Estimating the Counterfactual Impact of Conservation Programs on Land Cover Outcomes: The Role of Matching and Panel Regression Techniques

Deforestation and conversion of native habitats continues to be the leading driver of biodiversity and ecosystem service loss. A number of conservation policies and programs are implemented—from protected areas to payments for ecosystem services (PES)—to deter these losses. Currently, empirical evidence on whether these approaches stop or slow land cover change is lacking, but there is increasing interest in conducting rigorous, counterfactual impact evaluations, especially for many new conservation approaches, such as PES and REDD, which emphasize additionality. In addition, several new, globally available and free high-resolution remote sensing datasets have increased the ease of carrying out an impact evaluation on land cover change outcomes. While the number of conservation evaluations utilizing ‘matching’ to construct a valid control group is increasing, the majority of these studies use simple differences in means or linear cross-sectional regression to estimate the impact of the conservation program using this matched sample, with relatively few utilizing fixed effects panel methods—an alternative estimation method that relies on temporal variation in the data. In this paper we compare the advantages and limitations of (1) matching to construct the control group combined with differences in means and cross-sectional regression, which control for observable forms of bias in program evaluation, to (2) fixed effects panel methods, which control for observable and time-invariant unobservable forms of bias, with and without matching to create the control group. We then use these four approaches to estimate forest cover outcomes for two conservation programs: a PES program in Northeastern Ecuador and strict protected areas in European Russia. In the Russia case we find statistically significant differences across estimators—due to the presence of unobservable bias—that lead to differences in conclusions about effectiveness. The Ecuador case illustrates that if time-invariant unobservables are not present, matching combined with differences in means or cross-sectional regression leads to similar estimates of program effectiveness as matching combined with fixed effects panel regression. These results highlight the importance of considering observable and unobservable forms of bias and the methodological assumptions across estimators when designing an impact evaluation of conservation programs.     

The bias due to incomplete matching

Observational studies comparing groups of treated and control units are often used to estimate the effects caused by treatments. Matching is a method for sampling a large reservoir of potential controls to produce a control group of modest size that is ostensibly similar to the treated group. In practice, there is a trade-off between the desires to find matches for all treated units and to obtain matched treated-control pairs that are extremely similar to each other. We derive expressions for the bias in the average matched pair difference due to the failure to match all treated units--incomplete matching, and the failure to obtain exact matches--inexact matching. A practical example shows that the bias due to incomplete matching can be severe, and moreover, can be avoided entirely by using an appropriate multivariate nearest available matching algorithm, which, in the example, leaves only a small residual bias due to inexact matching.     

Directional penalties for optimal matching in observational studies

Multivariate matching in observational studies tends to view covariate differences symmetrically: a difference in age of 10 years is thought equally problematic whether the treated subject is older or younger than the matched control. If matching is correcting an imbalance in age, such that treated subjects are typically older than controls, then the situation in need of correction is asymmetric: a matched pair with a difference in age of 10 years is much more likely to have an older treated subject and a younger control than the opposite. Correcting the bias may be easier if matching tries to avoid the typical case that creates the bias. We describe several easily used, asymmetric, directional penalties and illustrate how they can improve covariate balance in a matched sample. The investigator starts with a matched sample built in a conventional way, then diagnoses residual covariate imbalances in need of reduction, and achieves the needed reduction by slightly altering the distance matrix with directional penalties, creating a new matched sample. Unlike penalties commonly used in matching, a directional penalty can go too far, reversing the direction of the bias rather than reducing the bias, so the magnitude of the directional penalty matters and may need adjustment. Our experience is that two or three adjustments, guided by balance diagnostics, can substantially improve covariate balance, perhaps requiring fifteen minutes effort sitting at the computer. We also explore the connection between directional penalties and a widely used technique in integer programming, namely Lagrangian relaxation of problematic linear side constraints in a minimum cost flow problem. In effect, many directional penalties are Lagrange multipliers, pushing a matched sample in the direction of satisfying a linear constraint that would not be satisfied without penalization. The method and example are in an R package DiPs at CRAN .     

Integration of Sensory Force Feedback Is Disturbed in CRPS-Related Dystonia

Complex regional pain syndrome (CRPS) is characterized by pain and disturbed blood flow, temperature regulation and motor control. Approximately 25% of cases develop fixed dystonia. The origin of this movement disorder is poorly understood, although recent insights suggest involvement of disturbed force feedback. Assessment of sensorimotor integration may provide insight into the pathophysiology of fixed dystonia. Sensory weighting is the process of integrating and weighting sensory feedback channels in the central nervous system to improve the state estimate. It was hypothesized that patients with CRPS-related dystonia bias sensory weighting of force and position toward position due to the unreliability of force feedback. The current study provides experimental evidence for dysfunctional sensory integration in fixed dystonia, showing that CRPS-patients with fixed dystonia weight force and position feedback differently than controls do. The study shows reduced force feedback weights in CRPS-patients with fixed dystonia, making it the first to demonstrate disturbed integration of force feedback in fixed dystonia, an important step towards understanding the pathophysiology of fixed dystonia.     

Using Multiple Control Groups and Matching to Address Unobserved Biases in Comparative Effectiveness Research

Studies of large policy interventions typically do not involve randomization. Adjustments, such as matching, can remove the bias due to observed covariates, but residual confounding remains a concern. In this paper we introduce two analytical strategies to bolster inferences of the effectiveness of policy interventions based on observational data. First, we identify how study groups may differ and then select a second comparison group on this source of difference. Second, we match subjects using a strategy that finely balances the distributions of key categorical covariates and stochastically balances on other covariates. An observational study of the effect of parity on the severely ill subjects enrolled in the Federal Employees Health Benefits (FEHB) Program illustrates our methods.     

Mating tactics in external fertilizers when sperm is limited

Among externally fertilizing animals in aquatic habitats, theproportion of a female's egg clutch that is successfully fertilizedoften falls below 100%. In many such species, particularly incoral reef fishes, males spawn daily at high frequencies, oftenwith little or no sperm competition. A major evolutionary problemfor such males is how best to allocate sperm over successivespawns. Females face the problem of ensuring complete fertilizationof their egg clutch. Here we model male and female mating tacticswhen daily sperm production is limited and with various assumptionsconcerning how differences in the number of sperm released duringa mating influence the number of eggs fertilized. The modelsreveal conditions under which males can maximize daily reproductivesuccess, either by releasing a fixed number of sperm duringall successive spawns or by matching sperm numbers to the clutchsize of their mates. These patterns of sperm allocation exertdifferent pressures on females, which may respond evolutionarilyby developing various mating tactics of their own.     

aCGH.Spline--an R package for aCGH dye bias normalization

MOTIVATION: The careful normalization of array-based comparative genomic hybridization (aCGH) data is of critical importance for the accurate detection of copy number changes. The difference in labelling affinity between the two fluorophores used in aCGH-usually Cy5 and Cy3-can be observed as a bias within the intensity distributions. If left unchecked, this bias is likely to skew data interpretation during downstream analysis and lead to an increased number of false discoveries. RESULTS: In this study, we have developed aCGH.Spline, a natural cubic spline interpolation method followed by linear interpolation of outlier values, which is able to remove a large portion of the dye bias from large aCGH datasets in a quick and efficient manner. Conclusions: We have shown that removing this bias and reducing the experimental noise has a strong positive impact on the ability to detect accurately both copy number variation (CNV) and copy number alterations (CNA).     

Inverse sampling of controls in a matched case control study

A method of inverse sampling of controls in a matched case-control study is described in which, for each case, controls are sampled until a discordant set is achieved. For a binary exposure, inverse sampling is used to determine the number of controls for each case. When most individuals in a population have the same exposure, standard case-control sampling may result in many case-control sets being concordant with respect to exposure and thus uninformative in the conditional logistic analysis. The method using inverse control sampling is proposed as a solution to this problem in situations when it is practically feasible. In many circumstances, inverse control sampling is found to offer improved statistical efficiency relative to a comparable study with a fixed number of controls per case.     

Comparison of photo‐matching algorithms commonly used for photographic capture–recapture studies

Photographic capture–recapture is a valuable tool for obtaining demographic information on wildlife populations due to its noninvasive nature and cost‐effectiveness. Recently, several computer‐aided photo‐matching algorithms have been developed to more efficiently match images of unique individuals in databases with thousands of images. However, the identification accuracy of these algorithms can severely bias estimates of vital rates and population size. Therefore, it is important to understand the performance and limitations of state‐of‐the‐art photo‐matching algorithms prior to implementation in capture–recapture studies involving possibly thousands of images. Here, we compared the performance of four photo‐matching algorithms; Wild‐ID, I3S Pattern+, APHIS, and AmphIdent using multiple amphibian databases of varying image quality. We measured the performance of each algorithm and evaluated the performance in relation to database size and the number of matching images in the database. We found that algorithm performance differed greatly by algorithm and image database, with recognition rates ranging from 100% to 22.6% when limiting the review to the 10 highest ranking images. We found that recognition rate degraded marginally with increased database size and could be improved considerably with a higher number of matching images in the database. In our study, the pixel‐based algorithm of AmphIdent exhibited superior recognition rates compared to the other approaches. We recommend carefully evaluating algorithm performance prior to using it to match a complete database. By choosing a suitable matching algorithm, databases of sizes that are unfeasible to match “by eye” can be easily translated to accurate individual capture histories necessary for robust demographic estimates.     

A model-based analysis of microarray experimental error and normalisation

A statistical model is proposed for the analysis of errors in microarray experiments and is employed in the analysis and development of a combined normalisation regime. Through analysis of the model and two-dye microarray data sets, this study found the following. The systematic error introduced by microarray experiments mainly involves spot intensity-dependent, feature-specific and spot position-dependent contributions. It is difficult to remove all these errors effectively without a suitable combined normalisation operation. Adaptive normalisation using a suitable regression technique is more effective in removing spot intensity-related dye bias than self-normalisation, while regional normalisation (block normalisation) is an effective way to correct spot position-dependent errors. However, dye-flip replicates are necessary to remove feature-specific errors, and also allow the analyst to identify the experimentally introduced dye bias contained in non-self-self data sets. In this case, the bias present in the data sets may include both experimentally introduced dye bias and the biological difference between two samples. Self-normalisation is capable of removing dye bias without identifying the nature of that bias. The performance of adaptive normalisation, on the other hand, depends on its ability to correctly identify the dye bias. If adaptive normalisation is combined with an effective dye bias identification method then there is no systematic difference between the outcomes of the two methods.     

Matching and ideal free distributions

Ideal free distributions characterise how a group of animals should be distributed between sources of food. In a simple case, the ratio of the number of animals matches the ratio of input rates, a result known as input matching. The matching law characterises how an animal should allocate responses to sources of food. If matching holds then the ratio of behaviour allocated to the sources matches the ratio of rewards obtained. Several authors have drawn attention to the analogy between input matching and the matching law. I present a critical review of this topic, and go on to investigate the claim that the distribution of animals should be analyzed in the same way that the generalized matching law is analyzed. This involves assuming that the relationship between the ratio of animals and the ratio of resources obeys a power law with two parameters, one corresponding to bias and the other to sensitivity. On this view, a plot of the log of the ratio of animals against the log of the ratio of input rates will be a straight line and its intercept and slope will provide estimates of bias and sensitivity. I show that this approach can give a good fit even when the underlying model does not result in a power law. A consequence is that the parameters estimated cannot be interpreted as bias and sensitivity. I conclude that there are no compelling reasons for analyzing the distribution of animals using log-log plots, and the method has the disadvantage of encouraging the mistaken view that deviations from input matching follow a power law.     

Inverse Sampling in Case Control Studies Involving a Rare Exposure

In case-control studies where the outcome is not uncommon but the exposure is rare, inverse sampling may be used to reduce the total number of subjects required to find a fixed number of exposed cases and controls. The sampling distribution is negative binomial rather than binomial. Logistic regression for adjustment of covariates may be implemented on the computer program GLIM by the appropriate use of macros. An example is given.     

Impact of random and systematic recall errors and selection bias in case--control studies on mobile phone use and brain tumors in adolescents (CEFALO study)

Whether the use of mobile phones is a risk factor for brain tumors in adolescents is currently being studied. Case--control studies investigating this possible relationship are prone to recall error and selection bias. We assessed the potential impact of random and systematic recall error and selection bias on odds ratios (ORs) by performing simulations based on real data from an ongoing case--control study of mobile phones and brain tumor risk in children and adolescents (CEFALO study). Simulations were conducted for two mobile phone exposure categories: regular and heavy use. Our choice of levels of recall error was guided by a validation study that compared objective network operator data with the self-reported amount of mobile phone use in CEFALO. In our validation study, cases overestimated their number of calls by 9% on average and controls by 34%. Cases also overestimated their duration of calls by 52% on average and controls by 163%. The participation rates in CEFALO were 83% for cases and 71% for controls. In a variety of scenarios, the combined impact of recall error and selection bias on the estimated ORs was complex. These simulations are useful for the interpretation of previous case-control studies on brain tumor and mobile phone use in adults as well as for the interpretation of future studies on adolescents.     

Some properties of an estimator for the basic reproduction number of the general epidemic model.

In this paper some properties of a convenient estimator, derived from a martingale estimating function, for the basic reproduction number of the general epidemic model are given for both finite and large samples. These properties give some guidelines for using this convenient estimator. It is shown that it underestimates the parameter and that the bias tends to zero when the population size and the initial number of infectives are increased simultaneously. The bias cannot be removed for a fixed number of introductory infectives. However, the estimator is asymptotically unbiased, conditional on a major outbreak. A simulation study shows that the central limit theorem applies for moderate population sizes.     

Correcting for measurement error in individual-level covariates in nonlinear mixed effects models

The nonlinear mixed effects model is used to represent data in pharmacokinetics, viral dynamics, and other areas where an objective is to elucidate associations among individual-specific model parameters and covariates; however, covariates may be measured with error. For additive measurement error, we show substitution of mismeasured covariates for true covariates may lead to biased estimators for fixed effects and random effects covariance parameters, while regression calibration may eliminate bias in fixed effects but fail to correct that in covariance parameters. We develop methods to take account of measurement error that correct this bias and may be implemented with standard software, and we demonstrate their utility via simulation and application to data from a study of HIV dynamics.     

Description of a new computer wire coding method and its application to evaluate potential control selection bias in the Savitz et al. childhood cancer study

We developed a new computer wire coding method and then applied it to investigate the suggestion that control selection bias might explain the observed association between wire codes and childhood cancer made in the study conducted by Savitz et al. in the Denver area. The computer wire coding method used a geographic information system approach with data on the local distribution electric system and from tax assessor records. Individual residences were represented as a circle scaled to the ground floor area of the residence and centered on the lot centroid. The wire code of the residence was determined from the distance between the circle and the relevant power line, and from the current carrying capacity of that line. Using this method, wire codes were generated for 238 290 residences built before 1986, the time of the Savitz et al. study, in the Denver metropolitan area. We then attempted to reconstruct the 1985 population of hypothetically eligible control children in the Denver metropolitan area by using 1980 census data. Since data were not available to locate the children in each residence within a census block, uniform, Poisson, and negative binomial distributions were used to randomly assign children to residences. To evaluate the likelihood of the wire code distribution of the controls selected by Savitz et al., 100 random trials were conducted for each distribution, matching two controls to each case. The odds ratios between childhood cancer and very high current configuration (VHCC) wire codes were reduced when the assigned controls were used, suggesting control selection bias may have been present. However, control selection bias is unlikely to account for all the reported association between childhood cancer and wire codes in the Savitz et al. study.     

Dynamic sizing of multilayer perceptrons

This article proposes a stochastic method for determining the number of hidden nodes of a multilayer perceptron trained by a backpropagation algorithm. During the learning process, an auxiliary markovian algorithm controls the sizing of the hidden layers. As usual, the main idea is to promote the addition of nodes the closer the net is to a stall configuration, and to remove those units not sufficiently “lively”. The combined algorithm produces families of nets which converge fast towards well trained nets with a small number of nodes. Numerical experiments are performed both on conventional benchmarks and on realistic learning problems.These experiments show that for learning tasks of sufficiently high complexity, the additional (with respect to the conventional fixed architecture methods) complexity of our method is compensated by a greater velocity and a higher success percentage in obtaining the minimum of the error function. Received: 7 December 1992/Accepted in revised form: 23 September 1993     

Sex ratios observed in 80 species of parrots

A number of potential evolutionary and physiological factors may be involved in avian sex ratio bias so that under certain conditions a sex ratio bias may favour males or females within a population. In addition different factors may be important in manipulating sex ratio bias through the different life stages. In this study sex ratio bias was examined in a total of 16 570 captive parrots, representing 80 species, many of which are endangered in the wild, using database records originating form commercial laboratories that offer genetic sexing. Within the species examined 72% showed a male bias this was significant in three species, when adjusted for multiple comparisons. This preliminary study is limited due to lack of data on the age of the individuals sampled. However, the large dataset do suggest that this phenomenon should be further considered by investigators working at a species level where such data can be collected.     

A Multistage Gene Normalization System Integrating Multiple Effective Methods

Gene/protein recognition and normalization is an important preliminary step for many biological text mining tasks. In this paper, we present a multistage gene normalization system which consists of four major subtasks: pre-processing, dictionary matching, ambiguity resolution and filtering. For the first subtask, we apply the gene mention tagger developed in our earlier work, which achieves an F-score of 88.42% on the BioCreative II GM testing set. In the stage of dictionary matching, the exact matching and approximate matching between gene names and the EntrezGene lexicon have been combined. For the ambiguity resolution subtask, we propose a semantic similarity disambiguation method based on Munkres'' Assignment Algorithm. At the last step, a filter based on Wikipedia has been built to remove the false positives. Experimental results show that the presented system can achieve an F-score of 90.1%, outperforming most of the state-of-the-art systems.     

The relationship between third-codon position nucleotide content, codon bias, mRNA secondary structure and gene expression in the drosophilid alcohol dehydrogenase genes Adh and Adhr

To gain insights into the relationship between codon bias, mRNA secondary structure, third-codon position nucleotide distribution, and gene expression, we predicted secondary structures in two related drosophilid genes, Adh and Adhr, which differ in degree of codon bias and level of gene expression. Individual structural elements (helices) were inferred using the comparative method. For each gene, four types of randomization simulations were performed to maintain/remove codon bias and/or to maintain or alter third-codon position nucleotide composition (N3). In the weakly expressed, weakly biased gene Adhr, the potential for secondary structure formation was found to be much stronger than in the highly expressed, highly biased gene Adh. This is consistent with the observation of approximately equal G and C percentages in Adhr ( approximately 31% across species), whereas in Adh the N3 distribution is shifted toward C (42% across species). Perturbing the N3 distribution to approximately equal amounts of A, G, C, and T increases the potential for secondary structure formation in Adh, but decreases it in Adhr. On the other hand, simulations that reduce codon bias without changing N3 content indicate that codon bias per se has only a weak effect on the formation of secondary structures. These results suggest that, for these two drosophilid genes, secondary structure is a relatively independent, negative regulator of gene expression. Whereas the degree of codon bias is positively correlated with level of gene expression, strong individual secondary structural elements may be selected for to retard mRNA translation and to decrease gene expression.