This dissertation investigates the interaction of population invariance, equating assumptions, and equating accuracy with group differences. In addition, matched samples equating methods are considered as a possible way to improve equating accuracy with large group differences. Data from one administration of four mixed-format Advanced Placement (AP) Exams were used to create pseudo old and new forms sharing common items. Population invariance analyses were conducted based on levels of examinee parental education using a single group equating design. Old and new form groups with common item effect sizes (ESs) ranging from 0 to 0.75 were created by sampling examinees based on their level of parental education. Equating was conducted for four common item nonequivalent group design equating methods: frequency estimation, chained equipercentile, IRT true score, and IRT observed score. Additionally, groups with ESs greater than zero were matched using three different matching techniques including exact matching on parental education level and propensity score matching with several other background variables. The accuracy of equating results was evaluated by comparing each equating relationship with an ES greater than zero to the equating relationship where the ES equaled zero. Differences between comparison and criterion equating relationships were quantified using the root expected mean squared difference (REMSD) statistic, classification consistency, and standard errors of equating (SEs). The accuracy of equating results and the adequacy of equating assumptions was compared for unmatched and matched samples. As ES increased, equating results tended to become less accurate and less consistent across equating methods. However, there was relatively little population dependence of equating results, despite large subgroup performance differences. Large differences between criterion and comparison equating relationships appeared to be caused instead by violations of equating assumptions. As group differences increased, the degree to which frequency estimation and chained equipercentile assumptions held decreased. In addition, all four AP Exams showed some evidence of multidimensionality. Because old and new form groups were selected to differ in terms of their respective levels of parental education, the matching methods that included parental education appeared to improve equating accuracy and the degree to which equating assumptions held, at least for very large ESs.

Methamphetamine is the most common amphetamine used and, along with 3,4-methylenedioxymethamphetamine (MDMA, Ecstasy), is considered part of a worldwide drug epidemic. Monitoring metham-phetamine levels in the body is important for purposes of drug screening for employment, criminal investigations, and therapeutic drug monitoring. While methamphetamine is suitable for detection using immunoassay techniques, these methods tend to have significant cross reactivity with other compounds. Over the last decade, more than eighty different quantitative, confirmatory analytical methods for measuring methamphetamine in biological samples have been published in the scientific literature. Analytical instrumentation used in these methods includes gas chromatography-mass spectrometry (GC-MS), liquid chromatography-mass spectrometry (LC-MS), capillary electrophoresis (CE), among others. These assays are capable of quantifying methamphetamine concentrations in a variety of biological matrices, including blood, plasma, urine, hair, and fingernails. Some of these techniques can achieve detection as low as 0.1 ng/mL (1 ppb) concentra-tions. The strengths and limitations of these methodologies will be discussed in the context of methamphetamine analysis. Additionally, methods that can simultaneously measure methamphetamine levels as well as metabolites and other drugs of abuse will be highlighted.

A fundamental issue in educational measurement is what frame of reference to use when interpreting students’ performance on an assessment. One frame of reference that is often used to enhance interpretations of test scores is normative, which adds meaning to test score interpretations by indicating the rank of an individual’s score within a distribution of test scores of a well-defined reference group. One of the most commonly used frames of reference on student achievement provided by test publishers of large-scale assessments is national norms, whereby students’ test scores are referenced to a distribution of scores of a nationally representative sample. A national probability sample can fail to fully represent the population because of student and school nonparticipation. In practice, this is remedied by weighting the sample so that it better represents the intended reference population. The focus of this study was on weighting and determining the extent to which weighting grade 4 and grade 8 student records that are not fully representative of the nation can recover distributions of reading and math scores in a national probability sample. Data from a statewide testing program were used to create six grade 4 and grade 8 datasets, each varying in its degree of representativeness of the nation, as well as in the proximity of its reading and math distributions to those of a national sample. The six datasets created for each grade were separately weighted to different population totals in two different weighting conditions using four different bivariate stratification designs. The weighted distributions were then smoothed and compared to smoothed distributions of the national sample in terms of descriptive statistics, maximum absolute differences between the relative cumulative frequency distributions, and chi-square effect sizes. The impact of using percentile ranks developed from the state data was also investigated. By and large, the smoothed distributions of the weighted datasets were able to recover the national distribution in each content area, grade, and weighting condition. Weighting the datasets to the nation was effective in making the state test score distributions more similar to the national distributions. Moreover, the stratification design that defined weighting cells by the joint distribution of median household income and ethnic composition of the school consistently produced desirable results for the six datasets used in each grade. Log-linear smoothing using a polynomial of degree 4 was effective in making the weighted distributions even more similar to those in the national sample. Investigation of the impact of using the percentile ranks derived from the state datasets revealed that the percentile ranks of the distributions that were most similar to the national distributions resulted in a high percentage of agreement when classifying student performance based on raw scores associated with the same percentile rank in each dataset. The utility of having a national frame of reference on student achievement, and the efficacy of estimating such a frame of reference from existing data are also discussed.

The vast and complex wealth of information available to researchers often leads to a systematic review, which involves a detailed and comprehensive plan and search strategy with the goal of identifying, appraising, and synthesizing all relevant studies on a particular topic. A meta–analysis, conducted ideally as part of a comprehensive systematic review, statistically synthesizes evidence from multiple independent studies to produce one overall conclusion. The increasingly widespread use of meta–analysis has led to growing interest in meta–analytic methods for rare events and sparse data. Conventional approaches tend to perform very poorly in such settings. Recent work in this area has provided options for sparse data, but these are still often hampered when heterogeneity across the available studies differs based on treatment group. Heterogeneity arises when participants in a study are more correlated than participants across studies, often stemming from differences in the administration of the treatment, study design, or measurement of the outcome. We propose several new exact methods that accommodate this common contingency, providing more reliable statistical tests when such patterns on heterogeneity are observed. First, we develop a permutation–based approach that can also be used as a basis for computing exact confidence intervals when estimating the effect size. Second, we extend the permutation–based approach to the network meta–analysis setting. Third, we develop a new exact confidence distribution approach for effect size estimation. We show these new methods perform markedly better than traditional methods when events are rare, and heterogeneity is present.

Soil microorganisms play important roles in maintaining soil quality and ecosystem health. Development of effective methods for studying the composition, diversity, and behavior of microorganisms in soil habitats is essential for a broader understanding of soil quality. Forest management strategies and practices are of vital significance for sustainable forest production. How the different forest management measures will influence soil microbial communities is a widespread concern of forest industry and scientific communities. Only a small proportion (~0.1%) of the bacteria from natural habitats can be cultured on laboratory growth media. Direct extraction of whole-community DNA from soil, followed by polymerase chain reaction (PCR) and other analysis circumvents the problems of the culture-dependent methods and may shed light on a broader range of microbial communities in the soil. DNA-based molecular methods rely on high quality soil microbial DNA as template, and thus extraction of good quality DNA from soil samples has been a challenge because of the complex and heterogeneous nature of the soil matrix. The objectives of this research were to establish a set of DNA-based molecular methods and to apply them to investigate forest soil microbial composition and diversity. Soil samples were collected from different forest ecosystems, i.e., the natural forest (YNF) and the first rotation (~ 50 years) (Y1R) and the second rotation (~ 1 year) (Y2R) of hoop pine plantations at Yarraman, and from different forest residue management practices (the experiments had established 6.4 years before the samples were collected) at Gympie, two long-term experimental sites of the Queensland Department of Primary Industry-Forestry in subtropical Queensland, Australia. Some DNA-based molecular techniques, including DNA extraction and purification, PCR amplification, DNA screening, cloning, sequencing and phylogenetic analyses, were explored using Yarraman soil samples, which were high in organic matter, clay and iron oxide contents. A set of methods was assembled based on the recommendations of the method development experiments and applied to the investigations of the microbial composition and diversity of the Yarraman and Gympie soil samples. Four soil DNA extraction methods, including the Zhou method (Zhou et al., 1996), the Holben method (Holben, 1994), the UltraClean (Mo Bio) and FastDNA (Bio 101) soil DNA extraction kits, were explored. It was necessary to modify these methods for Yarraman soil. I designed and introduced a pre-lysis buffer washing step, to partially remove soil humic substances and promote soil dispersion. This modification greatly improved the quality of the extracted DNA, decreasing co-extracted humic substances by 31% and increasing DNA yield by 24%. The improved Holben method was recommended for fungal community studies, and the improved Zhou method for bacterial community studies. The extracted DNA was good in quality, with a consistent size of ~20 kb and a yield of 48-87 g g-1 soil, and could be successfully used for 16S (Zhou method) and 18S (Holben method) rDNA amplifications. For less difficult environmental samples, UltraClean kits could be a good option, because they are simple and fast and the extracted DNA are also of good quality. Screening of the DNA PCR products using TGGE, Heteroduplex-TGGE and SSCP was also explored. These methods were not so effective for the screening of the soil DNA PCR products, owing to the difficulty in interpretation of the results. Cloning was a necessary step to obtain a single sequence at species level in soil microbial community studies. The screening of the clone library by TGGE, Heteroduplex-TGGE and SSCP could only separate the clones into several major bands, although SSCP gave better separation. Sequencing of selected clones directly from the clone library obtained ultimate results of microbial taxonomic composition and diversity through well-established sequence analysis software packages and the databases. It was recommended that, in this project with the target of microbial community composition and diversity, soil DNA PCR products were directly cloned to construct clone libraries and a sample of clones were sequenced to achieve an estimate of the taxonomic composition of the soil. Fungal communities of the Yarraman soil samples under the natural forest (YNF) and the hoop pine plantations (YHP) were investigated using 18S rDNA based cloning and sequencing approaches. Twenty-eight clone sequences were obtained and analysed. Three fungal orders, i.e., Zygomycota, Ascomycota and Basidiomycota were detected from the YNF and YHP samples. By contrast, culture-based analyses of fungi in the literature were mostly Ascomycetes. YNF appeared to have more Ascomycota but less Zygomycota than YHP, and within the Zygomycota order, YHP had more unidentified species than YNF. Bacterial communities of Yarraman soil samples of YNF, Y1R and Y2R were investigated using 16S rDNA-based cloning and sequencing approaches. 305 16S rDNA clone sequences were analysed and showed an overall bacterial community composition of Unclassified bacteria (34.4%), Proteobacteria (22.0%), Verrucomicrobia (15.7%), Acidobacteria (10.2%), Chloroflexi (6.9%), Gemmatimonadetes (5.6%), and Actinobacteria (5.2%). There was a significant difference among YNF, Y1R and Y2R in the taxonomic group composition. YNF had a greater proportion of Acidobacteria (18.0%), Verrucomicrobia (23.0%) and Chloroflexi (9.0%) than Y1R and Y2R (corresponding to 6.3%, 12.1% and 5.9%, respectively), while Y1R and Y2R had a higher percentage of the Unclassified group (38.5% for Y1R and 46.5% for Y2R) than YNF (18.0%). For the Proteobacteria group, YNF had more Alpha-subdivision but Y1R and Y2R had more Delta-subdivision. From YNF to Y1R to Y2R, the clone sequence variable site ratios, 5% and 10% OTU numbers and Shannon's diversity index H' values tended to decrease, indicating the soil bacterial diversity decreased from the natural forest to the first and the second rotation hoop pine plantations. The large amount of unclassified clone sequences could imply a novel group of bacteria in the soil, particularly in the hoop pine soil samples. Alternatively they may result from artefacts during the PCR process. Bacterial communities of the Gympie soil under different residue management practices, i.e., residue (litter plus logging residue) removed (G0R), residue retained (G1R), and residue doubled (G2R), were also investigated using the 16S rDNA-based cloning and sequencing approaches. Acidobacteria (37.6%) and Proteobacteria (35.6%, including Alpha-subdivision of 29.9% and Gamma-subdivision of 5.7%) were dominant components of the communities, followed by Actinobacteria (14.7%), Verrucomicrobia (7.3%) and Unclassified bacteria. There was no significant difference among G0R, G1R and G2R in the bacterial community compositions and diversity. These findings provided an in-depth vision of the soil microbial communities under different forest management practices. Their combination with other soil analysis results, such as physical and chemical properties, and forest production data, could provide an improved understanding of sustainable forest management strategies.

The aims of the work this thesis is based upon were to develop convenient analytical procedures for determining resin acids in biological and environmental matrices, and apply them to enhance understanding of the occurrence, exposure to and uptake by exposed individuals of resin acids. Particular focus has been on the workplace environment of the Swedish wood pellets industry. Sample extraction procedures and high-performance liquid chromatography/electrospray ionisation-mass spectrometry (HPLC/ESI-MS) methodologies were developed for measuring resin acids in dust, skin and urine samples. Chromatographic separation of abietic (AA) and pimaric acid was achieved by using a polar-embedded C12 stationary phase. The HPLC/ESI-MS method avoids undesirable oxidation of AA, which was found to occur during the derivatisation step in the standard MDHS 83/2 gas chromatography/flame ionisation detection (GC/FID) methodology, leading to false observations of both AA and the oxidation product 7-oxodehydroabietic acid (7-OXO). Personal exposures to resin acids in the Swedish wood pellet production industry were found to be lower, on average, than the British Occupational Exposure Limit for rosin (50 µg/m3). The oxidised resin acid 7-OXO, was detected in both dust and skin samples indicating the presence of allergenic resin acids. A correlation between air and post-shift urinary concentrations of dehydroabietic acid (DHAA), and a trend towards an increase in urinary 7-OXO during work shifts, were also observed. Whether the increase in 7-OXO was due to direct uptake or metabolism of other resin acids cannot be concluded from the results. An efficient HPLC/UV methodology with diode-array detection was developed for screening commercial products for rosin that could be used in laboratories lacking mass spectrometers. Very high concentrations of free resin acids were detected in depilatory wax strips using the method. / At the time of doctoral defense, the following paper was unpublished and had a status as follows: Paper 4: Submitted.

Boron (B) is a rare element on Earth crust with a natural abundance of 0.001%. However, boron content of water and soils may be significantly high in the regions with rich boron reserves. In addition, extensive use of agrochemicals in soils as well as various natural processes increases the boron concentration in water. Despite B is an essential element for all living creatures, it may pose risks at high level exposures. World Health Organization (WHO) has recommended a daily intake of 1 to 13 mg B for adults. Turkey has almost 70% of world boron reserves principally in four regions: K&uuml / tahya, Emet / Balikesir, Bigadi&ccedil / Eskisehir, Kirka and Bursa, Kemalpasa. The boron content of water in these regions may go up to significant levels. Therefore, it is important to determine B in drinking water from these regions. Electrothermal atomic absorption spectrometry (ETAAS) is a relatively sensitive technique for determination of boron. However, the technique suffers from formation of molecular boron compounds. Therefore, use of chemical modifiers and pyrolytically coated graphite tubes modified with refractory carbide forming elements (Ta, W, Zr, Pd, Ru, Os) were utilized to develop a reliable and sensitive method. Based on optimization studies, Tantalum (Ta) coated tube and co-injection of 5.0 &micro / L 0.01 mol/L Ca(NO3)2, 5.0 &micro / L 0.05 mol/L citric acid together with 15.0 &micro / L sample solution prepared in 1000 mg/L Mg(NO3)2 have been chosen as optimum conditions. Optimum temperatures for pyrolysis and atomization temperatures were determined as 1100 and 2700 &deg / C, respectively. Under these conditions, a detection limit of 0.088 mg/L and a characteristic mass of 186 pg for 15.0 &micro / L sample volume were obtained. The accuracy of the method was checked by EnviroMAT-Waste Water EU-L-1 CRM and NIST 1573a Tomato Leaves SRM analyses. Drinking water samples were collected from Balikesir, Bigadi&ccedil / and K&uuml / tahya, Emet and analyzed by the developed method. Samples were also analyzed by more sensitive techniques / ICP-OES and ICP-MS for a comparison study. The results are compatible with each other.

With the advent of high-throughput genomic and proteomic technologies, in conjunction with the difficulty in obtaining even moderately sized samples, small-sample classifier design has become a major issue in the biological and medical communities. Training-data error estimation becomes mandatory, yet none of the popular error estimation techniques have been rigorously designed via statistical inference or optimization. In this investigation, we place classifier error estimation in a framework of minimum mean-square error (MMSE) signal estimation in the presence of uncertainty, where uncertainty is relative to a prior over a family of distributions. This results in a Bayesian approach to error estimation that is optimal and unbiased relative to the model. The prior addresses a trade-off between estimator robustness (modeling assumptions) and accuracy. Closed-form representations for Bayesian error estimators are provided for two important models: discrete classification with Dirichlet priors (the discrete model) and linear classification of Gaussian distributions with fixed, scaled identity or arbitrary covariances and conjugate priors (the Gaussian model). We examine robustness to false modeling assumptions and demonstrate that Bayesian error estimators perform especially well for moderate true errors. The Bayesian modeling framework facilitates both optimization and analysis. It naturally gives rise to a practical expected measure of performance for arbitrary error estimators: the sample-conditioned mean-square error (MSE). Closed-form expressions are provided for both Bayesian models. We examine the consistency of Bayesian error estimation and illustrate a salient application in censored sampling, where sample points are collected one at a time until the conditional MSE reaches a stopping criterion. We address practical considerations for gene-expression microarray data, including the suitability of the Gaussian model, a methodology for calibrating normal-inverse-Wishart priors from unused data, and an approximation method for non-linear classification. We observe superior performance on synthetic high-dimensional data and real data, especially for moderate to high expected true errors and small feature sizes. Finally, arbitrary error estimators may be optimally calibrated assuming a fixed Bayesian model, sample size, classification rule, and error estimation rule. Using a calibration function mapping error estimates to their optimally calibrated values off-line, error estimates may be calibrated on the fly whenever the assumptions apply.

In this work, a practical approach to assessing bias and uncertainty using patient samples in a clinical laboratory is presented. The scheme is essentially a splitsample setup where one instrument is appointed to being the “master” instrument which other instruments are compared to. The software presented automatically collects test results from a Laboratory Information System in production and couples together the results of pairwise measurements. Partitioning of measurement results by user-defined criteria and how this can facilitate isolation of variation sources are also discussed. The logic and essential data model are described and the surrounding workflows outlined. The described software and workflow are currently in considerable practical use in several Swedish large-scale distributed laboratory organizations. With the appropriate IT-support, split-sample testing can be a powerful complement to external quality assurance.

This essay studies the written performance of 43 Swedish junior high school students. Relative clauses, prepositional usage and subject-verb agreement are studied and analysed in order to see what and how many errors the students make and then finally to see if there is any difference in the performance of boys and girls. Previous research in the area has shown an advantage in favour of girls and this study confirmed this. Even though the differences were not marked, the girls performed better than the boys in the majority of the cases studied. The data further indicated that there is great variation within the gender groups as well as between them.