Downloads & Free Reading Options - Results

PF

We study the spectra of MANOVA estimators and other quadratic estimators for variance component covariance matrices in multivariate random effects models. When the dimensionality of the observations is large and comparable to both the number of samples and the number of realizations of each random effect, the empirical eigenvalue distributions of such estimators are well-approximated by deterministic laws. The Stieltjes transforms of these laws may be characterized by systems of fixed-point equations, which are numerically solvable by a simple iterative procedure. Our derivation of this characterization uses operator-valued free probability theory, and we establish a general asymptotic freeness result for families of rectangular orthogonally-invariant random matrices, which is of independent interest. Our work is motivated by the estimation of components of covariance between multiple phenotypic traits in quantitative genetics, and we specialize our results to common experimental designs that arise in this application.

We study the spectra of MANOVA estimators and other quadratic estimators for variance component covariance matrices in multivariate random effects models. When the dimensionality of the observations is large and comparable to both the number of samples and the number of realizations of each random effect, the empirical eigenvalue distributions of such estimators are well-approximated by deterministic laws. The Stieltjes transforms of these laws may be characterized by systems of fixed-point equations, which are numerically solvable by a simple iterative procedure. Our derivation of this characterization uses operator-valued free probability theory, and we establish a general asymptotic freeness result for families of rectangular orthogonally-invariant random matrices, which is of independent interest. Our work is motivated by the estimation of components of covariance between multiple phenotypic traits in quantitative genetics, and we specialize our results to common experimental designs that arise in this application.

Philip Morris Records; publication

Philip Morris Records; publication

In the paper, the problem of the existence of the maximum likelihood estimate and the REML estimate in the variance components model is considered. Errors in the proof of Theorem 3.1 in the article of Demidenko and Massam (Sankhy\=a 61, 1999), giving a necessary and sufficient condition for the existence of the maximum likelihood estimate in this model, are pointed out and corrected. A new proof of Theorem 3.4 in the Demidenko and Massam's article, concerning the existence of the REML estimate of variance components, is presented.

Traditional methods for covariate adjustment of treatment means in designed experiments are inherently conditional on the observed covariate values. In order to develop a coherent general methodology for analysis of covariance, we propose a multivariate variance components model for the joint distribution of the response and covariates. It is shown that, if the design is orthogonal with respect to (random) blocking factors, then appropriate adjustments to treatment means can be made using the univariate variance components model obtained by conditioning on the observed covariate values. However, it is revealed that some widely used models are incorrectly specified, leading to biased estimates and incorrect standard errors. The approach clarifies some issues that have been the source of ongoing confusion in the statistics literature.

This article is from Ecology and Evolution , volume 4 . Abstract A trait must genetically correlate with fitness in order to evolve in response to natural selection, but theory suggests that strong directional selection should erode additive genetic variance in fitness and limit future evolutionary potential. Balancing selection has been proposed as a mechanism that could maintain genetic variance if fitness components trade off with one another and has been invoked to account for empirical observations of higher levels of additive genetic variance in fitness components than would be expected from mutation–selection balance. Here, we used a long-term study of an individually marked population of North American red squirrels (Tamiasciurus hudsonicus) to look for evidence of (1) additive genetic variance in lifetime reproductive success and (2) fitness trade-offs between fitness components, such as male and female fitness or fitness in high- and low-resource environments. “Animal model” analyses of a multigenerational pedigree revealed modest maternal effects on fitness, but very low levels of additive genetic variance in lifetime reproductive success overall as well as fitness measures within each sex and environment. It therefore appears that there are very low levels of direct genetic variance in fitness and fitness components in red squirrels to facilitate contemporary adaptation in this population.

Two bootstrapping or resampling strategies were investigated to determine their applicability to estimating standard errors and ensuing confidence intervals on variance components in two-factor random analysis of variance models. In light of prior negative findings regarding the application of bootstrapping to this particular problem, a recommendation of an "optimal" approach to resampling was sought. The study used Monte Carlo simulations to test the variance component estimation accuracy under simultaneous resampling of all effect factors in a random model versus resampling a single factor. Results indicate that single-factor was a preferable method of resampling and produced reasonable estimates of both standard errors and confidence intervals (parametric and non-parametric). Suggestions for the appropriate application of the technique to educational measurement are discussed. (Author/TJH)

The P.E.P. Report 1969-1973 focuses on the various findings and activities of the Program Evaluation Project. The study in this chapter was designed to conduct a statistical analysis of the Goal Attainment Score, and estimate variance components due to choice of material in the followup guide, followup interviewer bias or error, and the client's actual long-term deviation from expectation. These factors together determine the reliability of the Goal Attainment score as it was applied in this Program Evaluation Project study, and, in addition, provide some useful indication of its potential reliability in other evaluative applications. (Author/RC)

Taking Albert's (1976) formulation of a mixed model Analysis of variance, we consider improved estimation of the variance components for balanced designs under squared error loss. Two approaches are presented. One extends the ideas of Stein (1964). The other is developed from the fact that variance components can be expressed as linear combinations of chi-scale parameters. Encouraging simulation results are presented. Keywords: Linear regression analysis, Chi square scale parameters.

This paper discusses variables influencing written composition quality and how they can best be controlled to improve the reliability of assessment of writing ability. The study aimed to concentrate on the effects of different imposed time limits on student performance in written composition at grade 10 level using methods of generalizability theory; however, as the study progresses the major focus becomes methodological issues. Research studies and literature on this subject are reviewed before the design of the study is described. The sample of 320 students from Iowa and Newfoundland schools who were divided into eight groups (each group being assigned eight judges, who would assess the students' performance on two written topics under different time constraints) is presented. Scoring procedures are outlined. Discussed are the analytic methods of both Model 1 (a seven way analysis of variance model) and its limitations, and Model 2, which was adopted to analyze the variance for a confounded model incorporating elements of topic, time limit, and group. The results and limitations of the study are documented in the concluding section. Data analysis statistics are appended. (AEF)

This paper discusses variables influencing written composition quality and how they can best be controlled to improve the reliability of assessment of writing ability. The study aimed to concentrate on the effects of different imposed time limits on student performance in written composition at grade 10 level using methods of generalizability theory; however, as the study progresses the major focus becomes methodological issues. Research studies and literature on this subject are reviewed before the design of the study is described. The sample of 320 students from Iowa and Newfoundland schools who were divided into eight groups (each group being assigned eight judges, who would assess the students' performance on two written topics under different time constraints) is presented. Scoring procedures are outlined. Discussed are the analytic methods of both Model 1 (a seven way analysis of variance model) and its limitations, and Model 2, which was adopted to analyze the variance for a confounded model incorporating elements of topic, time limit, and group. The results and limitations of the study are documented in the concluding section. Data analysis statistics are appended. (AEF)

In this project confidence intervals were determined for the functions of variance components. These included one-sided and two-sided confidence intervals. Since no method is available for exact (exact confidence coefficients) for confidence intervals on the functions of variance components considered, confidence intervals with approximate confidence coefficients were examined. In each case large simulation studies were conducted to determine how good the confidence intervals are. In cases where alternative methods are available, they were compared to the procedures in this project.

In this project confidence intervals were determined for the functions of variance components. These included one-sided and two-sided confidence intervals. Since no method is available for exact (exact confidence coefficients) for confidence intervals on the functions of variance components considered, confidence intervals with approximate confidence coefficients were examined. In each case large simulation studies were conducted to determine how good the confidence intervals are. In cases where alternative methods are available, they were compared to the procedures in this project.

Consider the two-factor nested components of variance model Y sub ijk = mu + A sub i + B sub ij + C sub ijk, where Var(A sub i) = sigma squared sub A Var(B sub ij) = sigma-squared sub B, Var(C sub ijk) = sigma-squared sub C. Confidence intervals are derived for sigma-squared sub A/sigma-squared sub C, sigma-squared sub A (sigma-squared sub A + sigma-squared sub C) and sigma-squared sub C/(sigma-squared sub A + sigma-squared sub C). (Author)

PhD dissertation, Ohio State University

Previous studies into airframe acquisition cost estimation do not explicitly recognize the existence of correlation in the historical data. If one believes this data problem exists, then it is possible to develop a components of variance model that takes the problem into account. It is a more general model that recognizes two sources of error: (1) error due to different types of airframes and (2) overall or ordinary regression error. The variance of these two errors can be estimated and then can be utilized along with the technique of generalized least squares to obtain a cost estimating relationship which explicitly accounts for the data correlation. This modeling technique, when compared to techniques presently in service, shows that present estimating relationships underestimate the variance of the cost prediction of a new type airframe and overestimate the variance of the cost prediction of a follow-on airframe. Also, those existing techniques which implicitly recognize data correlation do not make use of all the data information available and therefore produce estimates with poor confidence prediction intervals.

Generalizability theory is based upon analysis of variance (ANOVA) and requires estimation of variance components for the ANOVA design under consideration in order to compute either G (Generalizability) or D (Decision) coefficients. Estimation of variance components has a number of alternative methods available using SAS, BMDP, and ad hoc procedures. This study examines the effect of these methods on coefficeints for a large sample of subjects under complex designs. The Wechsler Adult Intelligence Scale-Revised national standardization sample provided the data base. Subject (n=1880) variables (age, gender, etc.) provided variety for the construction of complex factoral designs. Three designs were used to investigate variance component estimation and resultant g-coefficeints. The designs included: (1) a two-factor subject x item design; (2) a three-factor subject x item x age design; and (3) a four-factor subject x item x age x gender design. These designs allowed estimation of both within and between subject effects. Suggestions are made concerning equal and unequal situations. (Author/PN)

Click here to view the University of Florida catalog record

This paper documents a check standard wind tunnel test conducted in the Langley 0.3-Meter Transonic Cryogenic Tunnel (0.3M TCT) that was designed and analyzed using the Modern Design of Experiments (MDOE). The test designed to partition the unexplained variance of typical wind tunnel data samples into two constituent components, one attributable to ordinary random error, and one attributable to systematic error induced by covariate effects. Covariate effects in wind tunnel testing are discussed, with examples. The impact of systematic (non-random) unexplained variance on the statistical independence of sequential measurements is reviewed. The corresponding correlation among experimental errors is discussed, as is the impact of such correlation on experimental results generally. The specific experiment documented herein was organized as a formal test for the presence of unexplained variance in representative samples of wind tunnel data, in order to quantify the frequency with which such systematic error was detected, and its magnitude relative to ordinary random error. Levels of systematic and random error reported here are representative of those quantified in other facilities, as cited in the references.

This article is from Genetics, Selection, Evolution : GSE , volume 43 . Abstract Background: The pre-weaning growth rate of lambs, an important component of meat market production, is affected by maternal and direct genetic effects. The French genetic evaluation model takes into account the number of lambs suckled by applying a multiplicative factor (1 for a lamb reared as a single, 0.7 for twin-reared lambs) to the maternal genetic effect, in addition to including the birth*rearing type combination as a fixed effect, which acts on the mean. However, little evidence has been provided to justify the use of this multiplicative model. The two main objectives of the present study were to determine, by comparing models of analysis, 1) whether pre-weaning growth is the same trait in single- and twin-reared lambs and 2) whether the multiplicative coefficient represents a good approach for taking this possible difference into account. Methods: Data on the pre-weaning growth rate, defined as the average daily gain from birth to 45 days of age on 29,612 Romane lambs born between 1987 and 2009 at the experimental farm of La Sapinière (INRA-France) were used to compare eight models that account for the number of lambs per dam reared in various ways. Models were compared using the Akaike information criteria. Results: The model that best fitted the data assumed that 1) direct (maternal) effects correspond to the same trait regardless of the number of lambs reared, 2) the permanent environmental effects and variances associated with the dam depend on the number of lambs reared and 3) the residual variance depends on the number of lambs reared. Even though this model fitted the data better than a model that included a multiplicative coefficient, little difference was found between EBV from the different models (the correlation between EBV varied from 0.979 to 0.999). Conclusions: Based on experimental data, the current genetic evaluation model can be improved to better take into account the number of lambs reared. Thus, it would be of interest to evaluate this model on field data and update the genetic evaluation model based on the results obtained.

We present arguments in favour of the inequalities $var(X_n^2|X \in B_v(\rho)) \le 2\lambda_n E[X_n^2|X \in B_v(\rho)]$, where $X \sim N_v(0,\Lambda)$ is a normal vector in $v\ge 1$ dimensions, with zero mean and covariance matrix $\Lambda = \diag(\lambda)$, and $B_v(\rho)$ is a centered $v$-dimensional Euclidean ball of square radius $\rho$. Such relations lie at the heart of an iterative algorithm, proposed in ref. [1] to perform a reconstruction of $\Lambda$ from the covariance matrix of $X$ conditioned to $B_v(\rho)$. In the regime of strong truncation, i.e. for $\rho \lesssim \lambda_n$, the above inequality is easily proved, whereas it becomes harder for $\rho \gg \lambda_n$. Here, we expand both sides in a function series controlled by powers of $\lambda_n/\rho$ and show that the coefficient functions of the series fulfill the inequality order by order if $\rho$ is sufficiently large. The intermediate region remains at present an open challenge.

tecnologie segrete

This study describes the development and initial psychometric evaluation of a Recognizing Effective Special Education Teachers (RESET) teacher observation instrument. Specifically, the study uses generalizability theory to compare two versions of a rubric, one with general descriptors of performance levels and one with item-specific descriptors of performance levels, to evaluate special education teacher implementation of explicit instruction. Eight raters participated in viewing and scoring videos of special education instruction. Data collected from raters were analyzed in a three facet, crossed, mixed-model design to estimate the variance components and reliability indices. Results show lower unwanted sources of variance and higher indices of reliability with the rubric with item-specific descriptors of performance levels. Contributions to the field of teacher evaluation are discussed. [This paper will be published in "Assessment for Effective Intervention" under the title "Variance and Reliability in Special Educator Observation Rubrics."]

This volume was digitized and made accessible online due to deterioration of the original print copy.

This paper provides a general methodology applicable to essentially any multistage survey in which the last stage units are drawn with equal probabilities. Specifically our formulas for the estimated variances of target parameter estimates will include all finite population corrections except those in the last stage which are usually negligible. Recent results in the estimation of components of variance in mixed linear models to achieve these results and are able to address the problem of estimability of variance components.

The paper describes a number of methods for estimating variance components in a general linear model. Explicit expressions are obtained for locally minimum variance unbiased estimators with and without the invariance condition. The principle of MINQE is described and a series of MIQE estimators satisfying one or more of the conditions - unbiasedness, invariance, nonnegative definiteness - are derived. Corresponding to each MINQE estimator, an iterated MINQE (IMINQE) is defined. It is shown that the ML (maximum Likelihood) estimator is IMINQE satisfying the invarianc e condition and the RML (restricted maximum likelihood estimator) is IMINQE satisfying both the invariance and unbiasedness conditions. Some comments are made on the numerical alorithms for computing MINQ, ML and RML estimators. (Author)

The purpose of this thesis was to examine the effects of applying the technique of control variates on the principal components of a given model. The investigation was done by comparing three sets of data as follows: (1) The set of principal components of the outputs of the model on which no variance reduction has been applied. (2) The set of principal components of variance-controlled outputs-control variates was performed prior to principal components analysis being done. (3) The set of variance-controlled principal components-control variates was performed on the principal components of the outputs. The comparison of the effects was carried out by examining the percentage of variance explained by the principal components and be reviewing the scatter plots of the first two principal components.

Introduction [1] : The Kermani sheep breed is one of the most important native breeds in Iran. It is well adapted to the harsh and undesirable environmental conditions of the southeastern part of the country. Therefore, consideration to the breeding of Kermani sheep and the improvement of the environmental and genetic conditions of this breed can contributed greatly to providing a part of the livelihood needs of the people. The longevity in the ewe is a high economic trait, because increasing the longevity will lead to a reduction in the culling rate as well as replacement costs in the herd and increasing of number of lambs. Considering that the investigation and identification of the important factors influence culling of ewes from the herd, which leads to a decrease in the yield and longevity of the animal, can be used to provide and develop a suitable strategy to promote the survival and longevity of the ewes. In addition, due to the absence of genetic parameters of production life length in Kermani sheep, the aim of this study was to study changes, prediction of components of variance and genetic parameters of survival distribution function in Kermani ewes. Materials and Methods: In this research, longevity records of 772 heads of Kermani ewes (collected from 1989 to 2008) at the Kermani sheep station (Shahr Babak, Kerman province) were studied. The overall production life length was equal to the time between births to ewe culling. Frequency distribution for causes of culling and survival distribution function for ewes were estimated using R software. The Weibull model and Matvec software were also used for estimating variance component and genetic parameters of longevity traits as survival. Results and Discussion: The results showed that the mean of ewe’s longevity was 4.72 years with average lambing of 3.4 in life time. The most important cause of culling was disease (53%), especially in the first three parity of the ewe. After disease, oldness, reproductive problems, low production, physical phenomenon, experimental slaughter, selling additional ewes and bad type (0.5%) are the accounted causes for remaining ewe culling, respectively. Considering that culling due to disease accounts for more than half of the deaths, prevention and treatment of diseases are important in this breed. Therefore, the type and timing of the prevalence should be investigated and preventive and therapeutic proceeding should be taken for diseases. The total amount of non-optional culling (illness, physical incidence, and aging) was about 83% of total casualties. It can be concluded that the deletion due to the low production was very small. At the age of two and three, the mortality rate was low, reaching its peak at the age of four, has declined since this age, with partial fluctuations between the ages of 5 and 6, and at the age of 7 and 8 years decline have happened. Therefore, most ewes were culled from the herd before reaching the age of 8. Most culling has occurred in pre-maturity ages, so it should be taken into consideration because it indicates that the ewes were eliminated before they reach the peak of production. The conditional culling probability increases with age, while cumulative survival rate decreased with age of ewe. The hazard ratio of death increased with age up to 6 years of old steadily and then increased strongly, and reached two at 7 to 8 years of old. The high hazard ratio indicated an early death or short longevity. Normally, inverse of this ratio expresses useful information. The reversal of this ratio at a certain time interval, provided that the hazard level remains constant at this interval, will indicate the time it takes to occur (death). For example, at a time interval of 3 to 4 years, the ewe's longevity will be equal to a reversal of 0.5298, equivalent to two years. But, for 7 to 8 years old, the reciprocal of two will be 0.5 years that indicating if the dangers remain constant at this interval, the ewe's life will be only 6 months, and ewes remaining in this group will be lost after 6 months. Trend of survival and hazard ratio at different ages of ewes can be provided as a model to breeders of this breed. Heritability on the logarithmic scale, the original scale and effective heritability of the risk ratio obtained from the sire model with the Weibull function was predicted at low to high (0.002-0.145). Although heritability estimates for longevity are low, but genetic improvement may be increased by selecting animals with higher inbreeding values. Conclusion: The results of this study showed that the most effective factor in shortening the longevity and decreasing survival in Kermani ewes is culling due to disease, especially in pre-maturity. Therefore, in order to improve the longevity, necessary studies and necessary preventive should be taken regarding the type and timing of the disease. Genetically, if possible, it would seek to create genetic resistance to diseases and environmental conditions. On the other hand, the results of models with the Weibull function indicate that in order to improve the survival rate and longevity, in addition to improving the non-genetic factors, the selection should be considered within the breed. Also, high heritability traits with high genetic correlation with longevity traits are identified and indirect selection be done based on them.  

A study on sampling errors of variance components was conducted within the framework of generalizability theory by P. L. Smith (1978). The study used an intuitive approach for solving the problem of how to allocate the number of conditions to different facets in order to produce the most stable estimate of the universe score variance. Optimization techniques are proposed as a promising approach for solving allocation problems in generalizability studies. An important consequence of the conclusion that allocation problems in generalizability studies have much in common with allocation problems in decision studies would be a considerable reduction in the effort required to develop optimization procedures and computer programs. This examination of optimization techniques is the beginning of a procedure that enables the researcher to design a generalizability study in line with stated limits on sampling errors as well as resources. (SLD)

This document shows that, by a reparametrization, the problem of estimating a linear combination of variance components can be reduced to that of estimating single variance component. Such a reduction is used to obtain some characterizations of nonnegatively estimable linear combinations of various components, Characterization of nonnegative estimability using MINQUE is also discussed. Additional keywords: quadratic subspace, QUE's (quadratic unbiased estimators).

We extend the results concerning the upper bounds for the maximum likelihood degree and the REML degree of the one-way random effects model presented in Gross et al. [Electron. J. Stat. 6 (2012), pp. 993-1016] to the case of the normal linear mixed model with two variance components. Then we prove that both parts of Conjecture 1 in the paper of Gross et al., which concerns a certain extension of the one-way random effects model, are true under fairly mild conditions.

Introduction Tomato is a self-pollinated crop and has a high potential for heterosis production. Tomato has a wide range of diversity in terms of vegetative and fruit traits. Therefore, learning information about the genetics of the tomato plant and the inheritance of its various traits to the next generation will help plant breeders to use appropriate breeding methods to improve them. One of the methods that is used to know the genetic structure of plants, identify parental lines and determine their combining ability is line × tester analysis. Line × tester analysis provides information about general and specific combining of parents and can be useful in estimating different types of gene effects such as additive and non-additive effects. In most of the developed countries, many researches have been done in relation to hybrid production and combining ability among tomato lines, and sometimes the inferred results are different from each other. In Iran, few studies have been done about crossing cultivars and their hybrids, and most of the seeds used by farmers are imported from other countries. Therefore, this study intends to evaluate genetic variance components, general and specific combining ability of some quantitative traits in a number of tomato lines and testers and their hybrids by using line × tester analysis.   Materials and Methods This research was conducted in Sari Agricultural Sciences and Natural Resources University, Mazandaran Province, Iran in 2022. Two modified cultivars SC and V as lines and three modified cultivars L, R and MZ as testers were crossed with each other to create F 1 hybrids. Six F 1 genotypes and their parents (11 treatments in total) were cultivated in the farm in a randomized complete block design with three replications. The evaluated traits included the number of days to the first flowering, earliness, number of fruits per plant, fruit weight per plant (g), fruit yield (g), fruit length and width (cm). In order to analyze the variance of the experimental design to search for diversity between treatments, to separate the effects of treatments into their components based on line × tester analysis, to mean comparison with Duncan's test, and also to calculate the general and specific combining ability, R statistical software was used. Also, in order to calculate additive and non-additive variances, Singh and Chaudhary's method was used.   Results and Discussion The results of line × tester variance analysis showed that the mean squares of parents and testers were significant for all traits except fruit length and width, and the mean squares of crosses and lines were significant for all traits except fruit length. The effect of line × tester was significant for all traits except the number of fruits per plant and fruit length. The line of SC to improve the number of days to first flowering, earliness, plant height, fruit weight per plant, and fruit width, and the line of V to improve the number of fruit per plant were the best general combiners with testers. The tester of L for improve all traits except yield, and the tester of MZ for improve plant height were the best general combiners with the maternal lines. Among the crosses, the SC×L cross for improve earliness and fruit width, and the SC×R and V×MZ crosses for improve plant height and fruit weight per plant, respectively, were favorable specific combiners. The mean comparison of the genotypes for some important traits showed that among the parental cultivars, the line of SC and among the crosses, the SC×L genotype had the lowest means for the number of days to first flowering and earliness. Also, the line of SC for the number of fruits per plant and the SC×L genotype for fruit weight per plant, yield and fruit width had the highest means. Also, the estimation of additive and non-additive variances indicated that in plant height and fruit weight per plant traits, additive variance plays the main role. While for the traits of the number of days to first flowering, earliness and yield, the contribution of non-additive variance was more than the additive variance.   Conclusion According to the results obtained from this study, in future projects it is recommended to use parents that have significant general combining ability (GCA) for traits. Because such parents easily transfer the trait to their next generation. In this way, the line of SC was a good general combiner for the number of days to first flowering, earliness, plant height, fruit weight per plant and fruit width, and the line of V was a good general combiner for the number of fruits per plant. Among the testers, the tester of L was a good general combiner for improve the number of days to first flowering, earliness, number of fruits per plant, fruit weight per plant, and fruit width, and the tester of MZ recorded a high GCA for the plant height. Also, for the improvement of earliness and fruit width, the SC×L cross and for plant height and fruit weight per plant, SC×R and V×MZ crosses were favorable specific combiner. Mean comparison of genotypes showed that the SC×L cross is superior to its parents for the number of days to first flowering, earliness, fruit weight per plant, fruit yield, and fruit length and width. The traits of plant height and fruit weight per plant are more affected by additive variance, so the best breeding method to improve plant height and fruit weight per plant is selection from among the segregating population. The traits of number of days to first flowering, earliness and yield were affected by non-additive variance, so hybrid production is recommended to improve the mentioned traits.

This text is a reprint of the first English-language book on rhetoric, including an introduction and notes from Frederic Ives Carpenter of the University of Chicago. Cox's original brief treatise covers various themes and topics related to oratory and argument. Carpenter's introductory material adds context and analysis. (Summary by Amelia Chesley)

2Brownies and Prince Florimel

By Palmer Cox

After Prince Florimel flees his home (and a particularly nasty uncle) he has a series of adventures with the mischevious but helpful Brownies, Queen Titania and her fairies, and the ill-tempered enchanter, Dragonfel.<p> Reader's note: Although some terms in this 1918 book are not really acceptable anymore, the book’s overall message that people should not be judged by their size is still a worthy lesson today. - Summary by Jude Somers

The Crusades were a series of religious wars fought between 1096 and 1272 to recover the Holy Land from Islamic rule. According to the Latin Church, Crusaders were penitent pilgrims whose sins were forgiven. British historian, George Cox, writes of the churchmen, great and small, who inspired the Crusades, of the warriors who left families and lands behind, of the wily Venetian merchants and Byzantine emperors who exploited the knights, and of the valor of the Saracens. Here are accounts of sublime sacrifice and bestial ferocity, of dynastic conflict within the Crusader States, of sieges, starvation, pestilence, and ambush, and of the clash and interpenetration of two cultures. - Summary by Pamela Nagami, M.D.

The Brownies are delightful, helpful, mischievous characters and this series of rollicking rhymes tells many stories about their adventures. - Summary by Jude Somers

5Another Brownie Book

By Palmer Cox

Brownies, like fairies and goblins, are imaginary little sprites, who are supposed to delight in harmless pranks and helpful deeds. They work and sport while weary households sleep, and never allow themselves to be seen by mortal eyes. Summary by Palmer Cox

6Frontier Humor in Verse, Prose and Picture

By Palmer Cox

Also known for his "Brownies" books, Canadian humorist Palmer Cox give us a delightful collection of humorous verse and short prose vignettes. From the publisher's preface, "thrice happy is the man who, having seen, can tell the fun; and having told, can picture it for others’ eyes and so roll on the rollicking humor, for the brightening of a world already far too sad." - Summary by Larry Wilson

A limbless reptilian monster that propels itself through the swamps with a propellor-tipped tail? A creature so ashamed of its monstrous appearance that it dissolves into tears when captured? Learn about the snoligoster, the squonk and many other 'fearsome critters' in this field guide written and illustrated by two North American foresters who know them well. Listeners who suspect that these creatures are the stuff of tall tales, will nevertheless do well to look out for slide-rock bolters when vacationing in the Colorado mountains. - Summary by Phil Benson

The British historian, George W. Cox writes that the "whole duration of the Athenian empire extends over little more than two generations." This confederation of Greek city states with Athens at its head was an attempt to present a unified front to the Persian Empire. But ultimately, the strife between democratic and oligarchic factions within each polis, combined with the tangle of alliances between the various city states, led to the ruinous and protracted Peloponnesian War. In this short work, Cox describes the battles on land and sea, and such personalities as the Athenian general, Nikias, his Spartan counterpart, Agis, and, above all, unprincipled, charismatic Alcibiades. (Summary by Pamela Nagami, M.D.)

In the darkest days of the Persian War when the armies of Xerxes were overrunning northern Greece, Athens faced destruction. The desperate Athenians consulted the oracle at Delphi, who answered, "For thus saith Zeus, that when all else within the land of Cecrops is wasted, the wooden wall alone shall not be taken." The British historian, George Cox agrees with many other authorities that the Greek statesman, Themistocles, bribed the oracle. Realizing that their survival depended upon a strong navy, he declared that the prophesy meant that the Persians would be defeated, not by fortifying Athens, but by "the fight at sea, for the fleet is your wooden wall." This book is a short history of the founding fathers of Athens, its lawgivers, tyrants, and generals: Solon, Peisistratos, Kleisthenes, Miltiades, Aristeides, and, above all, Themistocles. (Summary by Pamela Nagami, M.D.)

The object of this little book is to bring the reader nearer to the great masters in the realm of music, and make him or her acquainted with the man who worked, played, and suffered as do other men. Though confining themselves to a simple style, the authors have tried to enter into the life of each composer, and, wherever possible, by means of earnest study of letters and other data which have come down to us, to copy that composer's tricks of speech, manner of writing, and various idiosyncracies. - Summary by Preface

Rip Austin knew that curvaceous Sally Vaughn had been murdered for her insurance money, but how could he prove it when the local vice lords gave orders to the town council, police force and newspapers - and when he, himself, was neatly framed for two other murders? <br><br> Sex was the only coin that counted, and violence the only language spoken, until Rip Austin managed to blow the lid off! - Summary by Back cover, Art Enterprises, Inc. first edition