A two-phase sampling strategy for large-scale forest carbon budgets

The estimation of forest carbon sources and sinks is an important issue in forest inventories. For monitoring forest carbon budgets, a two-phase strategy is proposed. In the first phase, N points are thrown onto the study region by means of unaligned systematic sampling and then classified as forest or non-forest on the basis of aerial information. In the second phase, a sample of n points of those classified as forest is achieved by means of a sampling scheme with selection probabilities proportional to the sizes of an auxiliary variable. Subsequently, circular plots of adequate size are centred in the n selected points and visited in order to quantify the carbon amount within. Unbiased estimators of the forest area size as well as of the overall amount of carbon in that area and a conservative estimator of their variance–covariance matrix are derived. The strategy is applied to estimate carbon budgets in the forests of Trentino (North Italy).     

On the existence and uniqueness of the NPMLE in biased sampling models

Identifiability, existence and uniqueness of the non-parametric maximum likelihood estimator (NPMLE) in biased sampling models are thoroughly investigated. The conditions for existence and uniqueness are considered from a different perspective and proofs are simplified. An extension to semi-parametric models is provided.     

Two-stage informative cluster sampling—estimation and prediction with applications for small-area models

This paper considers the effects of informative two-stage cluster sampling on estimation and prediction. The aims of this article are twofold: first to estimate the parameters of the superpopulation model for two-stage cluster sampling from a finite population, when the sampling design for both stages is informative, using maximum likelihood estimation methods based on the sample-likelihood function; secondly to predict the finite population total and to predict the cluster-specific effects and the cluster totals for clusters in the sample and for clusters not in the sample. To achieve this we derive the sample and sample-complement distributions and the moments of the first and second stage measurements. Also we derive the conditional sample and conditional sample-complement distributions and the moments of the cluster-specific effects given the cluster measurements. It should be noted that classical design-based inference that consists of weighting the sample observations by the inverse of sample selection probabilities cannot be applied for the prediction of the cluster-specific effects for clusters not in the sample. Also we give an alternative justification of the Royall [1976. The linear least squares prediction approach to two-stage sampling. Journal of the American Statistical Association 71, 657–664] predictor of the finite population total under two-stage cluster population. Furthermore, small-area models are studied under informative sampling.     

When does an imperfect sampling frame produce more efficient estimators than a perfect frame?

The issue of when imperfect sampling frames can result in more efficient estimators of population totals than perfect frames is explored. Our analysis is based on an expression we call the difference score. We show how, when properly expanded it provides an illuminating basis for comparing a weighted estimator under an imperfect frame with that of a conventional estimator assuming the frame has been corrected. Specifically, the circumstances (i.e., population and frame characteristics) under which an imperfect frame results in estimates of population totals that are more precise than those from a perfect frame can in many cases be discerned by analytically examining the terms in the expansion of this difference score. In addition, a classification tree methodology was used to further explore circumstances under which imperfect frames result in more precise estimators. The results of this analytical study complement, strengthen, and in many cases explain those discovered in an earlier empirical investigation that lead to recommendations as to when to correct a frame or when to adjust for imperfection using a weighting methodology called the arc weight estimator.     

Gibbs抽样条件下瓦格纳法则的中国有效性研究

 本文在Gibbs抽样条件下,利用我国1952-2006年经济增长和财政支出的时间序列数据,对瓦格纳法则在中国有效性问题进行了实证检验,并具体考察了中国政府公共支出规模扩张与经济增长之间的平滑变参数特征。实证结果表明中国经济增长与政府公共支出增长之间的因果关系并不确定,两者在数量上的关系具有明显的阶段性特点,其参数在具体形态上呈平滑变参数特征。     

Estimating network properties from snowball sampled data

This article addresses the estimation of topological network parameters from data obtained with a snowball sampling design. An approximate expression for the probability of a vertex to be included in the sample is derived. Based on this sampling distribution, estimators for the mean degree, the degree correlation, and the clustering coefficient are proposed. The performance of these estimators and their sensitivity with respect to the response rate are validated through Monte Carlo simulations on several test networks. Our approach has no complex computational requirements and is straightforward to apply to real-world survey data. In a snowball sample design, each respondent is typically enquired only once. Different from the widely used estimator for Respondent-Driven Sampling (RDS), which assumes sampling with replacement, the proposed approach relies on sampling without replacement and is thus also applicable for large sample fractions. From the simulation experiments, we conclude that the estimation quality decreases with increasing variance of the network degree distribution. Yet, if the degree distribution is not to broad, our approach results in good estimates for the mean degree and the clustering coefficient, which, moreover, are almost independent from the response rate. The estimates for the degree correlation are of moderated quality.     

Estimating network structure via random sampling: Cognitive social structures and the adaptive threshold method

This paper introduces and tests a novel methodology for measuring networks. Rather than collecting data to observe a network or several networks in full, which is typically costly or impossible, we randomly sample a portion of individuals in the network and estimate the network based on the sampled individuals’ perceptions on all possible ties. We find the methodology produces accurate estimates of social structure and network level indices in five different datasets. In order to illustrate the performance of our approach we compare its results with the traditional roster and ego network methods of data collection. Across all five datasets, our methodology outperforms these standard social network data collection methods. We offer ideas on applications of our methodology, and find it especially promising in cross-network settings.     

Measuring time use in surveys - Concordance of survey and experience sampling measures

It is still unclear to what extent time allocation retrospectively reported in questionnaires reflects people’s actual behavior. Addressing this research gap, we analyze the congruence of time use information assessed through retrospective questionnaires and through experience sampling methodology. Participants completed standard survey questions on time allocation. In addition, a mobile-phone-based experience sampling technology obtained snapshots of, on average, 54 momentary activities in which respondents participated while pursuing their normal daily routines.Results indicate that the associations between standard survey questions and experience sampling methods are quite substantial for long-lasting and externally structured activities, such as paid work. In contrast, associations between survey and experience sampling methods are somewhat weaker for less externally structured, short-term and infrequent activities, such as errands, housework, and leisure. However, further research is required to elucidate which method (experience sampling method or survey questions) results in more reliable and valid measures for short-term and sporadic activities.     

Mass imputation for two-phase sampling

Two-phase sampling is a cost-effective method of data collection using outcome-dependent sampling for the second-phase sample. In order to make efficient use of auxiliary information and to improve domain estimation, mass imputation can be used in two-phase sampling. Rao and Sitter (1995) introduce mass imputation for two-phase sampling and its variance estimation under simple random sampling in both phases. In this paper, we extend the Rao–Sitter method to general sampling design. The proposed method is further extended to mass imputation for categorical data. A limited simulation study is performed to examine the performance of the proposed methods.     

From grouped to de-grouped data: a new approach in distribution fitting for grouped data

Sampling within a given interval with a constraint has not been previously considered. Standard parametric simulation engines require knowledge of the parameters of the distribution from which a sample is drawn. These methods are limited if additional constrains are required for the simulated data. We propose a method that generates the targeted number of individual observations within a given interval with a constraint that the average value of observations is known. This method is further extended to a grouped data setting, as a way of data de-grouping, when the frequency and average value of observations are provided for each group. Several simulation studies are employed to evaluate the performance of the proposed method, in case of both a single interval and grouped data, for different simulation settings. Furthermore, the proposed method is evaluated in the parameter recovery when different distributions are fitted to the de-grouped data. This method is found to be superior to the uniform method previously used in data de-grouping. The results of the simulation study are promising and they show that this method can be used successfully in the applications where data de-grouping requires that the average value of observations is maintained in each group. The application of the proposed method is illustrated on a real data of insurance losses for bodily injury claims.