Commentary: Understanding practical lot quality assurance sampling

We adapted a rapid quality of care monitoring method to a fragile state with two aims: to assess the delivery of child health services in South Sudan at the time of independence and to strengthen local capacity to perform regular rapid health facility assessments. Using a two-stage lot quality assurance sampling (LQAS) design, we conducted a national cross-sectional survey among 156 randomly selected health facilities in 10 states. In each of these facilities, we obtained information on a range of access, input, process and performance indicators during structured interviews and observations. Quality of care was poor with all states failing to achieve the 80% target for 14 of 19 indicators. For example, only 12% of facilities were classified as acceptable for their adequate utilisation by the population for sick-child consultations, 16% for staffing, 3% for having infection control supplies available and 0% for having all child care guidelines. Health worker performance was categorised as acceptable in only 6% of cases related to sick-child assessments, 38% related to medical treatment for the given diagnosis and 33% related to patient counselling on how to administer the prescribed drugs. Best performance was recorded for availability of in-service training and supervision, for seven and ten states, respectively. Despite ongoing instability, the Ministry of Health developed capacity to use LQAS for measuring quality of care nationally and state-by-state, which will support efficient and equitable resource allocation. Overall, our data revealed a desperate need for improving the quality of care in all states.

Hypothesis tests

BackgroundNeonatal infections caused by unsafe umbilical cord practices account for the majority of neonatal deaths in Nigeria. We examined the trends in umbilical cord care practices between 2012 and 2015 that coincided with the introduction of chlorhexidine digluconate 7.1% gel in Bauchi and Sokoto States.MethodsWe obtained data from three rounds of lot quality assurance samples (LQAS) surveys conducted in 2012, 2013 and 2015. Households were randomly sampled in each round that totaled 1140 and 1311 households in Bauchi and Sokoto States respectively. Mothers responded to questions on cord care practices in the last delivery. Coverage estimates of practice indicators were obtained for each survey period. Local Government Area (LGA) estimates for each indicator were obtained with α ≤ 5%, and β ≤20% statistical errors and aggregated to State-level estimates with finite sample correction relative to the LGA population.ResultsOver 75 and 80% of deliveries in Bauchi and Sokoto States respectively took place at home. The proportion of deliveries in public facilities reported by mothers ranged from 19% in 2012 to 22.4% in 2015 in Bauchi State and from 12.9 to 13.2% in 2015 in Sokoto State. Approximately 50% of deliveries in Bauchi and more than 80% in Sokoto States were assisted by traditional birth attendants (TBAs) or relatives and friends, with little change in the survey periods. In Bauchi and in Sokoto States, over 75% and over 80% of newborn cords were cut with razor blades underscoring the pervasive role of the TBAs in the immediate postpartum period. Use of chlorhexidine digluconate 7.1% gel for cord dressing significantly increased to the highest level in 2015 in both States. Health workers who attended deliveries in health facilities switched from methylated spirit to chlorhexidine. There were no observable changes in cord care practices among the TBAs.ConclusionUnsafe umbilical cord care practices remained prevalent in Bauchi and Sokoto States of Nigeria, although a recent introduction of chlorhexidine digluconate 7.1% gel positively changed the cord care practices toward safer practices among public health providers. TBAs, friends and relatives played the strongest immediate postpartum roles and mostly retained the unsafe cord care practices such as use of ash, cow dung and hot compress. We recommend that existing TBAs are retrained and refocused to forge stronger links between communities and the primary health centers to increase mothers’ access to skilled birth attendants.

Statistical reasoning in clinical trials: Hypothesis testing

Hypothesis Testing

Deformation monitoring usually focuses on the detection of whether the monitored objects satisfy the given properties (e.g. being stable or not), and makes further decisions to minimise the risks, for example, the consequences and costs in case of collapse of artificial objects and/or natural hazards. With this intention, a methodology relying on hypothesis testing and utility theory is reviewed in this paper. The main idea of utility theory is to judge each possible outcome with a utility value. The presented methodology makes it possible to minimise the risk of an individual monitoring project by considering the costs and consequences of overall possible situations within the decision process. It is not the danger that the monitored object may collapse that can be reduced. The risk (based on the utility values multiplied by the danger) can be described more appropriately and therefore more valuable decisions can be made. Especially, the opportunity for the measurement process to minimise the risk is an important key issue. In this paper, application of the methodology to two of the classical cases in hypothesis testing will be discussed in detail: 1) both probability density functions (pdfs) of tested objects under null and alternative hypotheses are known; 2) only the pdf under the null hypothesis is known and the alternative hypothesis is treated as the pure negation of the null hypothesis. Afterwards, a practical example in deformation monitoring is introduced and analysed. Additionally, the way in which the magnitudes of utility values (consequences of a decision) influence the decision will be considered and discussed at the end.

Bayesian designs are increasingly used to conduct phase II clinical trials. However, stopping boundaries in most Bayesian designs are defined from posterior credible intervals. The use of designs based on posterior credible intervals results in a loss of efficiency when compared to formal stopping rules based on Bayesian hypothesis tests. Such designs also introduce an unnecessary element of subjectivity in the interpretation of trial results. We derive a new class of Bayesian designs based on formal hypothesis tests. The prior densities used to define the alternative hypotheses in these tests assign no mass to parameter values that are consistent with the null hypotheses and are called nonlocal alternative prior densities. We show that Bayesian designs based on hypothesis tests and nonlocal alternative prior densities are more efficient than common Bayesian designs based on posterior credible intervals and common frequentist designs. In contrast to trial designs based on Bayesian credible intervals, we demonstrate that the mis-specification of the prior densities used to describe the anticipated effect of the experimental treatment in designs based on hypothesis tests cannot increase the expected weight of evidence in favor of the trial agent. Extension of test-based designs to phase I-II designs and randomized phase II designs remains an open research question. Phase II single-arm trials designed using Bayesian hypothesis tests with nonlocal alternatives provide better operating characteristics, use fewer patients per correct decision, and provide more directly interpretable results than other commonly used Bayesian and frequentist designs. Because the mis-specification of the prior density on the effect of the experimental agent decreases the expected weight of evidence that is collected in favor of the experimental treatment, the use of Bayesian hypothesis tests to design clinical trials also eliminates a potential source of bias often associated with trials designed using posterior credible intervals.

Are there benefits from NHST?

Studying nonlinear and potentially chaotic phenomena in geophysics from measured signals is problematic when system noise interferes with the dynamic processes that one is trying to infer. In such circumstances, a framework for statistical hypothesis testing is necessary but the nonlinear nature of the phenomena studied makes the formulation of standard hypothesis tests, such as analysis of variance, problematic as they are based on underlying linear, Gaussian assumptions. One approach to this problem is the method of surrogate data, which is the technique explained in this paper. In particular, we focus on (i) hypothesis testing for nonlinearity by generating linearized surrogates as a null hypothesis, (ii) a variant of this that is perhaps more appropriate for image data where structural nonlinearities are common and should be retained in the surrogates, and (iii) gradual reconstruction where we systematically constrain the surrogates until there is no significant difference between data and surrogates and use this to understand geophysical processes. In addition to time series of sunspot activity, solutions to the Lorenz equations, and spatial maps of enstrophy in a turbulent channel flow, two examples are considered in detail. The first concerns gradual wavelet reconstruction testing of the significance of a specific vortical flow structure from turbulence time series acquired at a point. In the second, the degree of nonlinearity in the spatial profiles of river curvature is shown to be affected by the occurrence of meander cutoff processes but in a more complex fashion than previously envisaged.

In his Forum, P. Vermeesch (Eos, 90(47), 443, doi:10.1029/2009EO470004, 2009) applied Pearson's chi‐square test to a large catalog of earthquakes to test the hypothesis that earthquakes are uniformly distributed across day of week (the formal null hypothesis that an earthquake has equal probability of occurring on any day). In his analysis, this hypothesis is rejected, and he proposes that the statistical test implies that earthquakes are correlated with day of the week (with specifically high seismicity on Sunday), and therefore strong dependence ofp values on sample size makes them uninterpretable. It is a well‐known property of classical hypothesis tests that the power of a statistical test is a function of the degrees of freedom, so that a test with large degrees of freedom will always have the resolution to reject the null. Consideration for practical as well as statistical significance is essential. Selecting bins so that the chi‐square test fails to reject the null hypothesis is essentially formulating a test to agree with a foregone conclusion. To the point, this data set does not exhibit uniform seismicity across time, and the statistical test is summarizing the data correctly. With proper attention to the application setting, and formulation of the null and alternative hypotheses, summarizing with p values is technically sound.

This article recommends a refinement in teaching hypothesis testing in introductory, non-calculus-based statistics courses, focusing on how the null hypothesis is formulated in mean and proportion inference. The proposal is grounded in classroom experience at CSU Fullerton, where general education courses enroll liberal arts and natural science students in lecture–lab formats. Most mainstream textbooks at this level (e.g., Peck, Gould, De Veaux) present the null as an equality for one-sided tests. Instead, we suggest adopting the Neyman–Pearson framework, commonly used in mathematical statistics textbooks, which treats the null and alternative as complementary hypotheses, leading to an inequality in the null in one-sided tests. Framing the hypotheses in this way is more logical to students, allows fundamental hypothesis testing concepts such as Type I error and significance level to be taught more precisely, and helps students recognize the uncertainties inherent in hypothesis testing rather than viewing it as a purely algorithmic process. To support instructor implementation, we provide a classroom activity accompanied by an interactive Shiny app that allows students to explore Type I error probabilities empirically and visually. This pedagogical framework is consistent with GAISE recommendations to teach statistical thinking, focus on conceptual understanding, and use technology to explore ideas.

CHAPTER 9 - Tests of Hypotheses

In many engineering disciplines the interesting model parameters are estimated from a large number of heterogeneous and redundant observations by a least-squares adjustment. The significance of the model parameters, outlier detection and the model selection itself are checked within statistical hypothesis tests. The acceptance and the rejection of the hypothesis are strongly related with two types of errors. A type I error occurs if the null hypothesis is rejected, although it is true. A type II error occurs if the null hypothesis is accepted, although it is false. This paper proposes a general procedure to hypothesis testing in linear parameter estimation, if the uncertainty is considered by random variability and interval/fuzzy errors. The study focuses on the probability of type I and type II errors. The applied procedure is outlined in detail showing both theory and numerical examples for the parameterisation of a geodetic monitoring network (deformation analysis).

Shopping streets at local city in Japan became old and are generally declining. In this paper, we handle the area rebirth and/or regional revitalization of shopping street. We focus on Fuji city in Japan. Four big festivals are held at Fuji city. Many people visit these festivals including residents in that area. Therefore a questionnaire investigation to the residents and visitors is conducted during these periods in order to clarify residents and visitors’ needs for the shopping street and utilize them to the plan building of the area rebirth and/or regional revitalization of shopping street. Hypothesis testing was executed based on that. We have set 9 Null hypotheses. In the hypothesis testing, 4 cases out of 9 null hypotheses were rejected and nearly half of hypotheses were insisted clearly. We have obtained fruitful results. To confirm the findings by utilizing the new consecutive visiting records would be the future works to be investigated.

Because the probability of obtaining an experimental finding given that the null hypothesis is true [p(F\H0)] is not the same as the probability that the null hypothesis is true given a finding [p(H0\F)], calculating the former probability does not justify conclusions about the latter one. As the standard null-hypothesis significance-testing procedure does just that, it is logically invalid (J. Cohen, 1994). Theoretically, Bayes's theorem yields p(H0\F), but in practice, researchers rarely know the correct values for 2 of the variables in the theorem. Nevertheless, by considering a wide range of possible values for the unknown variables, it is possible to calculate a range of theoretical values for p(H0\F) and to draw conclusions about both hypothesis testing and theory evaluation.