Bootstrap Resampling Methods Research Articles

Knowledge transfer effectiveness of university‐industry alliances

Purpose – This paper attempts to propose an integrated model for measuring the knowledge transfer effectiveness in university‐industry alliances. The so‐called “RDCE” model is thereby proposed as an integrated model for measuring the knowledge transfer effectiveness. By combining inter‐organizational relations (IORs), knowledge‐based view (KBV) and resource‐based view (RBV) of firms, this paper aims to focus on the influence of determinant factors such as partner complementarities, partner attributes, the characteristics of the coordination and relationship quality between industrial companies and universities that may lead to the effectiveness of knowledge transfer.Design/methodology/approach – This framework thereby clarifies how mediating variables influenced the paths that constitute the direct, indirect and total effects of mediated models by integrating moderated regression analysis together with bootstrap resampling methods to ensure the precision in estimating confidence intervals of indirect effe...

Homogeneity test of rate ratios in stratified matched-pair studies

This paper investigates homogeneity test of rate ratios in stratified matched-pair studies on the basis of asymptotic and bootstrap-resampling methods. Based on the efficient score approach, we develop a simple and computationally tractable score test statistic. Several other homogeneity test statistics are also proposed on the basis of the weighted least-squares estimate and logarithmic transformation. Sample size formulae are derived to guarantee a pre-specified power for the proposed tests at the pre-given significance level. Empirical results confirm that (i) the modified score statistic based on the bootstrap-resampling method performs better in the sense that its empirical type I error rate is much closer to the pre-specified nominal level than those of other tests and its power is greater than those of other tests, and is hence recommended, whilst the statistics based on the weighted least-squares estimate and logarithmic transformation are slightly conservative under some of the considered settings; (ii) the derived sample size formulae are rather accurate in the sense that their empirical powers obtained from the estimated sample sizes are very close to the pre-specified nominal powers. A real example is used to illustrate the proposed methodologies.

Identifying cluster number for subspace projected functional data clustering

We propose a new approach, the forward functional testing (FFT) procedure, to cluster number selection for functional data clustering. We present a framework of subspace projected functional data clustering based on the functional multiplicative random-effects model, and propose to perform functional hypothesis tests on equivalence of cluster structures to identify the number of clusters. The aim is to find the maximum number of distinctive clusters while retaining significant differences between cluster structures. The null hypotheses comprise equalities between the cluster mean functions and between the sets of cluster eigenfunctions of the covariance kernels. Bootstrap resampling methods are developed to construct reference distributions of the derived test statistics. We compare several other cluster number selection criteria, extended from methods of multivariate data, with the proposed FFT procedure. The performance of the proposed approaches is examined by simulation studies, with applications to clustering gene expression profiles.

MPI-PHYLIP: parallelizing computationally intensive phylogenetic analysis routines for the analysis of large protein families.

BackgroundPhylogenetic study of protein sequences provides unique and valuable insights into the molecular and genetic basis of important medical and epidemiological problems as well as insights about the origins and development of physiological features in present day organisms. Consensus phylogenies based on the bootstrap and other resampling methods play a crucial part in analyzing the robustness of the trees produced for these analyses.MethodologyOur focus was to increase the number of bootstrap replications that can be performed on large protein datasets using the maximum parsimony, distance matrix, and maximum likelihood methods. We have modified the PHYLIP package using MPI to enable large-scale phylogenetic study of protein sequences, using a statistically robust number of bootstrapped datasets, to be performed in a moderate amount of time. This paper discusses the methodology used to parallelize the PHYLIP programs and reports the performance of the parallel PHYLIP programs that are relevant to the study of protein evolution on several protein datasets.ConclusionsCalculations that currently take a few days on a state of the art desktop workstation are reduced to calculations that can be performed over lunchtime on a modern parallel computer. Of the three protein methods tested, the maximum likelihood method scales the best, followed by the distance method, and then the maximum parsimony method. However, the maximum likelihood method requires significant memory resources, which limits its application to more moderately sized protein datasets.

Interaction between continuous variables in logistic regression model

Rothman argued that interaction estimated as departure from additivity better reflected the biological interaction. In a logistic regression model, the product term reflects the interaction as departure from multiplicativity. So far, literature on estimating interaction regarding an additive scale using logistic regression was only focusing on two dichotomous factors. The objective of the present report was to provide a method to examine the interaction as departure from additivity between two continuous variables or between one continuous variable and one categorical variable. We used data from a lung cancer case-control study among males in Hong Kong as an example to illustrate the bootstrap re-sampling method for calculating the corresponding confidence intervals. Free software R (Version 2.8.1) was used to estimate interaction on the additive scale.

Prospective validation of P2/MS noninvasive index using complete blood counts for detecting oesophageal varices in B-viral cirrhosis

Periodic endoscopic screening for oesophageal varices (OVs) and prophylactic treatment for high-risk OVs (HOVs; medium/large OVs or small OVs plus red sign/decompensation) are currently recommended for all cirrhotic patients. However, if a simple, noninvasive test is available, many low-risk patients may reliably avoid endoscopy. We conducted a large-scale validation study of a simple, noninvasive test called P2/MS based on complete blood counts, (platelet count)(2)/[monocyte fraction (%) x segmented neutrophil fraction (%)], and compared it with other predictive tests for HOVs in B-viral cirrhotic patients. From 2008 to 2009, we prospectively enrolled 318 consecutive B-viral cirrhotic patients. All underwent endoscopy and laboratory evaluation. An area under the receiver operating characteristic curve of P2/MS was 0.941 for HOVs, comparable with those of the age-spleen platelet ratio index (0.922, P=0.317) and spleen-platelet ratio index (0.922, P=0.324), and better than those of age-platelet index (0.653, P<0.001), aspartate aminotransferase (AST)-platelet ratio index (0.871, P<0.006) and AST-alanine aminotransferase ratio (0.644, P<0.001). P2/MS<11 reliably identified 83 patients as having HOVs (94.0% positive predictive value), while at a cutoff of 25 and 179 as not having HOVs (94.4% negative predictive value). Overall, P2/MS reliably determined the likelihood of HOVs in 262 patients (82.4%). These cutoffs were validated internally using bootstrap resampling methods, which showed good agreement. P2/MS is a simple, accurate and economical method, reducing the need for endoscopy in B-viral cirrhosis. Patients with P2/MS<11 should be considered for appropriate prophylactic treatments, while those with P2/MS>25 may avoid endoscopy reliably.

Purported flaws in management strategy evaluation: basic problems or misinterpretations?

Abstract Butterworth, D. S., Bentley, N., De Oliveira, J. A. A., Donovan, G. P., Kell, L. T., Parma, A. M., Punt, A. E., Sainsbury, K. J., Smith, A. D. M., and Stokes, T. K. 2010. Purported flaws in management strategy evaluation: basic problems or misinterpretations? – ICES Journal of Marine Science, 67: 567–574. Rochet and Rice, while recognizing management strategy evaluation (MSE) as an important step forward in fisheries management, level a number of criticisms at its implementation. Some of their points are sound, such as the need for care in representing uncertainties and for thorough documentation of the process. However, others evidence important misunderstandings. Although the difficulties in estimating tail probabilities and risks, as discussed by Rochet and Rice, are well known, their arguments that Efron's non-parametric bootstrap re-sampling method underestimates the probabilities of low values are flawed. In any case, though, the focus of MSEs is primarily on comparing performance and robustness across alternative management procedures (MPs), rather than on estimating absolute levels of risk. Qualitative methods can augment MSE, but their limitations also need to be recognized. Intelligence certainly needs to play a role in fisheries management, but not at the level of tinkering in the provision of annual advice, which Rochet and Rice apparently advocate, inter alia because this runs the risk of advice following noise rather than signal. Instead, intelligence should come into play in the exercise of oversight through the process of multiannual reviews of MSE and associated MPs. A number of examples are given of the process of interaction with stakeholders which should characterize MSE.

Asymptotic confidence interval construction for proportion difference in medical studies with bilateral data

Bilateral dichotomous data are very common in modern medical comparative studies (e.g. comparison of two treatments in ophthalmologic, orthopaedic and otolaryngologic studies) in which information involving paired organs (e.g. eyes, ears and hips) is available from each subject. In this article, we study various confidence interval estimators for proportion difference based on Wald-type statistics, Fieller theorem, likelihood ratio statistic, score statistics and bootstrap resampling method under the dependence or/and independence models for bilateral binary data. Performance is evaluated with respect to the coverage probability and expected width via simulation studies. Our empirical results show that (1) ignoring the dependence feature of bilateral data could lead to severely incorrect coverage probabilities; and (2) Wald-type, score-type and bootstrap confidence intervals based on the dependence model perform satisfactorily for small to large sample sizes in the sense that their empirical coverage probabilities are close to the pre-specified nominal confidence level and are hence recommended. A real data from an otolaryngologic study is used to illustrate the proposed methods.

Validation of FIB-4 and comparison with other simple noninvasive indices for predicting liver fibrosis and cirrhosis in hepatitis B virus-infected patients

To optimize management and predict long-term clinical courses in patients with chronic hepatitis B (CHB), noninvasive tests to determine the degree of hepatic fibrosis have been developed. This study aimed to validate a simple, noninvasive FIB-4 index, which was first derived from an HCV-HIV-co-infected population, in patients with CHB and to compare it with other noninvasive tests for predicting cirrhosis. From 2006-2008, a total of 668 consecutive CHB patients who underwent liver biopsies were enrolled. The fibrosis stage was assessed according to the Batts and Ludwig system by a single pathologist blinded to patients' data. For prediction of significant (F > or = 2) and severe (F > or = 3) fibrosis, and cirrhosis (F = 4), the area under the receiver-operating characteristic curves were 0.865, 0.910 and 0.926 respectively. In predicting cirrhosis, it demonstrated diagnostic values comparable to the age-spleen platelet ratio index (0.937, P=0.414) and age-platelet index (0.928, P=0.888), and better outcomes than spleen-platelet ratio index (0.882, P=0.007), aspartate aminotransferase (AST)-platelet ratio index (0.731, P<0.001) and AST-alanine aminotransferase ratio index (0.730, P<0.001). FIB-4 cut-offs of 1.6 and 3.6 provided 93.2% negative predictive value and 90.8% positive predictive value for detection of cirrhosis respectively. Based on these results, liver biopsy could be avoided in 70.5% of the study population. These cut-offs were validated internally using bootstrap resampling methods, showing good agreement. FIB-4 is a simple, accurate and inexpensive method of predicting cirrhosis, with outcomes comparable to other noninvasive tests and may reduce the need for liver biopsy in the majority of CHB patients.

FURTHER EVALUATION OF BOOTSTRAP RESAMPLING AS A TOOL FOR RADIO-INTERFEROMETRIC IMAGING FIDELITY ASSESSMENT

We report on a broader evaluation of statistical bootstrap resampling methods as a tool for pixel-level calibration and imaging fidelity assessment in radio interferometry. Pixel-level imaging fidelity assessment is a challenging problem, important for the value it holds in robust scientific interpretation of interferometric images, enhancement of automated pipeline reduction systems needed to broaden the user community for these instruments, and understanding leadingedge direction-dependent calibration and imaging challenges for future telescopes such as the Square Kilometer Array. This new computational approach is now possible because of advances in statistical resampling for data with long-range dependence and the available performance of contemporary high-performance computing resources. We expand our earlier numerical evaluation to span a broader domain subset in simulated image fidelity and source brightness distribution morphologies. As before, we evaluate the statistical performance of the bootstrap resampling methods against direct Monte Carlo simulation. We find both model-based and subsample bootstrap methods to continue to show significant promise for the challenging problem of interferometric imaging fidelityassessment, when evaluated over the broader domain subset. We report on their measured statistical performance and guidelines for their use and application in practice. We also examine the performance of the underlying polarization self-calibration algorithm used in this study over a range of parallactic angle coverage.

External Validation of P2/MS and Comparison with Other Simple Non-invasive Indices for Predicting Liver Fibrosis in HBV-Infected Patients

To optimize management strategies and predict the long-term clinical course in patients with chronic hepatitis B (CHB), non-invasive tests to determine the degree of hepatic fibrosis have been developed. We aimed to conduct a large-scale external validation of a simple, non-invasive test called P2/MS using CHB patients and to compare it to other non-invasive tests for the prediction of histological cirrhosis. From 2006 to 2009, we enrolled a total of 521 consecutive CHB patients who underwent liver biopsy. Fibrosis stage was assessed according to the Metavir scoring system by a single pathologist who was unaware of the patients' histories. For predictions of significant (p>or=2) and severe (p>or=3) fibrosis and cirrhosis (p=4), the areas under the receiver operating characteristic curves were 0.801, 0.856, and 0.906, respectively. In predicting cirrhosis, we found that diagnostic values were comparable to age-spleen platelet ratio index (0.931, p=0.063), spleen-platelet ratio index (0.923, p=0.145), age-platelet index (0.914, p=0.670), and FIB-4 (0.898. p=0.597) and had better outcomes than the aspartate aminotransferase (AST)-platelet ratio index (0.780, p<0.001), and AST-alanine aminotransferase ratio index (0.729, p<0.001). The cut-off points of P2/MS>83 and P2/MS<30 provided 91.1% of negative predictive value and 91.3% of positive predictive value, respectively. Based on these results, liver biopsies could be avoided in 67.0% of the population. These cut-offs were validated internally using bootstrap resampling methods, which showed good agreement. P2/MS is a simple, accurate, and inexpensive method with comparable outcomes to other non-invasive tests and may reduce the need for liver biopsy in the majority of CHB patients.

Nonlinear process modeling of fructosyltransferase (FTase) using bootstrap re-sampling neural network model

Recently, the increased demand of fructooligosaccharides (FOS) as a functional food has alarmed researchers to screen and identify new strains capable of producing fructosyltransferase (FTase). FTase is the enzyme that converts the substrate (sucrose) to glucose and fructose. The characterization of complex sugar such as table sugar, brown sugar, molasses, etc. will be carried out and the sugar that contained the highest sucrose concentration will be selected as a substrate. Eight species of macro-fungi will be screened for its ability to produce FTase and only one strain with the highest FTase activity will be selected for further studies. In this work, neural networks (NN) have been chosen to model the process based on their excellent 'resume' in coping with nonlinear process. Bootstrap re-sampling method has been utilized in re-sampling the data in this work. This method has successfully modeled the process as shown in the results.

Medication adherence levels and differential use of mental-health services in the treatment of schizophrenia

BackgroundAdherence to antipsychotics for schizophrenia is associated with favorable clinical outcomes. This study compared annual mental-health service utilization by recent medication adherence levels for patients treated for schizophrenia, and assessed whether adherence levels change from pre- to post-psychiatric hospitalization.MethodsWe analyzed data from a large prospective, non-interventional study of patients treated for schizophrenia in the United States, conducted between 7/1997 and 9/2003. Detailed mental-health resource utilization was systematically abstracted from medical records and augmented with patients' self report. Medication possession ratio (MPR) with any antipsychotic in the 6 months prior to enrollment was used to categorize patients as: adherent (MPR ≥ 80%, N = 1758), partially adherent (MPR ≥ 60% < 80%, N = 36), or non-adherent (MPR < 60%, N = 216). Group comparisons employed propensity score-adjusted bootstrap re-sampling methods with 1000 iterations, adjusting for baseline patient demographic and clinical characteristics identified a priori.ResultsAdherent patients had a lower rate of psychiatric hospitalization compared with partially adherent and non-adherent patients (p < 0.001) and were more likely than non-adherent to engage in group therapy, individual therapy, and medication management. Most patients (92.0%) who were adherent in the 6 months prior to hospital admission continued to be adherent 6 months following hospitalization. However, 75.0% of previously partially adherent became adherent, and 38.7% of previously non-adherent became adherent following hospitalization.ConclusionAdherence is associated with lower utilization of acute care services and greater engagement in outpatient mental-health treatment. Adherence is a potentially dynamic phenomenon, which may improve, at least temporarily, following patients' psychiatric hospitalizations.

De novo identification of in vivo binding sites from ChIP-chip data

Poster presentation Advances in high-throughput technologies such as ChIPchip and the completion of human genomic sequences allow analysis of the mechanisms of gene regulation on a systems level. In this study, we have developed a computational genomics approach (ChIPModules) and a motif discovery approach (ChIPMotifs) to mine the ChIP-chip data. The ChIPModules approach begins with experimentally determined binding sites and integrates positional weight matrices, a comparative genomics approach, and statistical learning methods to identify transcriptional regulatory modules. Using E2F1 ChIP-chip data performed on ENCODE regions in both HeLa and MCF7 cells, we have identified five regulatory modules for E2F1. One of modules was validated by using ChIP-chip with arrays containing ~14,000 human promoters. The ChIPMotifs approach incorporates a bootstrap re-sampling method to statistically infer the optimal cutoff threshold for a position weight matrix (PWM) of a motif identified from ChIP-chip data by ab initio motif discovery programs. Using OCT4 ChIP-chip data, we developed an in vivo OCT4 PWM. We then used this PWM and our ChIPModules to identify transcription factors co-localizing with OCT4 in a testicular germ cell tumor (Ntera2 cells).

Hospitalization rates before and after initiation of paliperidone ER in patients with schizophrenia: results from open-label extensions of the US double-blind trials

ABSTRACTObjective: To assess differences in the number of days hospitalized among schizophrenic patients receiving paliperidone extended-release (paliperidone ER) during the open-label extension (OLE) phases, compared to a similar time period prior to the screening for entry into the double-blind (DB) trials conducted in the United States.Methods: Mental health-related hospital days during the 52 weeks before entering the DB trials and during the OLE phases were compared. The mean number of hospital days per person per year in the pre- and post-periods was calculated and the statistical significance of pre-post differences was assessed using bootstrap resampling methods. Total person-years were also calculated for the pre- and post-periods to account for different lengths of observation.Results: Patients' (n = 215) mean ( ± SD) age was 41.2 ( ± 11.0) years; most were male (73.0 % ); and black (52.1 vs. 45.1 % white). The mean ( ± SD) paliperidone ER treatment duration during the OLE phase was 167.0 ( ± 145.0) days and the mean ( ± SD) daily dose was 10.5 ( ± 2.0) mg. Overall, paliperidone ER patients spent an average ( ± SD) of 13.2 ( ± 1.6) and 3.1 ( ± 0.7) hospital days per person-year in the pre-and post-periods, respectively (mean ± SD change 10.0 ± 1.8, 95 % CI 6.5, 13.4, p < 0.001). Using the 2007 Federal Per Diem Base Rate (i.e., $595.09 per day), this reduction in hospital days would result in an average ( ± SD) cost savings of $5951 ( ± 1071) per person per year.Conclusions: Patients had significantly fewer hospital days in the OLE phase compared to the 1-year period prior to entering the DB trial. Paliperidone ER may play a role in reducing mental health-related hospital days and associated costs. Important study limitations include the lack of a control group, the pre-post design comparing historical data with data collected in the trials which could create a bias due to the mismatch in settings, and patients having more frequent contact with treating physicians and investigators during the trial period, which could favor the outcomes in the OLE phase. Further studies are needed to confirm these findings.

Clinical Significance and Risk Factors of Delayed-Onset Neutropenia (DON) after Autologous Stem Cell Transplantation (ASCT) for B-Cell Non-Hodgkin Lymphoma: A Single-Institute Analysis.

Background: DON is known as one of the adverse events after ASCT, for which rituximab usage has been reported to be a risk factor. One of the limitations in previous reports is that confounding by indication for rituximab treatment, and thus it is yet unclear if rituximab is an independent risk factor for DON. To examine the impact of rituximab considering such limitation as well as to explore possible other determinants and clinical significance of DON, we conducted a retrospective cohort study.Method: Subjects were consecutive 97 patients with B-cell non-Hodgkin lymphoma receiving ASCT between 1991 and 2007 in Aichi cancer center hospital. We defined DON as absolute neutrophil count < 0.1 x 109/L at any point after 30 days post ASCT without apparent cause of neutropenia. We investigated adverse events and risk factors using propensity score (PS), step-wise method and bootstrap resampling method. PS for preference of rituximab treatment was calculated based upon fourteen clinical factors; rituximab usage, age, sex, performance status, clinical stage, B symptom, bulky lesions, no. of extra nodal sites, lactate dehydrogenase level, no. of prior chemothearpy and local radiation, existence of bone marrow involvement, status at ASCT, and types of conditioning regimens (total-body-irradiation (TBI) containging regimens or not). Unconditional logistic regression models adjusted for PS and other potential risk factors were applied to estimates odds ratios (ORs) and their 95% confidence intervals (CIs) for occurrence of DON. Selection of potential risk factors in validation analysis was done by forward step-wise method (p-value for removal=0.10 and that for inclusion=0.05). Finally, we applied bootstrap resampling method for validation of our analyses (repeating 10,000 times with 97 sampling from original cohort).Results: 56.7% patients were treated with rituximab combined chemotherapy before ASCT, of which 26.3% had prior rituximab exposure. The median number of prior chemotherapy was one (range; one to five), and 47.4% patients received three or more prior regimens before ASCT. 22.7% patients received TBI regimen. With a median follow-up of 3.42 years, DON was observed in 46.7% patients. The lowest neutrophil count was 446 x 106/L (range; 9–992). Of the patients developing DON, 17.2% patients experienced febrile infectious events. Interstitial pneumonia (IP) was observed in 10% of the patients with DON and 5.5% without DON. No patients experienced Grade 4 or 5 events. By step-wise method, selected factors for final validation model were rituximab usage, TBI, female sex, B symptom, and prior radiation therapy. In the validation, usage of rituximab (OR=3.28: 1.07–10.0, p=0.037), and prior radiation before ASCT (1.81: 1.24–2.63, p=0.002) were identified as independent risk factors for DON. Other factors failed to demonstrate significant impact on DON.Conclusion: DON was associated with febrile infectious events in 17% patients and most of the cases achieved rapid clinical improvement. Our retrospective cohort analysis considering confounding by indication for rituximab confirmed significant impact of usage of rituximab on DON among B-cell lymphoma patients receiving autologous HSCT. Further evaluation of the impact of radiation therapy before HSCT on DON is also warranted.

Age and growth of big skate (Raja binoculata) and longnose skate (R. rhina) in the Gulf of Alaska

In 2003, big skates, Raja binoculata, and longnose skates, Raja rhina, were the target of a commercial fishery around Kodiak Island in the Gulf of Alaska (GOA) for the first time. The sudden development of a fishery for these species prompted the need for improved life history information to better inform fishery managers. Due to the selective nature of the skate fishery, mostly larger individuals were captured. Back-calculation from skate vertebral measurements was used to estimate size-at-age for younger skates. Because back-calculated age-length data within individuals were highly correlated, bootstrap resampling methods were used to test for differences between male and female growth curves. Results from bootstrapping indicated that differences between male and female growth were statistically significant for both species. This investigation indicates that growth of big skates in the GOA (max size 178 cm total length, max age 15 years) is similar to that in California, but different from that in British Columbia. For longnose skates, our GOA results agree with those reported in British Columbia, but were considerably older (max size 130 cm, max age 25 years) than those reported in California, which may not be surprising because longnose skates in the present study were generally larger. This life history information suggests that both big and longnose skates are at risk of unsustainable exploitation by targeted fisheries.

Identification of critical genes in microarray experiments by a Neuro-Fuzzy approach

Gene expression profiling by microarray technology is usually difficult to interpret into a simpler pattern. One approach to resolve the complexity of gene expression profiles is the application of artificial neural networks (ANNs). A potential difficulty in this strategy, however, is that the non-linear nature of ANN makes it essentially a ‘black-box’ computation process. Addition of a fuzzy logic approach is useful because it can complement ANN by explicitly specifying membership function during computation. We employed a hybrid approach of neural network and fuzzy logic to further analyze a published microarray study of gene responses to eight bacteria in human macrophages. The original analysis by hierarchical clustering found common gene responses to all bacteria but did not address individual responses. Our method allowed exploration of the gene response of the host to individual bacterium. We implemented a two-layer, feed-forward neural network containing the principle of ‘competitive learning’ (i.e. ‘winner-take-all’). The weights of the trained neural network were fed into a fuzzy logic inference system. A new measurement, called the impact rating (IR) was also introduced to explore the degree of importance of each gene. To assess the reliability of the IR value, a bootstrap re-sampling method was applied to the dataset and a confidence level for each IR was obtained. Our approach has successfully uncovered the unique features of host response to individual bacterium. Further, application of gene ontology (GO) annotation to the genes of high IR values in each response has suggested new biological pathways for individual host–pathogen interactions.

Impacts of irrigation on 20th century temperature in the northern Great Plains

Land use change can modify root zone moisture distribution, energy partitioning and subsequently, near surface energy balance. Various modeling studies provided evidence of these changes. For example, land use change from natural grass land to irrigated land use would significantly increase and decrease latent and sensible energy flux, respectively. This type of long-term modification of energy balance would in turn change near surface temperatures. The Great Plains of North America experienced significant overturning of land from natural grass land to irrigated land use during the 20th century. This study provides assessment on the changes in the historical near surface temperature records in Nebraska, USA. Long-term mean monthly maximum, minimum, and monthly mean air temperature data from 5 irrigated and 5 non-irrigated sites were analyzed. Length and homogeneity of time series and stability of stations were primary determinants in selection of these stations. The time series include Cooperative Weather Observation Network (COOP) and Historical Climate Network (HCN) data sets. Pairwise comparisons of temperatures between irrigated and non-irrigated locations for pre- and post-1945, -1950, and -1955 periods were completed for both data sets. These breakdowns of time series helped to identify periods of widespread land use change. Results show notably cooler temperatures over irrigated areas. For example, mean maximum growing season temperature at irrigated Alliance was 0.64 °C and 1.65 °C cooler compared to non-irrigated Halsey during pre- and post-1945 period, respectively. Hence, there was a 1.01 °C cooling during post-1945 years. Moreover, there has been a greater cooling during the second half of 20th century. The bootstrap re-sampling method was applied and trend analyses were completed for further verification of results. These assessments largely show a decreasing trend in mean maximum growing season temperatures over irrigated areas. To further verify the results and to determine the impacts of extreme values (including extremely cool temperatures), the 20% trimmed mean approach was applied. The impacts of extreme values have been minimal and based on the results obtained we conclude that land use change in the northern Great Plains has modified near surface temperature records.

Advances in testing the statistical significance of mediation effects.

P. A. Frazier, A. P. Tix, and K. E. Barron (2004) highlighted a normal theory method popularized by R. M. Baron and D. A. Kenny (1986) for testing the statistical significance of indirect effects (i.e., mediator variables) in multiple regression contexts. However, simulation studies suggest that this method lacks statistical power relative to some other approaches. The authors describe an alternative developed by P. E. Shrout and N. Bolger (2002) based on bootstrap resampling methods. An example and step-by-step guide for performing bootstrap mediation analyses are provided. The test of joint significance is also briefly described as an alternative to both the normal theory and bootstrap methods. The relative advantages and disadvantages of each approach in terms of precision in estimating confidence intervals of indirect effects, Type I error, and Type II error are discussed.