Search

We illustrate the application of mixture IRT [item response theory] models to evaluate the possibility of respondent confusion due to the negative wording of certain items on a social-emotional learning (SEL) assessment. Using actual student self-report ratings on four social-emotional learning scales collected from students in grades 3-12 from CORE districts in the state of California, we also evaluate the consequences of the potential confusion in biasing student- and school-level scores as well as correlational relationships between SEL and student-level variables. Models of both full and partial confusion are examined. Our results suggest that (1) rating scale confusion due to negatively-worded items does appear to be present; (2) the confusion is most prevalent at lower grade levels (3rd-5th); and (3) the occurrence of confusion is positively related to reading proficiency and ELL status, as anticipated, and bias estimates of SEL correlations with these student-level variables. For these reasons, we suggest future iterations of the SEL measures use only positively oriented items. To maintain measurement continuity, we suggest bias corrections based on the studied mixture model may be useful, although the precision of such corrections is sensitive to the nature of confusion (e.g., full versus partial).

Phosphorus has historically been a limiting nutrient in the Florida Everglades. Increased P loading to the Everglades over the past several decades has led to significant changes in water quality and plant communities. Stormwater runoff that drains agricultural lands and enters the Water Conservation Areas (WCAs) are known to contain elevated levels of P, but the exact source of this P has not been fully determined. Here the results of an O isotope study of dissolved inorganic phosphate (DIP) in both polluted and relatively pristine (or reference) areas of the Everglades are reported. The data reveal spatial and temporal variations in the delta 18O signature of DIP, reflecting the source and the degree of cycling of P. The delta 18O values of DIP collected from the Everglades National Park were close or equal to the predicted delta 18O values of DIP formed in situ in equilibrium with ambient water, indicating that P is quickly cycled in the water column in oligotrophic ecosystems with very low P concentrations. However, most DIP samples collected from areas impacted by agricultural runoff yielded delta 18O values that deviated from the predicted equilibrium DIP delta 18O values based on the delta 18O of water and water temperature, suggesting that biological cycling of P was not rapid enough to remove the fertilizer ?18O signature in the DIP pool from areas receiving high P loading. The delta 18O signature of DIP in impacted areas reflects a mixing of fertilizer P and biologically cycled P, where the relative proportions of biologically cycled vs. fertilizer DIP are controlled by both biological (microbial activities and plant uptake) and hydrologic factors (loading rate and residence time). Using a two-end-member (i.e., fertilizer P and biologically cycled P) mixing model, fertilizers were estimated to contribute about 15 100% of the DIP pool in the highly impacted areas of the northern Everglades, whereas the DIP pool in the reference (i.e., relatively pristine) wetlands in the Everglades National Park was dominated by biologically cycled P. The study shows that O isotopic measurements of dissolved PO(exp 3-, sub 4) can be a useful tool for tracing the fertilizer P inputs to freshwater ecosystems.

In this paper, calculations of the hole initiated interband impact ionization rate in bulk silicon and GaAs are presented based on an ensemble Monte Carlo simulation with the inclusion of a wave-vector-dependent numerical transition rate formulation. The ionization transition rate is determined for each of the three valence bands, heavy, light, and split-off, using Fermi's golden rule with a two-body, screened Coulomb interaction. The dielectric function used within the calculation is assumed to be wave-vector-dependent. Calculations of the field-dependent impact ionization rate as well as the quantum yield are presented. It is found from both the quantum yield results and examination of the hole distribution function that the effective threshold energy for hole initiated impact ionization is relatively soft, similar to that predicted for the corresponding electron initiated ionization events occur more frequently than either heavy or split-offf initiated ionization events in bulk silicon over the applied electric field strengths examined here, 250-500 kV/cm. Conversely,in GaAs, the vast majority of hole initated ionization events originate from holes within the split-off band.

The hole initiated impact ionization rate in bulk silicon and GaAs is calculated using a numerical formulation of the impact ionization transition rate incorporated into an ensemble Monte Carlo simulation. The transition rate is calculated from Fermi's golden rule using a two-body screened Coulomb interaction including a wavevector dependent dielectric function. It is found that the effective threshold for hole initiated ionization is relatively soft in both materials, that the split-off band dominates the ionization process in GaAs. and that no clear dominance by any one band is observed in silicon, though the rate out of the light hole band is greatest.

The reliability of doped-barrier AlGaAs/GsAs multi-quantum well avalanche photodiodes fabricated by molecular beam epitaxy is investigated via accelerated life tests. Dark current and breakdown voltage were the parameters monitored. The activation energy of the degradation mechanism and median device lifetime were determined. Device failure probability as a function of time was computed using the lognormal model. Analysis using the electron beam induced current method revealed the degradation to be caused by ionic impurities or contamination in the passivation layer.

The electron interband impact-ionization rate for both silicon and gallium arsenide is calculated using an ensemble Monte Carlo simulation with the expressed purpose of comparing different formulations of the interband ionization transition rate. Specifically, three different treatments of the transition rate are examined: the traditional Keldysh formula, a new k-dependent analytical formulation first derived by W. Quade, E. Scholl, and M. Rudan (1993), and a more exact, numerical method of Y. Wang and K. F. Brennan (1994). Although the completely numerical formulation contains no adjustable parameters and as such provides a very reliable result, it is highly computationally intensive. Alternatively, the Keldysh formular, although inherently simple and computationally efficient, fails to include the k dependence as well as the details of the energy band structure. The k-dependent analytical formulation of Quade and co-workers overcomes the limitations of both of these models but at the expense of some new parameterization. It is found that the k-dependent analytical method of Quade and co-workers produces very similar results to those obtained with the completely numerical model for some quantities. Specifically, both models predict that the effective threshold for impact ionization in GaAs and silicon is quite soft, that the majority of ionization events originate from the second conduction band in both materials, and that the transition rate is k dependent. Therefore, it is concluded that the k-dependent analytical model can qualitatively reproduce results similar to those obtained with the numerical model yet with far greater computational efficiency. Nevertheless, there exist some important drawbacks to the k-dependent analytical model of Quade and co-workers: These are that it does not accurately reproduce the quantum yield data for bulk silicon, it requires determination of a new parameter, related physically to the overlap intergrals of the Bloch state which can only be adjusted by comparison to experiment, and fails to account for any wave-vector dependence of the overlap integrals. As such the transition rate may be overestimated at those points for which 'near vertical,' small change in k, transitions occur.

Ensemble Monte Carlo calculations of the electron interband impact ionization rate in bulk GaAs are presented using a wave-vector (k)-dependent formulation of the ionization transition rate. The transition rate is evaluated through the use of numerically generated wavefunctions determined via a k-p calculation within the first two conduction bands at numerous points within a finely spaced three-dimensional grid in k space. The transition rate is determined to be greatest for states within the second conduction band. Is is found that the interband impact ionization transition rate in bulk GaAs is best characterized as having an exceedingly soft threshold energy. As a consequence, the dead space, defined as the distance over which the ionization probability for a given carrier is assumed to be zero, is estimated to be much larger than that estimated using a harder threshold. These results have importance in the design of the multiquantum-well avalanche photodiodes.

A model is presented describing the factors and processes which determine the measured C-14 ages of soil calcium carbonate. Pedogenic carbonate forms in isotopic equilium with soil CO2. Carbon dioxide in soils is a mixture of CO2 derived from two biological sources: respiration by living plant roots and respiration of microorganisms decomposing soil humus. The relative proportion of these two CO2 sources can greatly affect the initial C-14 content of pedogenic carbonate: the greater the contribution of humus-derived CO2, the greater the initial C-14 age of the carbonate mineral. For any given mixture of CO2 sources, the steady-state (14)CO2 distribution vs. soil depth can be described by a production/diffusion model. As a soil ages, the C-14 age of soil humus increases, as does the steady-state C-14 age of soil CO2 and the initial C-14 age of any pedogenic carbonate which forms. The mean C-14 age of a complete pedogenic carbonate coating or nodule will underestimate the true age of the soil carbonate. This discrepancy increases the older a soil becomes. Partial removal of outer (and younger) carbonate coatings greatly improves the relationship between measured C-14 age and true age. Although the production/diffusion model qualitatively explains the C-14 age of pedogenic carbonate vs. soil depth in many soils, other factors, such as climate change, may contribute to the observed trends, particularily in soils older than the Holocene.

We examine the basic design issues in the optimization of GaAs delta-doped and AlGAs delta-doped quantum-well avalanche photodiode (APD) structures using a theoretical analysis based on an ensemble Monte Carlo simulation. The devices are variations of the p-i-n doped quantum-well structure previously described in the literature. They have the same low-noise, high-gain and high-bandwidth features as the p-i-n doped quantum-well device. However, the use of delta doping provides far greater control or the doping concentrations within each stage possibly enhancing the extent to which the device can be depleted. As a result, it is expected that the proposed devices will operate at higher gain levels (at very low noise) than devices previously developed.

We present calculations of the interband impact ionization rate calculated using a wave vector dependent (k-dependent) semiclassical formulation of the transition rate. The transition rate is determined using Fermi's golden rule from a two-body screened Coulomb interaction assuming energy and momentum conservation. The transition rate is calculated for the first two conduction bands of silicon by numerically integrating over the full Brillouin zone. The overlap integrals in the expression for the transition rate are determined numerically using a 15 band k-p calculation. It is found that the transition rate depends strongly on the initiating electron wave vector (k vector) and that the transition rate is greatest for electrons originating within the second conduction band than the first conduction band. An ensemble Monte Carlo simulation, which includes the numerically determined ionization transition rate as well as the full details of the first two conduction bands, is used to calculate the total impact ionization rate in bulk silicon. Good agreement with the experimentally determined electron ionization rate data is obtained.