Jackknife for bias reduction in predictive regressions

One of the fundamental econometric models in finance is predictive regression. The standard least squares method produces biased coefficient estimates when the regressor is persistent and its innovations are correlated with those of the dependent variable. This article proposes a general and convenient method based on the jackknife technique to tackle the estimation problem. The proposed method reduces the bias for both single- and multiple-regressor models and for both short- and long-horizon regressions. The effectiveness of the proposed method is demonstrated by simulations. An empirical application to equity premium prediction using the dividend yield and the short rate highlights the differences between the results by the standard approach and those by the bias-reduced estimator. The significant predictive variables under the ordinary least squares become insignificant after adjusting for the finite-sample bias. These discrepancies suggest that bias reduction in predictive regressions is important in practical applications.

Approximation of Bayesian predictive p-values with regression ABC

In the Bayesian framework a standard approach to model criticism is to compare some function of the observed data to a reference predictive distribution. The result of the comparison can be summarized in the form of a p-value, and it's well known that computation of some kinds of Bayesian predictive p-values can be challenging. The use of regression adjustment approximate Bayesian computation (ABC) methods is explored for this task. Two problems are considered. The first is the calibration of posterior predictive p-values so that they are uniformly distributed under some reference distribution for the data. Computation is difficult because the calibration process requires repeated approximation of the posterior for different data sets under the reference distribution. The second problem considered is approximation of distributions of prior predictive p-values for the purpose of choosing weakly informative priors in the case where the model checking statistic is expensive to compute. Here the computation is difficult because of the need to repeatedly sample from a prior predictive distribution for different values of a prior hyperparameter. In both these problems we argue that high accuracy in the computations is not required, which makes fast approximations such as regression adjustment ABC very useful. We illustrate our methods with several samples.

Comparison of three expert elicitation methods for logistic regression on predicting the presence of the threatened brush-tailed rock-wallaby petrogale penicillata

Numerous expert elicitation methods have been suggested for generalised linear models (GLMs). This paper compares three relatively new approaches to eliciting expert knowledge in a form suitable for Bayesian logistic regression. These methods were trialled on two experts in order to model the habitat suitability of the threatened Australian brush-tailed rock-wallaby (Petrogale penicillata). The first elicitation approach is a geographically assisted indirect predictive method with a geographic information system (GIS) interface. The second approach is a predictive indirect method which uses an interactive graphical tool. The third method uses a questionnaire to elicit expert knowledge directly about the impact of a habitat variable on the response. Two variables (slope and aspect) are used to examine prior and posterior distributions of the three methods. The results indicate that there are some similarities and dissimilarities between the expert informed priors of the two experts formulated from the different approaches. The choice of elicitation method depends on the statistical knowledge of the expert, their mapping skills, time constraints, accessibility to experts and funding available. This trial reveals that expert knowledge can be important when modelling rare event data, such as threatened species, because experts can provide additional information that may not be represented in the dataset. However care must be taken with the way in which this information is elicited and formulated.

Beyond statistical procedures for predictive modelling: Data mining algorithms and support for university research at QUT

In a seminal data mining article, Leo Breiman [1] argued that to develop effective predictive classification and regression models, we need to move away from the sole dependency on statistical algorithms and embrace a wider toolkit of modeling algorithms that include data mining procedures. Nevertheless, many researchers still rely solely on statistical procedures when undertaking data modeling tasks; the sole reliance on these procedures has lead to the development of irrelevant theory and questionable research conclusions ([1], p.199). We will outline initiatives that the HPC & Research Support group is undertaking to engage researchers with data mining tools and techniques; including a new range of seminars, workshops, and one-on-one consultations covering data mining algorithms, the relationship between data mining and the research cycle, and limitations and problems with these new algorithms. Organisational limitations and restrictions to these initiatives are also discussed.

Predictive factors of contact lens case contamination

PURPOSE: To examine the relationship between contact lens (CL) case contamination and various potential predictive factors. METHODS: 74 subjects were fitted with lotrafilcon B (CIBA Vision) CLs on a daily wear basis for 1 month. Subjects were randomly assigned one of two polyhexamethylene biguanide (PHMB) preserved disinfecting solutions with the corresponding regular lens case. Clinical evaluations were conducted at lens delivery and after 1 month, when cases were collected for microbial culture. A CL care non-compliance score was determined through administration of a questionnaire and the volume of solution used was calculated for each subject. Data was examined using backward stepwise binary logistic regression. RESULTS: 68% of cases were contaminated. 35% were moderately or heavily contaminated and 36% contained gram-negative bacteria. Case contamination was significantly associated with subjective dryness symptoms (OR 4.22, CI 1.37–13.01) (P<0.05). There was no association between contamination and subject age, ethnicity, gender, average wearing time, amount of solution used, non-compliance score, CL power and subjective redness (P>0.05). The effect of lens care system on case contamination approached significance (P=0.07). Failure to rinse the case with disinfecting solution following CL insertion (OR 2.51, CI 0.52–12.09) and not air drying the case (OR 2.31, CI 0.39–13.35) were positively correlated with contamination; however, did not reach statistical significance. CONCLUSIONS: Our results suggest that case contamination may influence subjective comfort. It is difficult to predict the development of case contamination from a variety of clinical factors. The efficacy of CL solutions, bacterial resistance to disinfection and biofilm formation are likely to play a role. Further evaluation of these factors will improve our understanding of the development of case contamination and its clinical impact.

Developing a predictive contractor satisfaction model (CoSMo) for construction projects

Over the last few decades, construction project performance has been evaluated due to the increase of delays, cost overruns and quality failures. Growing numbers of disputes, inharmonious working environments, conflict, blame cultures, and mismatches of objectives among project teams have been found to be contributory factors to poor project performance. Performance measurement (PM) approaches have been developed to overcome these issues, however, the comprehensiveness of PM as an overall approach is still criticised in terms of the iron triangle; namely time, cost, and quality. PM has primarily focused on objective measures, however, continuous improvement requires the inclusion of subjective measures, particularly contractor satisfaction (Co-S). It is challenging to deal with the two different groups of large and small-medium contractor satisfaction as to date, Co-S has not been extensively defined, primarily in developing countries such as Malaysia. Therefore, a Co-S model is developed in this research which aims to fulfil the current needs in the construction industry by integrating performance measures to address large and small-medium contractor perceptions. The positivist paradigm used in the research was adhered to by reviewing relevant literature and evaluating expert discussions on the research topic. It yielded a basis for the contractor satisfaction model (CoSMo) development which consists of three elements: contractor satisfaction (Co-S) dimensions; contributory factors and characteristics (project and participant). Using valid questionnaire results from 136 contractors in Malaysia lead to the prediction of several key factors of contractor satisfaction and to an examination of the relationships between elements. The relationships were examined through a series of sequential statistical analyses, namely correlation, one-way analysis of variance (ANOVA), t-tests and multiple regression analysis (MRA). Forward and backward MRAs were used to develop Co-S mathematical models. Sixteen Co-S models were developed for both large and small-medium contractors. These determined that the large contractor Malaysian Co-S was most affected by the conciseness of project scope and quality of the project brief. Contrastingly, Co-S for small-medium contractors was strongly affected by the efficiency of risk control in a project. The results of the research provide empirical evidence in support of the notion that appropriate communication systems in projects negatively contributes to large Co-S with respect to cost and profitability. The uniqueness of several Co-S predictors was also identified through a series of analyses on small-medium contractors. These contractors appear to be less satisfied than large contractors when participants lack effectiveness in timely authoritative decision-making and communication between project team members. Interestingly, the empirical results show that effective project health and safety measures are influencing factors in satisfying both large and small-medium contractors. The perspectives of large and small-medium contractors in respect to the performance of the entire project development were derived from the Co-S models. These were statistically validated and refined before a new Co-S model was developed. Developing such a unique model has the potential to increase project value and benefit all project participants. It is important to improve participant collaboration as it leads to better project performance. This study may encourage key project participants; such as client, consultant, subcontractor and supplier; to increase their attention to contractor needs in the development of a project. Recommendations for future research include investigating other participants‟ perspectives on CoSMo and the impact of the implementation of CoSMo in a project, since this study is focused purely on the contractor perspective.

Predictive modelling : parents’ decision making to use online child health information to increase their understanding and/or diagnose or treat their child’s health

Abstract Background The quantum increases in home Internet access and available online health information with limited control over information quality highlight the necessity of exploring decision making processes in accessing and using online information, specifically in relation to children who do not make their health decisions. Objectives To understand the processes explaining parents’ decisions to use online health information for child health care. Methods Parents (N = 391) completed an initial questionnaire assessing the theory of planned behaviour constructs of attitude, subjective norm, and perceived behavioural control, as well as perceived risk, group norm, and additional demographic factors. Two months later, 187 parents completed a follow-up questionnaire assessing their decisions to use online information for their child’s health care, specifically to 1) diagnose and/or treat their child’s suspected medical condition/illness and 2) increase understanding about a diagnosis or treatment recommended by a health professional. Results Hierarchical multiple regression showed that, for both behaviours, attitude, subjective norm, perceived behavioural control, (less) perceived risk, group norm, and (non) medical background were the significant predictors of intention. For parents’ use of online child health information, for both behaviours, intention was the sole significant predictor of behaviour. The findings explain 77% of the variance in parents’ intention to treat/diagnose a child health problem and 74% of the variance in their intentions to increase their understanding about child health concerns. Conclusions Understanding parents’ socio-cognitive processes that guide their use of online information for child health care is important given the increase in Internet usage and the sometimes-questionable quality of health information provided online. Findings highlight parents’ thirst for information; there is an urgent need for health professionals to provide parents with evidence-based child health websites in addition to general population education on how to evaluate the quality of online health information.

An analysis of regression models for predicting the speed of a wave glider autonomous surface vehicle

An important aspect of robotic path planning for is ensuring that the vehicle is in the best location to collect the data necessary for the problem at hand. Given that features of interest are dynamic and move with oceanic currents, vehicle speed is an important factor in any planning exercises to ensure vehicles are at the right place at the right time. Here, we examine different Gaussian process models to find a suitable predictive kinematic model that enable the speed of an underactuated, autonomous surface vehicle to be accurately predicted given a set of input environmental parameters.

A national satellite-based land-use regression model for air pollution exposure assessment in Australia

Land-use regression (LUR) is a technique that can improve the accuracy of air pollution exposure assessment in epidemiological studies. Most LUR models are developed for single cities, which places limitations on their applicability to other locations. We sought to develop a model to predict nitrogen dioxide (NO2) concentrations with national coverage of Australia by using satellite observations of tropospheric NO2 columns combined with other predictor variables. We used a generalised estimating equation (GEE) model to predict annual and monthly average ambient NO2 concentrations measured by a national monitoring network from 2006 through 2011. The best annual model explained 81% of spatial variation in NO2 (absolute RMS error=1.4 ppb), while the best monthly model explained 76% (absolute RMS error=1.9 ppb). We applied our models to predict NO2 concentrations at the ~350,000 census mesh blocks across the country (a mesh block is the smallest spatial unit in the Australian census). National population-weighted average concentrations ranged from 7.3 ppb (2006) to 6.3 ppb (2011). We found that a simple approach using tropospheric NO2 column data yielded models with slightly better predictive ability than those produced using a more involved approach that required simulation of surface-to-column ratios. The models were capable of capturing within-urban variability in NO2, and offer the ability to estimate ambient NO2 concentrations at monthly and annual time scales across Australia from 2006–2011. We are making our model predictions freely available for research.

Examining the predictive value of combining the theory of planned behaviour and the volunteer functions inventory

The functions of the volunteer functions inventory were combined with the constructs of the theory of planned behaviour (i.e., attitudes, subjective norms, and perceived behavioural control) to establish whether a stronger, single explanatory model prevailed. Undertaken in the context of episodic, skilled volunteering by individuals who were retired or approaching retirement (N = 186), the research advances on prior studies which either examined the predictive capacity of each model independently or compared their explanatory value. Using hierarchical regression analysis, the functions of the volunteer functions inventory (when controlling for demographic variables) explained an additional 7.0% of variability in individuals’ willingness to volunteer over and above that accounted for by the theory of planned behaviour. Significant predictors in the final model included attitudes, subjective norms and perceived behavioural control from the theory of planned behaviour and the understanding function from the volunteer functions inventory. It is proposed that the items comprising the understanding function may represent a deeper psychological construct (e.g., self-actualisation) not accounted for by the theory of planned behaviour. The findings highlight the potential benefit of combining these two prominent models in terms of improving understanding of volunteerism and providing a single parsimonious model for raising rates of this important behaviour.

An extended regression approach to estimating loads and their uncertainties in Great Barrier Reef catchments

There are numerous load estimation methods available, some of which are captured in various online tools. However, most estimators are subject to large biases statistically, and their associated uncertainties are often not reported. This makes interpretation difficult and the estimation of trends or determination of optimal sampling regimes impossible to assess. In this paper, we first propose two indices for measuring the extent of sampling bias, and then provide steps for obtaining reliable load estimates by minimizing the biases and making use of possible predictive variables. The load estimation procedure can be summarized by the following four steps: - (i) output the flow rates at regular time intervals (e.g. 10 minutes) using a time series model that captures all the peak flows; - (ii) output the predicted flow rates as in (i) at the concentration sampling times, if the corresponding flow rates are not collected; - (iii) establish a predictive model for the concentration data, which incorporates all possible predictor variables and output the predicted concentrations at the regular time intervals as in (i), and; - (iv) obtain the sum of all the products of the predicted flow and the predicted concentration over the regular time intervals to represent an estimate of the load. The key step to this approach is in the development of an appropriate predictive model for concentration. This is achieved using a generalized regression (rating-curve) approach with additional predictors that capture unique features in the flow data, namely the concept of the first flush, the location of the event on the hydrograph (e.g. rise or fall) and cumulative discounted flow. The latter may be thought of as a measure of constituent exhaustion occurring during flood events. The model also has the capacity to accommodate autocorrelation in model errors which are the result of intensive sampling during floods. Incorporating this additional information can significantly improve the predictability of concentration, and ultimately the precision with which the pollutant load is estimated. We also provide a measure of the standard error of the load estimate which incorporates model, spatial and/or temporal errors. This method also has the capacity to incorporate measurement error incurred through the sampling of flow. We illustrate this approach using the concentrations of total suspended sediment (TSS) and nitrogen oxide (NOx) and gauged flow data from the Burdekin River, a catchment delivering to the Great Barrier Reef. The sampling biases for NOx concentrations range from 2 to 10 times indicating severe biases. As we expect, the traditional average and extrapolation methods produce much higher estimates than those when bias in sampling is taken into account.