 mole % meta-oriented phenylene rings for para-oriented ones resulted in loss of the ultra-high strength and modulus, giving tensile properties comparable to those of fibers from wholly aromatic polyamides of the meta-oriented type. Fibers from the polyamide-hydrazides containing 50 mole % meta-oriented rings showed similar properties. Although ordered oxadiazole-amide copolymers were obtained from diamines containing two preformed oxadiazole linkages separated by m-phenylene rings, fibers could not be spun from them. Fiber of an ordered oxadiazole-amide copolymer was obtained, however, by heat treatment of the wholly p-phenylene ordered poly amide-hydrazide copolymer precursor fiber. Such a fiber exhibited a tenacity of 15.3 g/den, 3.6% elongation-to-break, and 564 g/den initial modulus.