Conditional Dependence Model Research Articles

How the Black Swan damages the harvest: Extreme weather events and the fragility of agriculture in development countries.

Climate change constitutes a rising challenge to the agricultural base of developing countries. Most of the literature has focused on the impact of changes in the means of weather variables on mean changes in production and has found very little impact of weather upon agricultural production. Instead, we focus on the relationship between extreme events in weather and extreme losses in crop production. Indeed, extreme events are of the greatest interest for scholars and policy makers only when they carry extraordinary negative effects. We build on this idea and for the first time, we adopt a conditional dependence model for multivariate extreme values to understand the impact of extreme weather on agricultural production. Specifically, we look at the probability that an extreme event drastically reduces the harvest of any of the major crops. This analysis, which is run on data for six different crops and four different weather variables in a vast array of countries in Africa, Asia and Latin America, shows that extremes in weather and yield losses of major staples are associated events. We find a high heterogeneity across both countries and crops and we are able to predict per country and per crop the risk of a yield reduction above 90% when extreme events occur. As policy implication, we can thus assess which major crop in each country is less resilient to climate shocks.

Evaluation of the sensitivity and specificity of the lateral flow assay, Rose Bengal test and the complement fixation test for the diagnosis of brucellosis in cattle using Bayesian latent class analysis

This study was conducted to evaluate the sensitivity (Se) and specificity (Sp) of the Rose Bengal test (RBT), complement fixation test (CFT), the serum lateral flow assay (LFAserum) and the blood lateral flow assay (LFAblood) for the detection of antibodies to Brucella spp. using Bayesian latent class models (BLCMs). Sera and whole blood were collected from naturally infected cattle reared in smallholder, small-scale commercial and large-scale commercial farms in Zimbabwe (n = 1022) and Botswana (n = 770). The BLCMs were fitted under the assumption that conditional dependences existed between the tests. Based on the conditional dependence model, the RBT had the highest Se of 0.897 (95 % Probability Intervals: 0.854; 0.932) compared to 0.827 (0.773; 0.872), 0.812 (0.76; 0.858) and 0.809 (0.785; 0.832) for the LFAserum, LFAblood and CFT, respectively. The CFT recorded a higher Sp of 0.999 (0.995; 1.000) than the LFAserum 0.996 (0.99; 1.000), the LFAblood 0.984 (0.976; 0.991) and the RBT 0.969 (0.959; 0.978). The data indicated that both the Se and Sp of RBT and CFT and the Sp of LFAserum and LFAblood were conditionally independent, while the Se appeared to be conditionally dependent. These results indicated that none of the evaluated tests had perfect Se and Sp and consequently could not be used alone for the diagnosis of brucellosis in cattle from the studied farming sectors. Thus, based on high Se and Sp, respectively, a brucellosis testing regimen using the RBT (screening) and the LFA (confirmatory) may be considered.

Single and double Wiebe function combustion model for a heavy-duty diesel engine retrofitted to natural-gas spark-ignition

The conversion of existing diesel engines to natural-gas spark-ignition operation would reduce the dependence on oil imports, reduce the burden on refining capacity, and increase U.S. energy security. The use of simple combustion models such as the Wiebe function to analyze or predict the combustion process in such retrofitted engines can accelerate and optimize the engine conversion. This study compared the standard single Wiebe function to a double-Wiebe function to investigate if the latter will improve the predicted mass fraction burned and, if yes, which of formats of the duel-Wiebe function in the literature described the best the mass fraction burned in such converted engine. The results showed that while the standard Wiebe function could not predict the mass fraction, a double-Wiebe function (one for the fast burn inside the piston bowl and a second for the slower burning process inside the squish region) predicted with good accuracy the mass fraction burned, heat release rate, in-cylinder pressure, and combustion phasing in such converted engines (∼90% reduction in error) for the conditions investigated here, especially if the start of the second Wiebe was delayed. Moreover, such a condition-dependent model as the double-Wiebe model is limited to the operating conditions used for determining the model parameters. However, the model can be developed into a condition-independent model if enough experimental data or CFD simulations are available to find the unique set (or sets) of parameters that minimizes the error at most conditions.

Modeling Nonlinear Conditional Dependence Between Response Time and Accuracy.

The most common process variable available for analysis due to tests presented in a computerized form is response time. Psychometric models have been developed for joint modeling of response accuracy and response time in which response time is an additional source of information about ability and about the underlying response processes. While traditional models assume conditional independence between response time and accuracy given ability and speed latent variables (van der Linden, 2007), recently multiple studies (De Boeck and Partchev, 2012; Meng et al., 2015; Bolsinova et al., 2017a,b) have shown that violations of conditional independence are not rare and that there is more to learn from the conditional dependence between response time and accuracy. When it comes to conditional dependence between time and accuracy, authors typically focus on positive conditional dependence (i.e., relatively slow responses are more often correct) and negative conditional dependence (i.e., relatively fast responses are more often correct), which implies monotone conditional dependence. Moreover, most existing models specify the relationship to be linear. However, this assumption of monotone and linear conditional dependence does not necessarily hold in practice, and assuming linearity might distort the conclusions about the relationship between time and accuracy. In this paper we develop methods for exploring nonlinear conditional dependence between response time and accuracy. Three different approaches are proposed: (1) A joint model for quadratic conditional dependence is developed as an extension of the response moderation models for time and accuracy (Bolsinova et al., 2017b); (2) A joint model for multiple-category conditional dependence is developed as an extension of the fast-slow model of Partchev and De Boeck (2012); (3) An indicator-level nonparametric moderation method (Bolsinova and Molenaar, in press) is used with residual log-response time as a predictor for the item intercept and item slope. Furthermore, we propose using nonparametric moderation to evaluate the viability of the assumption of linearity of conditional dependence by performing posterior predictive checks for the linear conditional dependence model. The developed methods are illustrated using data from an educational test in which, for the majority of the items, conditional dependence is shown to be nonlinear.

Determinants of multiple paternity in a fluctuating population of ground squirrels

Multiple paternity is common in vertebrates that produce several offspring in the same reproductive bout, but the rate often varies among and within populations. Three primary explanations for this variation have been advanced: null models based on encounter rate of mates, socioecological models dependent on the ability of males to monopolize females, and age- or condition-dependent models of female choice. We used 18 years of genetic and demographic data to examine the mating system and patterns of multiple paternity in a free-living population of golden-mantled ground squirrels (Callospermophilus lateralis). The mating system was polygynandrous, but opportunity for sexual selection was lower for females than for males. Annual reproductive success of males was low for yearlings and new immigrants and increased with breeding tenure in the population. Multiple paternity was evident in 62% of litters. In accordance with the socioecological model of male monopolization, rates of multiple paternity decreased with female spatial clustering, unless male–male competition, as indicated by male density, was also high. From Bateman gradients, we found no direct fitness benefit of multiple paternity for females. Though not statistically significant, multiple paternity appeared to decrease with maternal age and peri-oestrous mass, in possible support of the female choice model. Together, our results suggest that variation in the rate of multiple paternity in golden-mantled ground squirrels was determined by density and the active strategies of males and females. Since the advent of molecular parentage assignment several decades ago, we have known that females of many species produce offspring with different fathers. Several theories have been developed for why females produce multiply-sired clutches or litters, but rarely are we able to identify the environmental, social, or individual conditions under which they do so. In this study, we genotyped offspring produced in one population of ground squirrels over 18 years, and found that the frequency of multiple paternity varied considerably from year to year, that density of female kin interacted with male density to influence multiple paternity, and that older and heavier females tended to be less likely to produce multiply-sired litters. These results demonstrate how dynamic population and individual characteristics of breeding males and females contribute to mating system variation in the same population over time.

A multi‐scale framework for flood risk analysis at spatially distributed locations

AbstractThis paper presents a multi‐scale framework for flood risk analysis from fluvial and coastal sources at broad (including national) scales. The framework combines an extreme value spatial model of fluvial and coastal flood hazards using the Heffernan and Tawn conditional dependence model, with a new Markov approach to representing the spatial variability of flood defences. The nested multi‐scale structure enables spatial and temporal dependence at a national scale to be combined with detailed local analysis of inundation and damage. By explicitly considering each stage of the process, potential uncertainties in the risk estimate are identified and can be communicated to end users to encourage informed decision making. The framework is demonstrated by application to an insurance portfolio of static caravan sites across the UK worth over £2bn. In the case study, the largest uncertainties are shown to derive from the spatial structure used in the statistical model and limited data on flood defences and receptor vulnerability.

Estimation of the sensitivity and specificity of a Leptospira spp. in-house ELISA through Bayesian modelling

The microscopic agglutination test (MAT) is still considered the gold standard for the diagnosis of leptospirosis, although studies have shown that the test is an imperfect gold standard for clinical samples and unsuitable for epidemiological studies. Here, test characteristics of an in-house ELISA were identified for both subclinical and clinical populations by Bayesian latent class models. A conditional dependence model for two diagnostic tests and two populations was adapted to analyse a clinical and a subclinical scenario, respectively. These Bayesian models were used to estimate the sensitivity and specificity of the in-house ELISA and the MAT as well as the prevalences. The Bayesian estimates of the in-house ELISA were: clinical sensitivity=83.0%, clinical specificity=98.5%, subclinical sensitivity=85.7% and subclinical specificity=99.1%. In contrast, the estimates of the MAT were: clinical sensitivity=65.6%, clinical specificity=97.7%, subclinical sensitivity 54.9% and subclinical specificity=97.3%. The results show the suitability of the in-house ELISA for both clinical investigations and epidemiological studies in mildly endemic areas.

Evolved individual differences: Advancing a condition-dependent model of personality

The field of personality psychology offers a wealth of robust empirical research and a successful descriptive taxonomy, but neither explains the origins of the structure of human personality nor elaborates a generative framework for predicting the specific conditions that evoke the development of distinct personality traits. Exploration of traditional personality constructs within an evolutionary adaptive individual differences framework may help fill this explanatory gap. Personality traits exhibit functional features and patterns of variation expected from psychological adaptations designed to solve survival- and reproduction-related challenges recurrently faced during our species’ evolutionary history. Condition-dependent evolutionary models of personality have been proposed for decades, but only recently have begun to see empirical investigation. These models posit that species-typical psychological mechanisms take as input cues from the individual’s phenotype that would have been ancestrally linked to differential cost–benefit tradeoffs of alternative personality strategies, and produce as output personality trait levels with the greatest probabilistic net benefit for the individual. This paper elaborates a more nuanced conceptual framework that builds on earlier conceptualizations of condition-dependent traits to yield new and untested hypotheses about personality trait variation and covariation. It then describes clear future research directions for empirically investigating these readily testable hypotheses.

A practical approach for incorporating dependence among fields in probabilistic record linkage

BackgroundMethods for linking real-world healthcare data often use a latent class model, where the latent, or unknown, class is the true match status of candidate record-pairs. This commonly used model assumes that agreement patterns among multiple fields within a latent class are independent. When this assumption is violated, various approaches, including the most commonly proposed loglinear models, have been suggested to account for conditional dependence.MethodsWe present a step-by-step guide to identify important dependencies between fields through a correlation residual plot and demonstrate how they can be incorporated into loglinear models for record linkage. This method is applied to healthcare data from the patient registry for a large county health department.ResultsOur method could be readily implemented using standard software (with code supplied) to produce an overall better model fit as measured by BIC and deviance. Finding the most parsimonious model is known to reduce bias in parameter estimates.ConclusionsThis novel approach identifies and accommodates conditional dependence in the context of record linkage. The conditional dependence model is recommended for routine use due to its flexibility for incorporating conditional dependence and easy implementation using existing software.

On the Tuning Parameter for the Adaptive Bonferroni Procedure under Positive Dependence

Guo introduced an adaptive Bonferroni procedure and he proved that his adaptive Bonferroni procedure controls the familywise error rate under a conditional dependence model. However, how to choose the tuning parameter λ to control the familywise error rate in the procedure under positive dependence is not clear in his paper. In this paper, we suggest that λ = α . Simulation studies are provided.

Individual optimization of reproduction in a long-lived migratory bird: a test of the condition-dependent model of laying date and clutch size

Summary 1. Optimality theory predicts that both timing of arrival and arrival state on the breeding area will determine reproductive timing and investment in migratory organisms. We tested this idea using a condition-dependent individual optimization model (Ardea 68, 1980, 225 and The American Naturalist 143, 1994, 698) in common eider ducks through descriptive data, path analyses and experimental manipulation. 2. Our results support the causal pathways drawn from the optimization model indicating that individuals adjust their reproductive decisions as a function of their arrival date and body condition at arrival. 3. Independent of body condition, early-arriving females had a longer pre-laying period, but still initiated their nests earlier, and produced larger clutches than late-arriving birds. Independent of arrival date, females in good condition laid earlier than those in poor condition. Manipulation of pre-laying female body condition confirmed that the relationship between condition and laying date was causal. 4. Female common eiders appear to optimize reproductive decisions in response to both their external (i.e. environmental conditions affecting the egg-value) and internal (i.e. body condition) states. These adjustments seem to minimize the fitness costs of reproduction, in which higher clutch size is not associated with an apparent lower survival or future breeding probability. 5. Our study emphasizes the importance of (i) simultaneously considering the timing of migration, the state of individuals and the seasonal change in egg-value to understand clearly birds’ breeding decisions and (ii) appreciating the potential proximate and ultimate factors explaining why some individuals delay breeding and ⁄ or produce small clutches.

Interactions between lay date, clutch size, and postlaying energetic needs in a capital breeder

The condition-dependent model of optimal clutch size assumes body reserves required to initiate egg production include those for subsequent breeding phases. The threshold is expected to be similar among individuals, and hence postlaying condition should be independent of clutch size and lay date. Alternatively, the cost of incubation hypothesis predicts that females laying larger clutches should secure extra resources for incubation, and the expected fitness hypothesis suggests females adjust their condition according to the anticipated fitness benefits of the clutch. In these 2 cases, postlaying condition is predicted to be positively related to clutch size. We tested these predictions in common eider (Somateria mollissima), a precocial bird that produce eggs mostly from stored lipids and partly from endogenous proteins and rely extensively on reserves to incubate. We collected females at incubation onset and measured body condition indicators. Clutch size (number of eggs laid or number found in the nest at incubation onset) was not related to postlaying fat stores but females that laid fewer eggs maintained extra protein reserves. Timing of breeding was not related to postlaying body mass or protein reserves, whereas lay date's relationship with fat stores varied annually. Our results are generally consistent with the condition-dependent model and indicate variation in postlaying condition is mostly driven by factors other than clutch size and lay date. These data are inconsistent with the cost of incubation and the expected fitness hypotheses and suggest body store differences at incubation onset are mostly caused by environmental conditions encountered by laying females. Copyright 2011, Oxford University Press.

A note on adaptive Bonferroni and Holm procedures under dependence

SUMMARY Hochberg & Benjamini (1990) first presented adaptive procedures for controlling familywise error rate. However, until now, it has not been proved that these procedures control the familywise error rate. We introduce a simplified version of Hochberg & Benjamini’s adaptive Bonferroni and Holm procedures. Assuming a conditional dependence model, we prove that the former procedure controls the familywise error rate in finite samples while the latter controls it approximately.

On false discovery control under dependence

A popular framework for false discovery control is the random effects model in which the null hypotheses are assumed to be independent. This paper generalizes the random effects model to a conditional dependence model which allows dependence between null hypotheses. The dependence can be useful to characterize the spatial structure of the null hypotheses. Asymptotic properties of false discovery proportions and numbers of rejected hypotheses are explored and a large-sample distributional theory is obtained.

Pleiotropy and the genomic location of sexually selected genes.

Sexual selection drives the evolution of traits involved in the competition for mates. Although considerable research has focused on the evolution of sexually selected traits, their underlying genetic architecture is poorly resolved. Here I address the pleiotropic effects and genomic locations of sexually selected genes. These two important characteristics can impose considerable constraints on evolvability and may influence our understanding of the process of sexual selection. Theoretical models are inconsistent regarding the genomic location of sexually selected genes. Models that do not incorporate pleiotropic effects often predict sex linkage. Conversely, sex linkage is not explicitly predicted by the condition-dependent model (which considers pleiotropic effects). Evidence largely based on reciprocal crosses supports the notion of sex linkage. However, although they infer genetic contribution, reciprocal crosses cannot identify the genes or their pleiotropic effects. By surveying the genome of Drosophila melanogaster, I provide evidence for the genomic location and pleiotropic effects of 63 putatively sexually selected genes. Interestingly, most are pleiotropic (73%), and they are not preferentially sex linked. Their pleiotropic effects include fertility, development, life span, and viability, which may contribute to condition and/or fitness. My findings may also provide evidence for the capture of genetic variation in condition via the pleiotropic effects of sexually selected genes.

Multicenter comparison of serologic assays and estimation of human herpesvirus 8 seroprevalence among US blood donors.

As part of assessing the possibility of transfusion transmission of human herpesvirus 8 (HHV-8 or Kaposi's sarcoma-associated herpesvirus), HHV-8 seroprevalence was estimated among US blood donors, the performance of HHV-8 serologic tests was compared, and the presence of HHV-8 DNA was tested for in donated blood. Replicate panels of 1040 plasma specimens prepared from 1000 US blood donors (collected in 1994 and 1995) and 21 Kaposi's sarcoma patients were tested for antibodies to HHV-8 in six laboratories. HHV-8 PCR was performed on blood samples from 138 donors, including all 33 who tested seropositive in at least two laboratories and 22 who tested positive in at least one. The estimated HHV-8 seroprevalence among US blood donors was 3.5 percent (95% CI, 1.2%-9.8%) by a conditional dependence latent-class model, 3.0 percent (95% CI, 2.0%-4.6%) by a conditional independence latent-class model, and 3.3 percent (95% CI, 2.3%-4.6%) by use of a consensus-derived gold standard (specimens positive in two or more laboratories); the conditional dependence model best fit the data. In this model, laboratory specificities ranged from 96.6 to 100 percent. Sensitivities ranged widely, but with overlapping 95 percent CIs. HHV-8 DNA was detected in blood from none of 138 donors evaluated. Medical and behavioral screening does not eliminate HHV-8-seropositive persons from the US blood donor pool, but no viral DNA was found in donor blood. Further studies of much larger numbers of seropositive individuals will be required to more completely assess the rate of viremia and possibility of HHV-8 transfusion transmission. Current data do not indicate a need to screen US blood donors for HHV-8.

Sensorless sliding-mode control of induction motors using operating condition dependent models

A sensorless torque control system for induction motors has been developed. The system allows for fast and precise torque tracking over a wide range of speed. This paper also presents the identification and parameter estimation of an induction motor model with parameters varying as functions of the operating conditions encountered in hybrid electric vehicles applications. An adaptive sliding-mode speed-flux observer is developed and a cascade of discrete time sliding-mode controllers is used for flux and current control. Simulation and experimental results prove the validity of the approach.

Mathematical models for coordinated interpersonal timing in mother-infant interactions in the first year of life.

Two mathematical models of social contingency are examined in terms of their development over the first year of life. The interactions of 53 mothers and their infants were recorded at 6 weeks, 4 months, and 12 months. The infants' gazes at 6 weeks, the mothers' vocal behavior at 6 weeks, and the vocal behavior of the mother and infant at 4 and 12 months were automatically coded in terms of four states. The conditional dependence model and the response effects model were computed for each interaction at each age, and the coefficients of the models were examined as a function of age. The relative success of the models as estimates of moment-to-moment contingency as well as their variations with age are discussed.