Formal Hypothesis Research Articles

SPIRITT (study 20060141): A randomized phase II study of FOLFIRI with either panitumumab (pmab) or bevacizumab (bev) as second-line treatment (tx) in patients (pts) with wild-type (WT) KRAS metastatic colorectal cancer (mCRC).

454 Background: Pmab has demonstrated significant improvement in progression-free survival (PFS) in pts with WT KRAS mCRC as 2nd-line tx in a phase III trial comparing pmab + FOLFIRI vs FOLFIRI alone. Here, we describe the results of SPIRITT, a multicenter, randomized phase II study evaluating pmab + FOLFIRI and bev + FOLFIRI in pts with WT KRAS mCRC previously treated with a 1st-line bev + oxaliplatin (Ox)-based chemotherapy regimen. Methods: Pts were randomized 1:1 to pmab 6.0 mg/kg + FOLFIRI Q2W or to bev 5.0 or 10.0 mg/kg + FOLFIRI Q2W. Eligibility criteria included: WT KRAS mCRC, ECOG ≤ 1, no prior irinotecan or anti-EGFR tx, and tx failure of prior 1st-line bev + Ox-based therapy (≥ 4 cycles). The primary endpoint was PFS; secondary endpoints included overall survival (OS), objective response rate (ORR), and safety. No formal hypothesis was tested. Results: 182 pts with WT KRAS mCRC were randomized. All pts received tx. Efficacy results are shown (table). Worst grade of 3/4 adverse events (AE) occurred in 78% of pts in the pmab + FOLFIRI arm and 65% in the bev + FOLFIRI arm. Grade 5 AEs occurred in 7% of pts in the pmab + FOLFIRI arm and 7% in the bev + FOLFIRI arm. Tx discontinuation due to any AE was 29% in the pmab + FOLFIRI arm and 25% in the bev + FOLFIRI arm. Conclusions: In this estimation study of pts with WT KRAS mCRC that previously received bev + Ox-based tx, the PFS hazard ratio (HR) was 1.01 (95% CI: 0.68 - 1.50). The OS HR was 1.06 (95% CI: 0.75 - 1.49). The observed ORR was higher in the pmab + FOLFIRI arm. 54% of bev + FOLFIRI pts received subsequent anti-EGFR tx. The safety profile for both arms was similar to previously reported studies. Tx discontinuation rates due to AEs were similar between the arms. Clinical trial information: NCT00418938. [Table: see text]

Decreased small-world functional network connectivity and clustering across resting state networks in schizophrenia: an fMRI classification tutorial

Functional network connectivity (FNC) is a method of analyzing the temporal relationship of anatomical brain components, comparing the synchronicity between patient groups or conditions. We use functional-connectivity measures between independent components to classify between Schizophrenia patients and healthy controls during resting-state. Connectivity is measured using a variety of graph-theoretic connectivity measures such as graph density, average path length, and small-worldness. The Schizophrenia patients showed significantly less clustering (transitivity) among components than healthy controls (p < 0.05, corrected) with networks less likely to be connected, and also showed lower small-world connectivity than healthy controls. Using only these connectivity measures, an SVM classifier (without parameter tuning) could discriminate between Schizophrenia patients and healthy controls with 65% accuracy, compared to 51% chance. This implies that the global functional connectivity between resting-state networks is altered in Schizophrenia, with networks more likely to be disconnected and behave dissimilarly for diseased patients. We present this research finding as a tutorial using the publicly available COBRE dataset of 146 Schizophrenia patients and healthy controls, provided as part of the 1000 Functional Connectomes Project. We demonstrate preprocessing, using independent component analysis (ICA) to nominate networks, computing graph-theoretic connectivity measures, and finally using these connectivity measures to either classify between patient groups or assess between-group differences using formal hypothesis testing. All necessary code is provided for both running command-line FSL preprocessing, and for computing all statistical measures and SVM classification within R. Collectively, this work presents not just findings of diminished FNC among resting-state networks in Schizophrenia, but also a practical connectivity tutorial.

Validation of Ground-Motion Simulations for Historical Events Using SDoF Systems

The study presented in this paper is among the first in a series of studies toward the engineering validation of the hybrid broadband ground-motion simulation methodology by Graves and Pitarka (2010). This paper provides a statis- tical comparison between seismic demands of single degree of freedom (SDoF) systems subjected to past events using simulations and actual recordings. A number of SDoF systems are selected considering the following: (1) 16 oscillation periods between 0.1 and 6 s; (2) elastic case and four nonlinearity levels, from mildly inelastic to severely inelastic systems; and (3) two hysteretic behaviors, in particular, nondegrading-nonevolutionary and degrading-evolutionary. Demand spectra are de- rived in terms of peak and cyclic response, as well as their statistics for four historical earthquakes: 1979 Mw 6.5 Imperial Valley, 1989 Mw 6.8 Loma Prieta, 1992 Mw 7.2 Landers, and 1994 Mw 6.7 Northridge. The results of this study show that both elastic and inelastic demands from simu- lated and recorded motions are generally similar. However, for some structural sys- tems, the inelastic response to simulated accelerograms may produce median demands that appear different from those obtained using corresponding recorded motions. The magnitude of such differences depends on the SDoF period, the nonlinearity level, and, to a lesser extent, the hysteretic model used. In the case of peak response, these dis- crepancies are likely due to differences in the spectral shape, while the differences in terms of cyclic response can be explained by some integral parameters of ground motion (i.e., duration-related). Moreover, the intraevent standard deviation values of structural demands calculated from the simulations are generally lower than those given by recorded ground motions, especially at short periods. The assessment of the results using formal statistical hypothesis tests indicates that, in most cases, the dif- ferences found are not significant, increasing the trust in the use of simulated motions for engineering applications.

Efficiency of study designs in diagnostic randomized clinical trials

From the patients' management perspective, a good diagnostic test should contribute to both reflecting the true disease status and improving clinical outcomes. The diagnostic randomized clinical trial is designed to combine both diagnostic tests and therapeutic interventions. Evaluation of diagnostic tests is carried out with therapeutic outcomes as the primary endpoint rather than test accuracy. We lay out the probability framework for evaluating such trials. We compare two commonly referred designs-the two-arm design and the paired design-in a formal statistical hypothesis testing setup and identify the causal connection between the two tests. The paired design is shown to be more efficient than the two-arm design. The efficiency gains vary depending on the discordant rates of test results. We derive sample size formulas for both binary and continuous endpoints. We derive estimation of important quantities under the paired design and also conduct simulation studies to verify the theoretical results. We illustrate the method with an example of designing a randomized study on preoperative staging of bladder cancer.

Author's Reply

Author's Reply

1016O - Phase 2, Randomized Trial (CONCERT-2) of Panitumumab (PMAB) Plus Radiotherapy (PRT) Compared with Chemoradiotherapy (CRT) in Patients (PTS) with Unresected, Locally Advanced Squamous Cell Carcinoma of the Head and Neck (LASCCHN)

ABSTRACT Background We evaluated the safety and efficacy of pmab, a fully human monoclonal antibody against the epidermal growth factor receptor, by comparing PRT with CRT in pts with LASCCHN. Methods Pts with stage III, IVA, or IVB previously untreated LASCCHN of all sites excluding the nasopharynx were randomized 2:3 to open-label CRT or PRT. CRT included 2 cycles of cisplatin 100 mg/m2 during accelerated fractionation radiotherapy (XRT; 70-72 Gy over 6-6.5 weeks). Pmab included 3 cycles at a dose of 9.0 mg/kg with each cycle administered with XRT. This was an estimation study with no formal hypothesis testing. The primary endpoint was locoregional control (LRC) rate at 2 years; key secondary endpoints were progression free survival (PFS), overall survival (OS), and safety. Preplanned HPV subset analysis (determined by p16 immunohistochemistry) was performed on available samples. Results Of 151 treated pts (90 PRT, 61 CRT), 84% were men; median age was 58 years; ECOG PS 0: 64%. Overall, the 2-year LRC rate (95% CI) was 51% (40%-62%) for PRT and 61% (47%-72%) for CRT. For both PFS (hazard ratio [HR] = 1.73 [95% CI: 1.07-2.81]; p = 0.03) and OS (HR = 1.59 [95% Cl: 0.91-2.79]; p = 0.10), outcomes favored the CRT arm. Of 99 pts with tumors evaluable for HPV, 24% were HPV+. In 75 HPV- tumors, a difference was observed in PFS (HR = 2.04 [95% Cl: 1.05-3.96]; p = 0.04), with trends also favoring CRT in LRC and OS. Overall, dose intensity was high for all components of therapy in both arms (median relative dose intensity: 100% for pmab and 99% for cisplatin, median XRT dose: 100% in both arms). Grade 3+ adverse events (AEs) occurred in 85% PRT vs 81% CRT pts. Key differences in grade 3+ toxicity between arms (PRT, CRT) included skin disorders (which is a composite of skin-related AEs) (35%, 3%), neutropenia (0%, 13%), and febrile neutropenia (0%, 8%). Conclusions Trends favored the CRT arm for the primary endpoint (LRC at 2 years), as well as other measures of efficacy, in this predominantly HPV- LASCCHN population. Small numbers limit conclusions in the HPV+ group. Both PRT and CRT appeared well tolerated. Disclosure G. Hatoum: Advisory Board Member for Amgen K. Oliner: Amgen stock A. Vanderwalde: Corporate-sponsored research (Amgen), full-time employee of Amgen All other authors have declared no conflicts of interest.

The Economics of the Food System Revolution

A revolution in food systems—food supply chains upstream from farms, to the food industry in the midstream segments of processing and wholesale and in the downstream segment of retail, then on to consumers—has been under way in the United States for more than a century and in developing countries for more than three decades. The transformation includes extensive consolidation, very rapid institutional and organizational change, and progressive modernization of the procurement system. In this article we examine the economics of these system-wide changes. We argue that the steps of conceptualizing and empirically researching this transformation—its patterns and trends, determinants, and impacts on farms and processing small and micro enterprises—are still in their infancy because of (a) remaining limitations on data suitable for formal modeling and hypothesis testing and (b) the sheer complexity of food system–related decisions that need to be modeled and understood. With the rapid accumulation of high-quality data now under way, conceptual and theoretical progress is also likely to be rapid.

SPATIAL UNCERTAINTY IN LINE-SURFACE INTERSECTIONS WITH APPLICATIONS TO PHOTOGRAMMETRY

Abstract. The fields of photogrammetry and computer vision routinely use line-surface intersections to determine the point where a line intersects with a surface. The object coordinates of the intersection point can be found using standard geometric and numeric algorithms, however expressing the spatial uncertainty at the intersection point may be challenging, especially when the surface morphology is complex. This paper describes an empirical method to characterize the unknown spatial uncertainty at the intersection point by propagating random errors in the stochastic model using repeated random sampling methods. These methods accommodate complex surface morphology and nonlinearities in the functional model, however the penalty is the resulting probability density function associated with the intersection point may be non-Gaussian in nature. A formal hypothesis test is presented to show that straightforward statistical inference tools are available whether the data is Gaussian or not. The hypothesis test determines whether the computed intersection point is consistent with an externally-derived known truth point. A numerical example demonstrates the approach in a photogrammetric setting with a single frame image and a gridded terrain elevation model. The results show that uncertainties produced by the proposed empirical method are intuitive and can be assessed with conventional methods found in textbook hypothesis testing.

Challenge with a hepatitis B vaccine in two cohorts of 4–7-year-old children primed with hexavalent vaccines: An open-label, randomised trial in Italy

Background and aimsThe anamnestic response to a challenge dose of vaccine can assess immune memory and protection against hepatitis B infection. This study investigated responses to a challenge dose of monovalent hepatitis B vaccine in children immunised with three doses of either Hexavac or Infanrix-Hexa during infancy. MethodsThis open-label, randomised, controlled, four-arm study enrolled 410 healthy children aged 4–7 years who had received either Hexavac (n=201) or Infanrix-Hexa (n=209) at 3, 5 and 11 months of life. Children received a single intramuscular challenge dose of either hepatitis B vaccine, HBVaxPro (Hexavac, n=34; Infanrix-Hexa, n=28) or Engerix-B (Hexavac, n=167; Infanrix-Hexa, n=181). Hepatitis B surface antibody (anti-HBs) concentrations were measured before and 1 month after the challenge vaccine dose. The analysis was descriptive and no formal hypothesis was tested. ResultsOne month post-challenge, 91.2% of children in the Hexavac group (95% confidence interval [CI] 86.3, 94.8) and 98.0% (95% CI 94.9, 99.4) in the Infanrix-Hexa group had anti-HBs concentrations ≥10mIU/ml (primary endpoint). In a post hoc analysis, most children with pre-challenge anti-HBs concentration <10mIU/ml achieved anti-HBs concentrations ≥10mIU/ml (Hexavac group, 85.3% [95% CI 77.6, 91.2]; Infanrix-Hexa group, 91.9% [95% CI 78.1, 98.3]). Both challenge vaccines were well tolerated. ConclusionsThese data suggest that immune memory persists for long-term (5 years) after a primary vaccination in infancy with a hexavalent vaccine (Hexavac or Infanrix-Hexa).

The Search for Order: Understanding Hindu-Muslim Violence in Post-Partition India

One distinguishing feature of mainstream social science is its growing regard for model building and formal hypothesis testing. In South Asian studies this is most evident in accounts of ethnic riots or communal violence. This paper examines a model of votes and violence proposed by Steven Wilkinson. We first examine how well the model performs against a data set that we have assembled on the twenty worst incidents of communal violence in India since 1950. The Wilkinson model is consistent with some important key facts in our data set, most notably in terms of levels of urbanization and percentage Muslims in riotaffected towns and cities. However, proximity to national or state elections is not found to be a strong driver of prolonged ethnic rioting. Nor is it the case that India's worst instances of communal violence occurred mainly where there was direct electoral competition between less than 3.5 effective political parties, the other main predictive variable in the Wilkinson model. We then discuss the limitations more broadly of attempts to explain and even predict ethnic violence within the framework of a quantitative model. We pay attention to time inconsistencies, principal-agent problems, religiosity and the homogenization of riot events, and omitted variables (notably, memory work and ideological fervour). We conclude with some general remarks on the search for order in social science.

Defining spatial comparison metrics for evaluation of paleoclimatic field reconstructions of the Common Era

Climate field reconstructions (CFR) of the Common Era (the last two millennia) provide important insights into the dynamics of past climate change that, in turn, have implications for the future. Multiple CFR methods have emerged in the literature, and comparisons between these methods using pseudoproxy experiments have been performed. These experiments, however, have not fully quantified the spatial skill of the CFRs, particularly with regard to the relative performance of each. Toward such ends, a formal statistical hypothesis test is proposed as a means of evaluating the differences between two random fields that integrate the differences in both the mean and the dependence structure. This involves a careful selection of the statistical model for the CFR residual process and systematic comparisons over different spatial scales. Application of this method yields a systematic assessment of the spatial character of five widely applied CFRs in a pseudoproxy experiment context. The analyses indicate that spatial differences among the five CFRs are not statistically significant. Further rigorous statistical assessments will help elucidate the strength and weakness of each CFR method, while quantifying the degree to which their spatial dissimilarities can be ascribed to methodological choices. Copyright © 2012 John Wiley &amp; Sons, Ltd.

On the Analysis of Balanced Two‐Level Factorial Whole‐Plot Saturated Split‐Plot Designs

This paper considers an experimentation strategy when resource constraints permit only a single design replicate per time interval and one or more design variables are hard to change. The experimental designs considered are two‐level full‐factorial or fractional‐factorial designs run as balanced split plots. These designs are common in practice and appropriate for fitting a main‐effects‐plus‐interactions model, while minimizing the number of times the whole‐plot treatment combination is changed. Depending on the postulated model, single replicates of these designs can result in the inability to estimate error at the whole‐plot level, suggesting that formal statistical hypothesis testing on the whole‐plot effects is not possible. We refer to these designs as balanced two‐level whole‐plot saturated split‐plot designs. In this paper, we show that, for these designs, it is appropriate to use ordinary least squares to analyze the subplot factor effects at the ‘intermittent’ stage of the experiments (i.e., after a single design replicate is run); however, formal inference on the whole‐plot effects may or may not be possible at this point. We exploit the sensitivity of ordinary least squares in detecting whole‐plot effects in a split‐plot design and propose a data‐based strategy for determining whether to run an additional replicate following the intermittent analysis or whether to simply reduce the model at the whole‐plot level to facilitate testing. The performance of the proposed strategy is assessed using Monte Carlo simulation. The method is then illustrated using wind tunnel test data obtained from a NASCAR Winston Cup Chevrolet Monte Carlo stock car. Copyright © 2012 John Wiley &amp; Sons, Ltd.

Constrained penalized splines

AbstractThe penalized spline is a popular method for function estimation when the assumption of “smoothness” is valid. In this paper, methods for estimation and inference are proposed using penalized splines under additional constraints of shape, such as monotonicity or convexity. The constrained penalized spline estimator is shown to have the same convergence rates as the corresponding unconstrained penalized spline, although in practice the squared error loss is typically smaller for the constrained versions. The penalty parameter may be chosen with generalized cross‐validation, which also provides a method for determining if the shape restrictions hold. The method is not a formal hypothesis test, but is shown to have nice large‐sample properties, and simulations show that it compares well with existing tests for monotonicity. Extensions to the partial linear model, the generalized regression model, and the varying coefficient model are given, and examples demonstrate the utility of the methods.The Canadian Journal of Statistics40: 190–206; 2012 © 2012 Statistical Society of Canada

Multiple testing in statistical analysis of systems-based information retrieval experiments

High-quality reusable test collections and formal statistical hypothesis testing together support a rigorous experimental environment for information retrieval research. But as Armstrong et al. [2009b] recently argued, global analysis of experiments suggests that there has actually been little real improvement in ad hoc retrieval effectiveness over time. We investigate this phenomenon in the context of simultaneous testing of many hypotheses using a fixed set of data. We argue that the most common approaches to significance testing ignore a great deal of information about the world. Taking into account even a fairly small amount of this information can lead to very different conclusions about systems than those that have appeared in published literature. We demonstrate how to model a set of IR experiments for analysis both mathematically and practically, and show that doing so can cause p -values from statistical hypothesis tests to increase by orders of magnitude. This has major consequences on the interpretation of experimental results using reusable test collections: it is very difficult to conclude that anything is significant once we have modeled many of the sources of randomness in experimental design and analysis.

Conducting Scientific Research: Research Hypothesis and Null Hypothesis

What about research hypothesis? A hypothesis is a specific statement of prediction. It describes in concrete (rather than theoretical) terms what you expect will happen in your study. Not all studies have hypotheses. Sometimes a study is designed to be exploratory. There is no formal hypothesis, and perhaps the purpose of the study is to explore some area more thoroughly in order to develop some specific hypothesis or prediction that can be tested in future research.

The Relationship between Tuberculosis and Influenza Death during the Influenza (H1N1) Pandemic from 1918-19

The epidemiological mechanisms behind the W-shaped age-specific influenza mortality during the Spanish influenza (H1N1) pandemic 1918-19 have yet to be fully clarified. The present study aimed to develop a formal hypothesis: tuberculosis (TB) was associated with the W-shaped influenza mortality from 1918-19. Three pieces of epidemiological information were assessed: (i) the epidemic records containing the age-specific numbers of cases and deaths of influenza from 1918-19, (ii) an outbreak record of influenza in a Swiss TB sanatorium during the pandemic, and (iii) the age-dependent TB mortality over time in the early 20th century. Analyzing the data (i), we found that the W-shaped pattern was not only seen in mortality but also in the age-specific case fatality ratio, suggesting the presence of underlying age-specific risk factor(s) of influenza death among young adults. From the data (ii), TB was shown to be associated with influenza death (P = 0.09), and there was no influenza death among non-TB controls. The data (iii) were analyzed by employing the age-period-cohort model, revealing harvesting effect in the period function of TB mortality shortly after the 1918-19 pandemic. These findings suggest that it is worthwhile to further explore the role of TB in characterizing the age-specific risk of influenza death.

Improved Eigenanalysis of Discrete Subpopulations and Admixture Using the Minimum Average Partial Test

Principal components analysis of genetic data has benefited from advances in random matrix theory. The Tracy-Widom distribution has been identified as the limiting distribution of the lead eigenvalue, enabling formal hypothesis testing of population structure. Additionally, a phase change exists between small and large eigenvalues, such that population divergence below a threshold of F_ST is impossible to detect and above which it is always detectable. I show that the plug-in estimate of the effective number of markers in the EIGENSOFT software often exceeds the rank of the sample covariance matrix, leading to a systematic overestimation of the number of significant principal components. I describe an alternative plug-in estimate that eliminates the problem. This improvement is not just an asymptotic result but is directly applicable to finite samples. The minimum average partial test, based on minimizing the average squared partial correlation between individuals, can detect population structure at smaller F_ST values than the corrected test. The minimum average partial test is applicable to both unadmixed and admixed samples, with arbitrary numbers of discrete subpopulations or parental populations, respectively. Application of the minimum average partial test to the 11 HapMap Phase III samples, comprising 8 unadmixed samples and 3 admixed samples, revealed 13 significant principal components.

Valuing Dealers' Informational Advantage: A Study of Canadian Treasury Auctions

In many financial markets, dealers have the advantage of observing the orders of their customers. To quantify the economic benefit that dealers derive from this advantage, we study detailed data from Canadian Treasury auctions, where dealers observe customer bids while preparing their own bids. In this setting, dealers can use information on customer bids to learn about (i) competition, that is, the distribution of competing bids in the auction, and (ii) fundamentals, that is, the ex post value of the security being auctioned. We devise formal hypothesis tests for both sources of informational advantage. In our data, we do not find evidence that dealers are learning about fundamentals. We find that the “information about competition” contained in customer bids accounts for 13–27% of dealers' expected profits.

Evaluating the Performances of Software Cost Estimation Models through Prediction Intervals

The task of predicting accurately the cost required for the completion of a new software project is a challenging issue in the Software Cost Estimation area, since it is closely related with the activities of project management and the wise decision-making of organizations in order to bid, plan and budget a forthcoming system. However, the accurate prediction of the cost is often obtained with great uncertainty and for this reason there has been noted a lack of convergence in experimental studies. The main reason for the discrepancy can be derived from the inherent characteristic of prediction methodologies, since they produce point estimates without taking into account the risk covering the whole process. In this study, we propose a statistical framework, so as to focus on the construction of Prediction Intervals which provide an “optimistic” and a “pessimistic” guess for the true magnitude of the cost. The proposed framework that incorporates different accuracy indicators, formal hypothesis testing and graphical inspection of the predictive performance is applied on a dataset with real software projects.

MetaQC: objective quality control and inclusion/exclusion criteria for genomic meta-analysis

Genomic meta-analysis to combine relevant and homogeneous studies has been widely applied, but the quality control (QC) and objective inclusion/exclusion criteria have been largely overlooked. Currently, the inclusion/exclusion criteria mostly depend on ad-hoc expert opinion or naïve threshold by sample size or platform. There are pressing needs to develop a systematic QC methodology as the decision of study inclusion greatly impacts the final meta-analysis outcome. In this article, we propose six quantitative quality control measures, covering internal homogeneity of coexpression structure among studies, external consistency of coexpression pattern with pathway database, and accuracy and consistency of differentially expressed gene detection or enriched pathway identification. Each quality control index is defined as the minus log transformed P values from formal hypothesis testing. Principal component analysis biplots and a standardized mean rank are applied to assist visualization and decision. We applied the proposed method to 4 large-scale examples, combining 7 brain cancer, 9 prostate cancer, 8 idiopathic pulmonary fibrosis and 17 major depressive disorder studies, respectively. The identified problematic studies were further scrutinized for potential technical or biological causes of their lower quality to determine their exclusion from meta-analysis. The application and simulation results concluded a systematic quality assessment framework for genomic meta-analysis.