Justify your alpha

Facts and Fictions About Handling Multiple Comparisons.

Percutaneous nephrolithotomy (PCNL) is a well-established procedure for the management of urinary calculi and can be performed intercostally or subcostally. Favoring one approach vs the other is still debatable, and literature has been inconclusive regarding the efficacy and safety of both approaches. Hence, this study aims to assess the safety and efficacy of direct non-angled intercostal technique performed under full expiration and to compare it to the subcostal approach. PCNL was conducted among 361 patients during 2010-2015 at Saint George Hospital University Medical Center in Beirut, Lebanon. PCNL was done by one operator and by following a standard technique. After reviewing the medical records, 304 patients were included. Data analysis was conducted using Stata/IC 10.0. Bivariate analysis was conducted using Pearson's Chi-square, and logistic regression model was run. Alpha level was set at 0.05. Of the total patients, 54.6% and 45.4% underwent intercostal (Group I) and subcostal (Group II) access, respectively. Mean drop in hemoglobin in Group II was 1.9 g/dL vs 1.48 g/dL in Group I (p-value = 0.0040). The mean difference in operation time between group I (88.61 minutes) and group II (102.58 minutes) was statistically significant (p-value = 0.0064). Patients were stone free in 88.05% of the intercostal cases and 78.52% of the subcostal cases. Group II patients were twice more likely to have residual stones compared to Group I (p-value = 0.029). No statistical significance was observed in postoperative complications among both groups. In addition, no cases of pneumothorax were reported. Compared to subcostal access, intercostal approach under full expiration is a safe technique that provides optimal approach to the intrarenal collecting system and allows less angulation, less bleeding, and yields higher stone clearance with minimal complications. When performed by a well-trained urologist, intercostal access should be advocated in PCNL to obtain a direct non-angled access to the tip of the desired posterior calix.

COVID-19 related risk perception among taxi operators in Kingston and St. Andrew, Jamaica

ACL Research Retreat V

Teachers' perception of evaluation and testing in Nigerian secondary school system

Studies on play therapy in other disciplines have been done but post-operative pain management in pediatric patients has not been considered. This study determined the effects of play therapy using storytelling, bubble blowing, singing and film viewing on the pain management of post-operative pediatric patients. The single-group pretest-posttest design was employed. Post-operative pediatric patients of the orthopedic ward ages 2-7 years old assessed to have acute pain were chosen; seven out of 12 patients qualified and were given a pretest to assess the level of pain. Experimental treatment was administered; a posttest was administered thereafter. The level of pain was assessed before and after the intervention with the use of Wong-Baker FACES Pain Rating Scale, FLACC technique scale to assess the behavior and assessment of the vital signs. Play was used as the intervention and significant differences in the level of alleviated pain using self-assessment was computed by getting the decrement level of pain from pre to post assessment and subjecting it to the Kruskal-Wallis test. Storytelling was found to be the most effective resulting to lowered respiratory rate, heart rate and a normal blood pressure reading; however, it did not prove statistical significance as indicated by p-values greater than the 0.05 alpha level. Computed W coefficients associated with p-values revealed that pre-assessment data results were statistically different from the post-assessment data results in terms of self, behavioral and physiological assessments across the four types of play. The pain decrement from the pretest to posttest was statistically significant. A pain management program was proposed.

Mediterranean spotted fever (MSF) is a rickettsiosis endemic to Mediterranean countries caused by Rickettsia conorii, an obligate intracellular bacterium with endothelial cell tropism [1]. In a previous study we found an association between high plasma levels of tumor necrosis factor alpha (TNF-α) and severe MSF, raising the question of whether TNF-α is a causal contributor to disease pathogenesis or whether it is a nonspecific epiphenomenon [2]. In order to better characterize the role of proinflammatory cytokines in the pathogenesis of MSF we conducted the present study in which the plasma levels of interleukin-6 (IL-6) were analyzed and the variables significantly associated with disease severity were examined. The prospective study included 19 patients consecutively diagnosed with MSF at our hospital and 20 healthy controls. In all cases the diagnosis of MSF was confirmed by indirect immunofluorescence to R. conorii. Patients were considered to have severe MSF if they developed significant dysfunction of a major organ, as evidenced by neurological symptoms, or if they experienced a recent rise in serum creatinine levels, respiratory failure, or shock [3, 4]. Plasma IL-6 and TNF-α levels were determined using an enzyme-linked immunosorbent assay. The nonparametric Mann Whitney and Spearman’s correlation coefficient tests were used for the univariate analysis of clinical associations and analytical correlations, respectively. Multivariate analyses were performed to identify the variables significantly associated with disease severity in MSF patients. Plasma IL-6 levels were significantly higher in MSF patients on admission than in healthy controls (mean, 49.6 pg/ml [SD, 57.8], vs <4.0 pg/ml in all controls; P<0.001). Only two patients with active MSF had normal (undetectable) levels of IL-6. All patients had fever and rash, with the rash being purpuric in eight cases. Five patients developed severe MSF. Patients with severe MSF had higher plasma IL-6 levels than milder cases (mean, 97.1 pg/ml vs 32.7 pg/ml, respectively), but the difference was not statistically significant (P=0.10). Plasma IL-6 levels correlated significantly with plasma TNF-α (r=0.65, P=0.003) and serum C-reactive protein (r=0.49, P=0.03) levels, and inversely with the blood lymphocyte count (r= −0.76, P<0.001). In view of the association between severe MSF and increased levels of both IL-6 and TNF-α, we performed a stepwise logistic regression analysis to identify whether both cytokines contributed independently to the development of severe MSF. In this analysis, the only variable that approached statistical significance was plasma TNF-α (odds ratio of presenting severe MSF for each 10 pg/ml increase in TNF-α levels, 1.98; 95% confidence interval, 0.97–4.05; P=0.06), leaving plasma IL-6 outside the regression model. Several experimental studies give support to the hypothesis that proinflammatory cytokines might play a role in the pathogenesis of MSF. In vitro, activated macrophages and endothelial cells infected by R. conorii synthesize proinflammatory cytokines like TNF-α and IL6 [5, 6]. In addition, we reported high plasma levels of TNF-α and soluble TNF-α receptors in patients diagnosed with MSF, and the higher levels were associated with severe disease [2, 7]. In the present study, we also found high levels of IL-6 in plasma during the acute phase of MSF. As in many other inflammatory conditions, plasma IL-6 levels correlated J. Oristrell Internal Medicine Service, Corporacio Parc Tauli, C/ Parc Tauli s/n, 08208 Sabadell, Barcelona, Spain

Statistical significance tests are unnecessary even when properly done and properly interpreted: Reply to commentaries

A non-parametric statistical inference framework for Deep Learning in current neuroimaging

ABSTRACTThis study assessed whether a multi-ingredient energy supplement (MIES) could enhance cerebral-cortical activation and cognitive performance during an attention-switching task. Cerebral-cortical activation was recorded in 24 young adults (12 males, 12 females; 22.8 ± 3.8 yrs) via electroencephalography (EEG) both at rest and during the attention-switching task before (pretest) and 30 min after (posttest) consumption of a single serving of a MIES (MIES-1), two servings of a MIES (MIES-2), or a placebo (PL) in a double-blinded, randomized crossover experimental design. EEG upper-alpha power was assessed at rest and during the task, wherein d′ (Z[hit rate]–Z[false alarm rate]) and median reaction time (RT) for correct responses to targets on attention-hold and attention-switch trials were analyzed. For both d′ and RT, the Session (MIES-1, MIES-2, PL) × Time (pretest, posttest) interaction approached statistical significance (p = .07, η2p = 0.106). Exploring these interactions with linear contrasts, a significant linear effect of supplement dose on the linear effect of time was observed (ps ≤.034), suggesting the pretest-to-posttest improvement in sensitivity to task target stimuli (d′) and RT increased as a function of supplement dose. With respect to upper-alpha power, the Session × Time interaction was significant (p < .001, η2p = 0.422). Exploring this interaction with linear contrasts, a significant linear effect of supplement dose on the linear effect of time was observed (p < .001), suggesting pretest-to-posttest increases in cerebral-cortical activation were a function of supplement dose. In conclusion, our findings suggest that MIES can increase cerebral-cortical activation and RT during task performance while increasing sensitivity to target stimuli in a dose-dependent manner.

A testlet is a cluster of items that share a common passage, scenario, or other context. These items might measure something in common beyond the trait measured by the test as a whole; if so, the model for the item responses should allow for this testlet trait. But modeling testlet effects that are negligible makes the model unnecessarily complicated and risks capitalization on chance, increasing the error in parameter estimates. Checking each testlet to see if the items within the testlet share something beyond the primary trait could therefore be useful. This study included (a) comparison between a model with no testlets and a model with testlet g, (b) comparison between a model with all suspected testlets and a model with all suspected testlets except testlet g, and (c) a test of essential unidimensionality. Overall, Comparison b was most useful for detecting testlet effects. Model comparisons based on information criteria, specifically the sample-size adjusted Bayesian Information Criteria (SSA-BIC) and BIC, resulted in fewer false alarms than statistical significance tests. The test of essential unidimensionality had true hit rates and false alarm rates similar to the SSA-BIC when the testlet effect was zero for all testlets except the studied testlet. But the presence of additional testlet effects in the partitioning test led to higher false alarm rates for the test of essential unidimensionality.

Genome-wide association studies are carried out to identify unknown genes for a complex trait. Polymorphisms showing the most statistically significant associations are reported and followed up in subsequent confirmatory studies. In addition to the test of association, the statistical analysis provides point estimates of the relationship between the genotype and phenotype at each polymorphism, typically an odds ratio in case-control association studies. The statistical significance of the test and the estimator of the odds ratio are completely correlated. Selecting the most extreme statistics is equivalent to selecting the most extreme odds ratios. The value of the estimator, given the value of the statistical significance depends on the standard error of the estimator and the power of the study. This report shows that when power is low, estimates of the odds ratio from a genome-wide association study, or any large-scale association study, will be upwardly biased. Genome-wide association studies are often underpowered given the low alpha levels required to declare statistical significance and the small individual genetic effects known to characterize complex traits. Factors such as low allele frequency, inadequate sample size and weak genetic effects contribute to large standard errors in the odds ratio estimates, low power and upwardly biased odds ratios. Studies that have high power to detect an association with the true odds ratio will have little or no bias, regardless of the statistical significance threshold. The results have implications for the interpretation of genome-wide association analysis and the planning of subsequent confirmatory stages.

Are there benefits from NHST?

Objetivo: Demostrar que la exclusión de las altas con valores extremos de estancia (outliers) del sistema de clasificación de pacientes de los grupos diagnósticos relacionados para el cálculo de los indicadores clave de casuística&#13;\ny funcionamiento hospitalarios del Sistema Nacional de Salud carece de fundamento estadístico y es por tanto inadecuada.&#13;\nMétodos: Se ha aplicado una prueba de hipótesis con un nivel de confianza del 95% a cada uno de los 676 grupos diagnósticos relacionados que componen la versión All-Patient 23 para contrastar la distribución Normal de las estancias hospitalarias correspondientes a las 3 742 850 altas válidas del conjunto mínimo básico de datos del año 2009 en España. Además, se han calculado los coeficientes de asimetría y de curtosis de cada grupo diagnóstico relacionado.&#13;\nResultados: Sin tener en cuenta dos grupos que presentan un único registro, solamente un grupo diagnóstico no obtiene la significación estadística. Es decir, las estancias hospitalarias de la casi totalidad de los grupos diagnósticos relacionados no se distribuyen de forma Normal, para un error α de 0,05. Las causas de la no Normalidad son una asimetría con un marcado sesgo positivo y una curtosis con un pronunciado apuntamiento. Conclusiones: La detección y eliminación de los outliers se basan en la hipótesis de Normalidad de los datos. La distribución de las estancias hospitalarias no es Normal. En consecuencia, la exclusión de las altas con valores extremos de estancia de los grupos diagnósticos relacionados para la obtención de indicadores no es adecuada.

Experiments that find larger differences between groups than actually exist in the population are more likely to pass stringent tests of significance and be published than experiments that find smaller differences. Published measures of the magnitude of experimental effects will therefore tend to overestimate these effects. This bias was investigated as a function of sample size, actual population difference, and alpha level. The overestimation of experimental effects was found to be quite large with the commonly employed significance levels of 5 per cent and 1 per cent. Further, the recently recommended measure, ω 2 , was found to depend much more heavily on the alpha level employed than the true population ω 2 value. Hence, it was concluded that effect size estimation is impractical unless scientific journals drop the consideration of statistical significance as one of the criteria of publication.