Search results

Statistical methods of spatial analysis are often successful at either prediction or explanation, but not necessarily both. In a recent paper, Dearmon and Smith (2016) showed that by combining Gaussian Process Regression (GPR) with Bayesian Model Averaging (BMA), a modeling framework could be developed in which both needs are addressed. In particular, the smoothness properties of GPR together with the robustness of BMA allow local spatial analyses of individual variable effects that yield remarkably stable results. However, this GPR-BMA approach is not without its limitations. In particular, the standard (isotropic) covariance kernel of GPR treats all explanatory variables in a symmetric way that limits the analysis of their individual effects. Here we extend this approach by introducing a mixture of kernels (both isotropic and anisotropic) which allow different length scales for each variable. To do so in a computationally efficient manner, we also explore a number of Bayes-factor approximations that avoid the need for costly reversible-jump Monte Carlo methods.

To demonstrate the effectiveness of this Variable Length Scale (VLS) model in terms of both predictions and local marginal analyses, we employ selected simulations to compare VLS with Geographically Weighted Regression (GWR), which is currently the most popular method for such spatial modeling. In addition, we employ the classical Boston Housing data to compare VLS not only with GWR but also with other well-known spatial regression models that have been applied to this same data. Our main results are to show that VLS not only compares favorably with spatial regression at the aggregate level but is also far more accurate than GWR at the local level.

Past research has shown that forecast combination typically improves demand forecast accuracy even when only two component forecasts are used; however, systematic bias in the component forecasts can reduce the effectiveness of combination. This study proposes a methodology for combining demand forecasts that are biased. Data from an actual manufacturing shop are used to develop the methodology and compare its accuracy with the accuracy of the standard approach of correcting for bias prior to combination. Results indicate that the proposed methodology outperforms the standard approach.

This chapter examines the use of mathematical programming to remove systematic bias from demand forecasts. A debiasing methodology is developed and applied to demand data from an actual service operation. The accuracy of the proposed methodology is compared to the accuracy of a well-known approach that utilizes ordinary least squares regression. Results indicate that the proposed method outperforms the least squares approach.

This paper seeks to connect changes in the structure of wages at the occupation level to measures of the task content of jobs. We first present a simple model where skills are used to produce tasks, and changes in task prices are the underlying source of change in occupational wages. Using Current Population Survey (CPS) wage data and task measures from the O*NET, we document large changes in both the within and between dimensions of occupational wages over time, and find that these changes are well explained by changes in task prices likely induced by technological change and offshoring.

Standard stratified sampling (SSS) is a popular non-random sampling scheme. Maximum likelihood estimator (MLE) is inconsistent if some sampled strata depend on the response variable Y (‘endogenous samples’) or if some Y-dependent strata are not sampled at all (‘truncated sample’ – a missing data problem). Various versions of MLE have appeared in the literature, and this paper reviews practical likelihood-based estimators for endogenous or truncated samples in SSS. Also a new estimator ‘Estimated-EX MLE’ is introduced using an extra random sample on X (not on Y) to estimate the distribution EX of X. As information on Y may be hard to get, this estimator's data demand is weaker than an extra random sample on Y in some other estimators. The estimator can greatly improve the efficiency of ‘Fixed-X MLE’ which conditions on X, even if the extra sample size is small. In fact, Estimated-EX MLE does not estimate the full FX as it needs only a sample average using the extra sample. Estimated-EX MLE can be almost as efficient as the ‘Known-FX MLE’. A small-scale simulation study is provided to illustrate these points.

This research examines the effects of health, location, and other factors on receipt of wage income for young heads of households, aged 19 to 25, who lived in HUD-assisted housing and in other rental housing in 2011.

This chapter reports results of analyses of the 2011 American Housing Survey, merged with HUD administrative records, available as a public-use file at the U.S. Census Bureau.

Nineteen percent of young householders in assisted housing and 8% in other rental housing reported less than good health or a disability. Nearly two-thirds of young householders in assisted housing reported receipt of earned income. For respondents in assisted housing who reported good health and no disabilities, logistic regression models suggest that educational attainment beyond a high school diploma, more than one adult in the household, and living in metropolitan areas in the Midwest or West census regions were positively and statistically significant for receipt of earned income. For respondents in both assisted and other rental housing who reported less than good health and/or disabilities, residence in assisted housing or educational attainment beyond a high school diploma were positively associated with receipt of earned income, while residence in the metropolitan South lowered the odds of receipt of earned income.

Success of self-sufficiency programs will depend on accommodating the imperatives created by health, disability, and structural impediments created by a market economy.

This is the first analysis of health/disability and other barriers to paid employment that accurately identifies a nationally representative sample of young Millennials in HUD-assisted and other rental housing.

The payoff to schooling among the foreign born in the United States is only around one-half of the payoff for the native born. This paper examines whether this differential is related to the quality of the schooling immigrants acquired abroad. The paper uses the overeducation/required education/undereducation specification of the earnings equation to explore the transmission mechanism for the origin-country school-quality effects. It also assesses the empirical merits of two alternative measures of the quality of schooling undertaken abroad. The results suggest that a higher quality of schooling acquired abroad is associated with a higher payoff to schooling among immigrants in the US labor market. This higher payoff is associated with a higher payoff to correctly matched schooling in the United States, and a greater (in absolute value) penalty associated with years of undereducation. A set of predictions is presented to assess the relative importance of these channels, and the undereducation channel is shown to be the more influential factor. This channel is linked to greater positive selection in migration among those from countries with better quality schools. In other words, it is the impact of origin-country school quality on the immigrant selection process, rather than the quality of immigrants' schooling per se, that is the major driver of the lower payoff to schooling among immigrants in the United States.

This paper presents a mathematical programming model to reduce bias for both aggregate demand forecasts and lower echelon forecasts comprising a hierarchical forecasting system. Demand data from an actual service operation are used to illustrate the model and compare its accuracy with a standard approach for hierarchical forecasting. Results show that the proposed methodology outperforms the standard approach.

Customer experience is a relevant concept in marketing and tourism research since its correct understanding allows companies to achieve competitive advantage and service providers can reach several outcomes such as customer engagement, loyalty, and customer satisfaction. This chapter aims to analyze one of the main outcomes of the customer experience, the customer satisfaction through online reviews, and using spatial analysis as a tool to incorporate the contextual nature of the customer experience. Thus, our study considers online rating as a measure of customer satisfaction and tries to analyze the impact of actions under the control of the service provider (price and objective quality) and actions under the control of the customer (subjective quality) on customer satisfaction.

With the Spanish hotel industry as a study framework, an empirical study is developed to analyze, through geographically weighted regression techniques, the relationship between price, objective quality and subjective quality, and online ratings given by consumers with a sample of 1870 of geolocated hotels in Spain. The findings show how a premium price, depending on the geolocation, is an indicator for better customer experiences, and they also show that objective quality is the antecedent of customer experience whose positive effect on customer satisfaction is geographically more widespread. Results show contradictory effects of subjective quality, while in some areas subjective quality does not match the product fit of customers, in others it allows hotels to provide more satisfactory experiences.