Sources of selection bias in evaluating social programs: An interpretation of conventional measures and evidence on the effectiveness of matching as a program evaluation method

This paper decomposes the conventional measure of selection bias in observational studies into three components. The first two components are due to differences in the distributions of characteristics between participant and nonparticipant (comparison) group members: the first arises from differences in the supports, and the second from differences in densities over the region of common support. The third component arises from selection bias precisely defined. Using data from a recent social experiment, we find that the component due to selection bias, precisely defined, is smaller than the first two components. However, selection bias still represents a substantial fraction of the experimental impact estimate. The empirical performance of matching methods of program evaluation is also examined. We find that matching based on the propensity score eliminates some but not all of the measured selection bias, with the remaining bias still a substantial fraction of the estimated impact. We find that the support of the distribution of propensity scores for the comparison group is typically only a small portion of the support for the participant group. For values outside the common support, it is impossible to reliably estimate the effect of program participation using matching methods. If the impact of participation depends on the propensity score, as we find in our data, the failure of the common support condition severely limits matching compared with random assignment as an evaluation estimator.

Chemical phenotype matching between a plant and its insect herbivore

Two potential outcomes of a coevolutionary interaction are an escalating arms race and stable cycling. The general expectation has been that arms races predominate in cases of polygenic inheritance of resistance traits and permanent cycling predominates in cases in which resistance is controlled by major genes. In the interaction between Depressaria pastinacella, the parsnip webworm, and Pastinaca sativa, the wild parsnip, traits for plant resistance to insect herbivory (production of defensive furanocoumarins) as well as traits for herbivore “virulence” (ability to metabolize furanocoumarins) are characterized by continuous heritable variation. Furanocoumarin production in plants and rates of metabolism in insects were compared among four midwestern populations; these traits then were classified into four clusters describing multitrait phenotypes occurring in all or most of the populations. When the frequency of plant phenotypes belonging to each of the clusters is compared with the frequency of the insect phenotypes in each of the clusters across populations, a remarkable degree of frequency matching is revealed in three of the populations. That frequencies of phenotypes vary among populations is consistent with the fact that spatial variation occurs in the temporal cycling of phenotypes; such processes contribute in generating a geographic mosaic in this coevolutionary interaction on the landscape scale. Comparisons of contemporary plant phenotype distributions with phenotypes of herbarium specimens collected 9–125 years ago from across a similar latitudinal gradient, however, suggest that for at least one resistance trait—sphondin concentration—interactions with webworms have led to escalatory change.

Two-dimensional infrared spectroscopy of peptides by phase-controlled femtosecond vibrational photon echoes

Two-dimensional infrared spectra of peptides are introduced that are the direct analogues of two- and three-pulse multiple quantum NMR. Phase matching and heterodyning are used to isolate the phase and amplitudes of the electric fields of vibrational photon echoes as a function of multiple pulse delays. Structural information is made available on the time scale of a few picoseconds. Line narrowed spectra of acyl-proline-NH2 and cross peaks implying the coupling between its amide-I modes are obtained, as are the phases of the various contributions to the signals. Solvent-sensitive structural differences are seen for the dipeptide. The methods show great promise to measure structure changes in biology on a wide range of time scales.

Sequence tag identification of intact proteins by matching tanden mass spectral data against sequence data bases.

Molecular and fragment ion data of intact 8- to 43-kDa proteins from electrospray Fourier-transform tandem mass spectrometry are matched against the corresponding data in sequence data bases. Extending the sequence tag concept of Mann and Wilm for matching peptides, a partial amino acid sequence in the unknown is first identified from the mass differences of a series of fragment ions, and the mass position of this sequence is defined from molecular weight and the fragment ion masses. For three studied proteins, a single sequence tag retrieved only the correct protein from the data base; a fourth protein required the input of two sequence tags. However, three of the data base proteins differed by having an extra methionine or by missing an acetyl or heme substitution. The positions of these modifications in the protein examined were greatly restricted by the mass differences of its molecular and fragment ions versus those of the data base. To characterize the primary structure of an unknown represented in the data base, this method is fast and specific and does not require prior enzymatic or chemical degradation.

Two different but related mechanisms are used in plants for the repair of genomic double-strand breaks by homologous recombination.

Genomic double-strand breaks (DSBs) are key intermediates in recombination reactions of living organisms. We studied the repair of genomic DSBs by homologous sequences in plants. Tobacco plants containing a site for the highly specific restriction enzyme I-Sce I were cotransformed with Agrobacterium strains carrying sequences homologous to the transgene locus and, separately, containing the gene coding for the enzyme. We show that the induction of a DSB can increase the frequency of homologous recombination at a specific locus by up to two orders of magnitude. Analysis of the recombination products demonstrates that a DSB can be repaired via homologous recombination by at least two different but related pathways. In the major pathway, homologies on both sides of the DSB are used, analogous to the conservative DSB repair model originally proposed for meiotic recombination in yeast. Homologous recombination of the minor pathway is restricted to one side of the DSB as described by the nonconservative one-sided invasion model. The sequence of the recombination partners was absolutely conserved in two cases, whereas in a third case, a deletion of 14 bp had occurred, probably due to DNA polymerase slippage during the copy process. The induction of DSB breaks to enhance homologous recombination can be applied for a variety of approaches of plant genome manipulation.

Productivity and Credit Constraints: Firm-Level Evidence from Propensity Score Matching. Factor Markets Working Paper No. 3, September 2011

Drawing on a unique, farm-level panel dataset with 37,409 observations and employing a matching estimator, this paper analyses how farm access to credit affects farm input allocation and farm efficiency in the Central and Eastern European transition countries. We find that farms are asymmetrically credit constrained with respect to inputs. Farm use of variable inputs and capital investment increases up to 2.3% and 29%, respectively, per €1,000 of additional credit. Our estimates also suggest that farm access to credit increases total factor productivity up to 1.9% per €1,000 of additional credit, indicating that an improvement in access to credit results in an adjustment in the relative input intensities on farms. This finding is further supported by a negative effect of better access to credit on labour, suggesting that these two are substitutes. Interestingly, farms are found not to be credit constrained with respect to land.

Applying machine learning to catalogue matching in astrophysics

We present the results of applying automated machine learning techniques to the problem of matching different object catalogues in astrophysics. In this study, we take two partially matched catalogues where one of the two catalogues has a large positional uncertainty. The two catalogues we used here were taken from the H I Parkes All Sky Survey (HIPASS) and SuperCOSMOS optical survey. Previous work had matched 44 per cent (1887 objects) of HIPASS to the SuperCOSMOS catalogue. A supervised learning algorithm was then applied to construct a model of the matched portion of our catalogue. Validation of the model shows that we achieved a good classification performance (99.12 per cent correct). Applying this model to the unmatched portion of the catalogue found 1209 new matches. This increases the catalogue size from 1887 matched objects to 3096. The combination of these procedures yields a catalogue that is 72 per cent matched.

Crystallographic study of grain refinement in aluminum alloys using the edge-to-edge matching model

The edge-to-edge matching model for describing the interfacial crystallographic characteristics between two phases that are related by reproducible orientation relationships has been applied to the typical grain refiners in aluminum alloys. Excellent atomic matching between Al3Ti nucleating substrates, known to be effective nucleation sites for primary Al, and the Al matrix in both close packed directions and close packed planes containing these directions have been identified. The crystallographic features of the grain refiner and the Al matrix are very consistent with the edge-to-edge matching model. For three other typical grain refiners for Al alloys, TiC (when a = 0.4328 nm), TiB2 and AIB(2), the matching only occurs between the close packed directions in both phases and between the second close packed plane of the Al matrix and the second close packed plane of the refiners. According to the model, it is predicted that Al3Ti is a more powerful nucleating substrate for Al alloy than TiC, TiB2 and AlB2. This agrees with the previous experimental results. The present work shows that the edge-to-edge matching model has the potential to be a powerful tool in discovering new and more powerful grain refiners for Al alloys. (C) 2004 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

The correspondence framework for 3D surface matching algorithms

Beyond the inherent technical challenges, current research into the three dimensional surface correspondence problem is hampered by a lack of uniform terminology, an abundance of application specific algorithms, and the absence of a consistent model for comparing existing approaches and developing new ones. This paper addresses these challenges by presenting a framework for analysing, comparing, developing, and implementing surface correspondence algorithms. The framework uses five distinct stages to establish correspondence between surfaces. It is general, encompassing a wide variety of existing techniques, and flexible, facilitating the synthesis of new correspondence algorithms. This paper presents a review of existing surface correspondence algorithms, and shows how they fit into the correspondence framework. It also shows how the framework can be used to analyse and compare existing algorithms and develop new algorithms using the framework's modular structure. Six algorithms, four existing and two new, are implemented using the framework. Each implemented algorithm is used to match a number of surface pairs. Results demonstrate that the correspondence framework implementations are faithful implementations of existing algorithms, and that powerful new surface correspondence algorithms can be created. (C) 2004 Elsevier Inc. All rights reserved.

Matching of catalogues by probabilistic pattern classification

We consider the statistical problem of catalogue matching from a machine learning perspective with the goal of producing probabilistic outputs, and using all available information. A framework is provided that unifies two existing approaches to producing probabilistic outputs in the literature, one based on combining distribution estimates and the other based on combining probabilistic classifiers. We apply both of these to the problem of matching the HI Parkes All Sky Survey radio catalogue with large positional uncertainties to the much denser SuperCOSMOS catalogue with much smaller positional uncertainties. We demonstrate the utility of probabilistic outputs by a controllable completeness and efficiency trade-off and by identifying objects that have high probability of being rare. Finally, possible biasing effects in the output of these classifiers are also highlighted and discussed.

Receiving mutual impedance between two normal-mode helical antennas (NMHAs)

A new mutual impedance - the receiving mutual impedance - between two normal-mode helical antennas is defined, measured, and theoretically calculated. The variations of the receiving mutual impedance with antenna separation, with frequency, and with excitation source direction are critically investigated. An application of the receiving mutual impedance in direction finding demonstrates its more accurate description of the mutual coupling effect than that using the conventional mutual impedance.

Understanding the crystallography of the eutectoid microstructure in a Zn-Al alloy using the edge-to-edge matching model

The orientation relationship (OR) between the beta(Zn) phase and the alpha(Al) phase and the corresponding habit planes in a Zn-Al eutectoid alloy were accurately determined using convergent beam Kikuchi line diffraction patterns. In addition to the previously reported OR. [11 (2) over bar0](beta)parallel to[110](alpha), (0002)(beta)parallel to ((1) over bar 11)alpha, two new ORs were observed. They are: [11 (2) over bar0](beta)parallel to [110], ((1) over bar 101)(beta) 0.82 degrees from (002)(alpha) and [(1) over bar 100](beta)parallel to[112](alpha), (0002)(beta) 4.5 degrees from (111)(alpha). These ORs can be explained and understood using the recently developed edge-to-edge matching model. (c) 2006 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Imitating the neighbours: vocal dialect matching in a mimic-model system

Vocal mimicry provides a unique system for investigating song learning and cultural evolution in birds. Male lyrebirds produce complex vocal displays that include extensive and accurate mimicry of many other bird species. We recorded and analysed the songs of the Albert's lyrebird (Menura alberti) and its most commonly imitated model species, the satin bowerbird (Ptilonorhynchus violaceus), at six sites in southeast Queensland, Australia. We show that each population of lyrebirds faithfully reproduces the song of the local population of bowerbirds. Within a population, lyrebirds show less variation in song structure than the available variation in the songs of the models. These results provide the first quantitative evidence for dialect matching in the songs of two species that have no direct ecological relationship.

Edge-to-edge matching - The fundamentals

The basis of the present authors' edge-to-edge matching model for understanding the crystallography of partially coherent precipitates is the minimization of the energy of the interface between the two phases. For relatively simple crystal structures, this energy minimization occurs when close-packed, or relatively close-packed, rows of atoms match across the interface. Hence, the fundamental principle behind edge-to-edge matching is that the directions in each phase that correspond to the edges of the planes that meet in the interface should be close-packed, or relatively close-packed, rows of atoms. A few of the recently reported examples of what is termed edge-to-edge matching appear to ignore this fundamental principle. By comparing theoretical predictions with available experimental data, this article will explore the validity of this critical atom-row coincidence condition, in situations where the two phases have simple crystal Structures and in those where the precipitate has a more complex structure.

Machine learning for matching astronomy catalogues

An emerging issue in the field of astronomy is the integration, management and utilization of databases from around the world to facilitate scientific discovery. In this paper, we investigate application of the machine learning techniques of support vector machines and neural networks to the problem of amalgamating catalogues of galaxies as objects from two disparate data sources: radio and optical. Formulating this as a classification problem presents several challenges, including dealing with a highly unbalanced data set. Unlike the conventional approach to the problem (which is based on a likelihood ratio) machine learning does not require density estimation and is shown here to provide a significant improvement in performance. We also report some experiments that explore the importance of the radio and optical data features for the matching problem.