Small Sample Bias Research Articles

News Shocks across Countries: An Empirical Investigation

We estimate the role of news shocks to total factor productivity, foreign interest rates and commodity terms of trade in explaining the variance of output and other macro aggregates in a large sample of countries. To correct for the small-sample bias of the variance decomposition estimates we develop a Bootstrap-after-Bootstrap method. We find that the mean difference of variance share of output explained by news shocks between developing and developed countries is: I) Negligible for news shocks to total factor productivity. II) Positive for news shocks to foreign interest rates (6 p.p.) and to commodity terms of trade (8.3 p.p.). Using cross-sectional data, we find that countries with less financial development have a larger share of output variance explained by news shocks to foreign interest rates, and countries with higher total trade of commodities to output ratio and less developed financial markets exhibit a larger share of output variance explained by news shocks to commodity terms of trade. These results suggest that to study the role of news shocks in the economy, one-sector models with only shocks to total factor productivity are not adequate, and that there must be a structural distinction regarding financial markets' development when modeling developing countries as opposed to developed in a general equilibrium framework.

Conditional Fixed Income Correlations: Skews and Straddles

We apply extreme value and copula theory to estimate conditional correlations of various pairs of US fixed-income assets. This framework is applied to both comovements of fixed-income assets, and cross-movements of fixed-income assets where one, typically flight-to-quality, asset strongly outperforms and the other, typically a risky spread product, strongly underperforms. Our results show that for comovements between risky spread products the correlation structure resembles a straddle where conditional correlations monotonically increase both with increased market stress and with increased market rallies. This finding stands in contrast to the correlation patterns found in international equity markets and hedge fund strategies for which the correlation structure has the shape of a skew with increased correlations under extreme market stress and zero correlations during risk-on episodes. We also find this skew shape in the correlations between US fixed-income assets when we estimate conditional correlations for cross-movements with a flight-to-quality asset performing strongly during periods of stress and risky spread products considerably widen. Under these conditions, conditional correlations are very negative, but monotonically fade with market recoveries to eventually become zero during risk-on episodes. TOPICS:Fixed-income portfolio management, tail risks, statistical methods Key Findings • This study presents the first analysis of extreme value and copula theory applied to correlations estimations of various pairs of fixed-income assets conditioned on market sentiment. This method is immune to small sample biases present in subsampling methods. • Our results show for comovements between risky spread products that the correlation structure resembles a straddle where conditional correlations monotonically increase with increasing market stress and increasing market rallies. This finding is novel and contrasts patterns observed in equity markets. • We find a skew shape in conditional correlations in fixed-income markets between flight-to-quality assets and risky spread products in cross-movements. Conditional correlations are very negative during periods of stress but monotonically fade with market recoveries.

Evaluation of inferential methods for the net benefit and win ratio statistics

ABSTRACT General Pairwise Comparison (GPC) statistics, such as the net benefit and the win ratio, have been applied in clinical trial data analysis and design. In the literature, inferential methods based on re-sampling, asymptotic or exact methods have been proposed for these GPC statistics, but they have not been compared to each other. In this paper, the small sample bias of the variance estimation, Type I error control and 95% confidence interval coverage of the GPC inferential methods are evaluated using simulations. The exact permutation and bootstrap tests perform best in all evaluated aspects for the net benefit, while the exact bootstrap test performs best for the win ratio.

Finasteride and Suicide: A Postmarketing Case Series

Background: In 2011, depression was added to the product labeling of finasteride in the USA. The US Food and Drug Administration’s Adverse Event Reporting System database contains at least 36 death cases for finasteride. The aim of this study is to characterize the clinical histories and symptoms reported by a series of 6 suicide victims who took finasteride for treatment of androgenic alopecia. Methods: Medical records and autopsy reports were provided by family members of the cases. Relevant information was extracted according to guidelines for submitting adverse event reports. Results: An important pattern of symptoms was common among all cases who committed suicide in the setting of finasteride use – insomnia and persistent sexual dysfunction after medication discontinuation. Insomnia and fatigue/tiredness were some of the most debilitating symptoms. Apart from 1 case who had hyperlipidemia, there was no documentation of concomitant medication use with finasteride or any baseline medical or psychiatric diagnoses prior to starting finasteride. The findings of this postmarketing series may not be generalizable to the population of men who committed suicide in the setting of finasteride use due to small sample size and bias. Associations between medication use and symptoms cannot prove causality. Conclusion: Men under the age of 40 who use finasteride for alopecia are at risk for suicide if they develop persistent sexual adverse effects and insomnia. Further research is needed to establish whether finasteride has a causal relationship to suicide.

Bias-corrected estimators for the Vasicek model: an application in risk measure estimation

We evaluate the usefulness of bias-correction methods in enhancing the Vasicek model for market risk and counterparty risk management practices. The naive bias-corrected estimator, the Tang and Chen bias-corrected estimator and the Bao et al bias-corrected estimator are selected to be compared against the benchmark least squares (LS) estimator. Our Monte Carlo experiment shows that the bias-corrected estimators substantially reduce the small sample bias of the LS estimator for the Vasicek model and project much more accurate value-at-risk and potential future exposure estimations. Even if the sample length is as long as 30 years, the improvements are still significant, especially for the cases where the mean-reversion parameter is close to zero. The applications to real data further demonstrate that the small sample bias of the LS estimator cannot be ignored and one should consider bias-corrected estimators for the Vasicek model.

Biases in Long-Horizon Predictive Regressions

Analogous to Stambaugh (1999), this paper derives the small sample bias of estimators in J-horizon predictive regressions, providing a plug-in adjustment for these estimators. A number of surprising results emerge, including (i) a higher bias for overlapping than nonoverlapping regressions despite the greater number of observations, and (ii) particularly higher bias for an alternative long-horizon predictive regression commonly advocated for in the literature. For large J, the bias is linear in (J/T) with a slope that depends on the predictive variable’s persistence. The bias adjustment substantially reduces the existing magnitude of long-horizon estimates of predictability. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Mining for Oil Forecasts

We study the performance of many traditional and novel, text-based variables for in-sample and out-of-sample forecasting of oil spot, futures, and energy company stock returns, and changes in oil volatility, production, and inventories. After controlling for small-sample biases, we find evidence of in-sample predictability. Our text measures, derived using energy news articles, hold their own against traditional variables. While we cannot identify ex-ante rules for selecting successful out-of-sample forecasters, an analysis of all possible two-variable models reveals out-of-sample performance above that expected under random variation. Our findings provide new directions for identifying robust forecasting models for oil markets, and beyond.

An estimation of the prevalence of occult HBV infection in Western Europe and in Northern America: A meta-analysis.

Data on the prevalence of occult HBV infection (OBI) in Western Europe and in Northern America are few; hence, we conducted a systematic review and meta-analysis. All studies included had to fulfil the following inclusion criteria: (a) they investigated the prevalence of OBI (HBV DNA in liver tissue in HBsAg-negative subjects), (b) were carried out in Western Europe and in Northern America; (c) were available as a full-text manuscript, (d) written in English and (e) published up to December 2018. The exclusion criteria were as follows: (a) meta-analyses, letters, reviews, meeting abstracts or editorial comments; (b) studies investigating HBsAg-positive patients; (c) those investigating OBI outside Western Europe and in Northern America; and (d) to avoid small sample bias in the random-effects model, those enrolling less than five subjects. Thirty-four studies met the inclusion criteria, allowing a meta-analysis on 2729 patients. The overall prevalence of OBI was 34% (95% CI=26%-42%), 28% (CI 95%: 12%-48%) in 329 subjects without chronic liver disease and 35% (95% CI 26%-44%) in 2400 patients with chronic liver disease. The prevalence of OBI was 51% (95% CI 40%-62%) in the 823 anti-HBc-positive subjects and 19% (95% CI 10%-30%) in the 1,041 anti-HBc-negative subjects. Evaluating the data from 17 studies comparing anti-HBc-positive and negative subjects, the prevalence of OBI was higher in the 641 anti-HBc-positive subjects than in the 1041 anti-HBc-negative (prevalence ratio=2.29; 95% CI=1.61-3.26, P<.001). This meta-analysis showed that in HBsAg-negative subjects the prevalence of OBI was high and was associated with anti-HBc positivity.

Revisiting the persistence of real exchange rates

In this paper we provide new and robust evidence that panel regressions significantly reduce the estimated half-lives of real exchange rates in a sample of 21 countries. These results are obtained under dynamic panel models that control for non-purchasing power parity (PPP) equilibria. Small-sample bias does not have a significant impact on half-lives, since the increase in the half-life arising from the bias is more than compensated by the decrease of the half-life due to the control of non-PPP equilibria. Among the two sources of non-PPP equilibria, i.e., first, the Balassa-Samuelson effect and second, smooth shifts in mean, we find that, under a heterogeneous panel model, the first source results in a larger reduction of half-lives.

Challenges in using RCTs for evaluation of large-scale public programs with complex designs: Lessons from Peru

Abstract The use of randomized control trials (RCTs) to evaluate public policies and interventions in developing countries faces several challenges. These include limited budgets to finance sample designs and sample sizes required to evaluate multifaceted interventions, potential small-sample bias arising from such limited samples, and difficulties in random assignment when participants self-exclude from parts of the intervention. In addition, institutional challenges arise when seeking to evaluate large-scale interventions implemented within a state bureaucracy as compared to NGO small pilots’ evaluations. This short article seeks to discuss the practical challenges facing RCTs when used as a public policy and program evaluation mechanism. This discussion is based on the impact evaluation of a public project that offered several productive interventions to rural households who were already receiving conditional cash transfers.

What works in smoking cessation interventions for cancer survivors? A meta-analysis.

We conducted a meta-analysis of randomized controlled trials designed to promote smoking cessation among cancer survivors to (a) assess how effective interventions are at increasing quit rates, and (b) determine which intervention strategies are associated with effect sizes. Out of 10,848 records that were located using computerized searches and informal sources, 21 interventions met the inclusion criteria for the review. We developed a bespoke taxonomy of 36 categories of techniques designed to change smoking behavior, and coded sample, intervention, and methodological characteristics. Random effects meta-analysis and metaregressions were conducted. The sample-weighted average effect size for smoking cessation was d+ = .030, and was not significantly different from zero (95%CI = -.042 to .101). Effect sizes exhibited both publication bias and small sample bias. Metaregressions indicated that, out of the many potential moderators that were tested, just a single intervention feature was associated with effect sizes. Interventions delivered solely by nurses exhibited larger effects compared to interventions from other sources. The present review indicates that current smoking cessation interventions for cancer survivors are ineffective. High-quality and effective interventions are needed. We offer suggestions regarding promising intervention strategies. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Multistep forecast selection for panel data

We develop a new set of model selection methods for direct multistep forecasting of panel data vector autoregressive processes. Model selection is based on minimizing the estimated multistep quadratic forecast risk among candidate models. To attenuate the small sample bias of the least squares estimator, models are fitted using bias-corrected least squares. We provide conditions sufficient for the new selection criteria to be asymptotically efficient as n (cross sections) and T (time series) approach infinity. The new criteria outperform alternative selection methods in an empirical application to forecasting metropolitan statistical area population growth in the US.

An exploration of stroke survivors’ perspectives on cycling and the use of electric bikes

Background New and innovative approaches are needed to overcome the barriers to engaging people in physical and leisure activity after stroke. Outdoor cycling, including the use of adapted or electric bicycles, may be one approach. However, perceptions of stroke survivors on this topic have not yet been explored. Purpose To explore a sample of stroke survivors’ perspectives, who expressed an interest in cycling, about cycling and the use of electric bicycles. Methods A convenience sample of stroke survivors were identified through focus groups at a ‘Cycling after Stroke’ event, local stroke support groups, and structured interviews at a national conference for stroke survivors. Quantitative data were analysed descriptively and qualitative data analysed thematically. Results Data were collected from 21 stroke survivors, seven of whom were current cyclists. All participants were independently mobile with, or without, the use of a walking aid. Themes oriented around the value of cycling (e.g. getting out of the house, doing something for yourself, and feeling part of a community); concerns and challenges (safety and negotiating adaptations); and how they could be overcome (starting slowly and identifying sources of assistance). Conclusion Outdoor cycling may be a worthwhile approach to increasing physical and leisure activity after stroke. However, barriers still exist and need to be addressed to provide inclusive opportunities for adapted and electric cycling for stroke survivors. Due to the small sample size and bias population, further research is needed to explore stroke survivors’ perspectives on cycling to provide solutions to overcome the barriers identified. 181

Lung Density Changes Following SBRT: Association with Lung Dose and Clinical Pneumonitis

Radiation pneumonitis is an inflammation of the lungs that can develop following radiation therapy, the cause of which is not fully understood. The hypothesis of this study is that radiation induced lung changes are associated with radiation dose to the lung and pneumonitis. This work investigates the relationship between lung dose and the development of radiographic lung changes following stereotactic body radiotherapy (SBRT). In addition, the lung dose and radiographic changes were investigated for their ability to predict the development of clinical grade ≥2 radiation pneumonitis (RP≥2). Regions of consolidation (dense) and ground-glass opacities (hazy) were semi-automatically contoured on 262 3- and 6-month post treatment CT-scans for 145 subjects who underwent SBRT between 2008 and 2018 receiving 40-60 Gy in 3-10 fractions. 84% of the subjects received 48Gy in 4 fractions. RP≥2 was diagnosed in 9 patients (6%). The PTV volume, relative and absolute lung volumes irradiated within defined dose bins were extracted from the treatment plans. Volumes of dense and hazy radiographic changes, both absolute volumes in mL and volumes normalized by the PTV volume were assessed at 3- and 6-months. Correlations for all 3- and 6-month hazy and dense volumes with PTV volumes, relative and absolute lung volumes in defined dose bins, and treatment technique (3D conformal radiotherapy (3DCRT), static intensity modulated radiotherapy (IMRT), or volumetrically modulated arc therapy (VMAT)) were established through Spearman correlation using Bonferroni correction. Univariate logistic models, using the Firth small sample bias correction, were developed for predicting RP≥2. PTV volume and relative lung dose bins between 25 and 40 Gy were significantly correlated with the hazy and dense volumes on both the 3 and 6 month images (p=<0.0001-0.0019). There was no significant difference in the hazy or dense volumes at either 3- or 6-months by the treatment technique. Subjects developing RP≥2 on average demonstrated a 2- to 4-fold increase in the hazy and dense absolute and normalized volumes at both 3- and 6-months (n.s.). The 3-month dense (p=0.03), 6-month dense (p=0.02), 6-month dense normalized (p=0.04), and 6-month hazy volumes (p=0.03) were significantly correlated with RP≥2 in univariate logistic models (AUC 0.64-0.67). Lung V20-40Gyappears to be correlated with radiographic changes following lung SBRT. While irradiated lung volumes were not correlated with RP≥2, the amount of dense volume present at 3 and 6 months in particular appears to be related to the development of RP≥2. This observation indicates that in addition to dose, demographic and clinical factors may influence the development of RP and need to be further investigated.

Ingroup role models and underrepresented students’ performance and interest in STEM: A meta-analysis of lab and field studies

This meta-analysis synthesizes research on using ingroup role models to improve the performance and interest of underrepresented students in science, technology, engineering, and math (STEM). A systematic literature search resulted in forty-five studies that met the selection criteria, including the presence of a comparison group. Both lab and field studies suffered from small sample bias, with smaller sample sizes predicting larger effect sizes among lab studies, but smaller effect sizes among field studies. Correcting for small sample bias, ingroup role models had a small, but significant positive overall effect (d = 0.20) among field studies and a non-significant overall effect (d = 0.04) among lab studies. The only significant moderator was level of interaction, with in-person role models having smaller effects among lab studies (p = .008). Implications for interventions to increase the representation of female and underrepresented minority students in STEM and future directions for research are discussed.

Bias reduction in nonlinear and dynamic panels in the presence of cross-section dependence

Fixed effects estimation of nonlinear dynamic panel models is subject to the incidental parameter issue, leading to a biased asymptotic distribution. While this problem has been studied extensively in the literature, a general analysis allowing for both serial and cross-sectional dependence is missing. In this paper we investigate the large-N,T theory of the profile and integrated likelihood estimators, allowing for dependence across both dimensions. We show that under stronger dependence types the asymptotic bias disappears, but a Op(1∕T) small-sample bias remains. We provide bias correction and inference methods, and also obtain primitive conditions for asymptotic normality under various dependence settings.

US Risk Premia under Emerging Markets Constraints

USA market is the benchmark for empirical finance and considered the closest example of how an efficient market should behave. On the other hand, divergent results from the observed in the USA are often associated with unreliable and due deviations from efficient hypothesis. However, how would the US market results behave had the data the same constraints as an emerging market economy? To answer that question we analyze the risk premia market estimation under the typical constraints from emerging equity markets: the small number of assets and the short time-series sample available for estimation. We use parameters of time-series length, number of assets and accounting variables distribution from the Brazilian equity market. Surprisingly, we conclude that the US market risk premia convey the same data features as the Brazilian risk premia if under the same time constraints. Then, we evaluate two potential causes of problems in risk premia estimations with small T: i) small sample bias on betas, and ii) divergence between ex-post and ex-ante risk premia. Through Monte Carlo simulations, we conclude that for the T around 5 years the beta estimates are no longer a problem. However, it is necessary to analyze a time-series sample exceeding 40 years to obtain robust ex-ante risk premia.

BALLI: Bartlett-adjusted likelihood-based linear model approach for identifying differentially expressed genes with RNA-seq data

BackgroundTranscriptomic profiles can improve our understanding of the phenotypic molecular basis of biological research, and many statistical methods have been proposed to identify differentially expressed genes (DEGs) under two or more conditions with RNA-seq data. However, statistical analyses with RNA-seq data are often limited by small sample sizes, and global variance estimates of RNA expression levels have been utilized as prior distributions for gene-specific variance estimates, making it difficult to generalize the methods to more complicated settings. We herein proposed a Bartlett-Adjusted Likelihood-based LInear mixed model approach (BALLI) to analyze more complicated RNA-seq data. The proposed method estimates the technical and biological variances with a linear mixed-effects model, with and without adjusting small sample bias using Bartlkett’s corrections.ResultsWe conducted extensive simulations to compare the performance of BALLI with those of existing approaches (edgeR, DESeq2, and voom). Results from the simulation studies showed that BALLI correctly controlled the type-1 error rates at various nominal significance levels and produced better statistical power and precision estimates than those of other competing methods in various scenarios. Furthermore, BALLI was robust to variation of library size. It was also successfully applied to Holstein milk yield data, illustrating its practical value.Conclusions;BALLI is statistically more efficient and valid than existing methods, and we conclude that it is useful for identifying DEGs in RNA-seq analysis.

Effect of Postactivation Potentiation on Explosive Vertical Jump: A Systematic Review and Meta-Analysis.

Dobbs, WC, Tolusso, DV, Fedewa, MV, and Esco, MR. Effect of postactivation potentiation on explosive vertical jump: a systematic review and meta-analysis. J Strength Cond Res 33(7): 2009-2018, 2019-The primary aim of this systematic review and meta-analysis was to quantify the magnitude of the effect of postactivation potentiation (PAP) on explosive vertical power while accounting for the nesting of multiple effects within each study. This study was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Statement (PRISMA). Original research articles published by January 18, 2018, were located using an electronic search of 4 databases and yielded 759 original publications. Data were extracted and independently coded by 2 authors (W.C.D. and D.V.T.). The standardized mean effect size (ES) was calculated by subtracting the pre-treatment mean from the post-treatment mean and dividing by the pooled SD, adjusting for small sample bias. Multilevel random-effects model was used to aggregate a mean ES and 95% confidence interval (CI) for PAP on vertical jump performance. The cumulative results of 179 effects gathered from 36 studies indicate that PAP does not improve vertical jump performance (ES = 0.08, 95% CI -0.04 to 0.21, p = 0.197), with moderate heterogeneity. Moderator analysis indicated that rest intervals between 3 and 7 minutes provided favorable performance outcomes (ES = 0.18, 95% CI 0.05 to 0.31, p = 0.007). Conversely, rest intervals less than 3 minutes (ES = -0.15, 95% CI -0.31 to 0.01, p = 0.052) or performing isometric contractions (ES = -0.52, 95% CI -0.89 to -0.14, p = 0.007) may be detrimental to performance. Meta-regression indicated that rest interval was the only moderator significantly associated with ES (β = -0.04, 95% CI -0.57 to -0.02, R = 14.31%, p < 0.001). When appropriate PAP guidelines are followed, an increase in vertical jump performance may be achieved.

The extended nervous system: affect regulation, somatic and social change processes associated with mindful parenting

BackgroundA theoretical model of mindful parenting has the potential to succinctly summarise its various change processes. The primary aim of this study was to investigate some of the change processes associated with mindful parenting, namely, the affect regulation, somatic and social change processes. A secondary aim was to verify whether clinical insights are consistent with the change processes identified in a systematic review of mindful parenting.MethodInterpretative Phenomenological Analysis (IPA) was used to analyse semi-structured interviews with four Australian clinicians delivering Mindful Parenting (MP) programs. The clinicians had extensive personal meditation practice. This qualitative study is part of a mixed methods study, which commenced with a quantitative systematic review.ResultsSix higher-order themes identified as change processes included reflective functioning, attachment, cognitive, affective, somatic and social change processes.ConclusionThe anchor is a new theoretical model summarising the change processes associated with mindful parenting. The mother portrayed as the extended nervous system for the infant is a neologism that also has not been previously mentioned in the literature. Given the limitations with the small sample and potential bias with interpretation, the anchor is a starting point to developing a theoretical model of mindful parenting. Future research with larger sample sizes and objective measures is needed to confirm whether the anchor is a reasonable summary of the change processes.