Search results

In this chapter, we explicate two related techniques that help quantify the sensitivity of a given causal inference to potential omitted variables and/or other sources of unexplained heterogeneity. In particular, we describe the Impact Threshold of a Confounding Variable (ITCV) and the Robustness of Inference to Replacement (RIR). The ITCV describes the minimum correlation necessary between an omitted variable and the focal parameters of a study to have created a spurious or invalid statistical inference. The RIR is a technique that quantifies the percentage of observations with nonzero effects in a sample that would need to be replaced with zero effects in order to overturn a given causal inference at any desired threshold. The RIR also measures the percentage of a given parameter estimate that would need to be biased in order to overturn an inference. Each of these procedures is critical to help establish causal inference, perhaps especially for research urgently studying the COVID-19 pandemic when scholars are not afforded the luxury of extended time periods to determine precise magnitudes of relationships between variables. Over the course of this chapter, we define each technique, illustrate how they are applied in the context of seminal strategic management research, offer guidelines for interpreting corresponding results, and delineate further considerations.

This essay is a review of the recent literature on the methodology of economics, with a focus on three broad trends that have defined the core lines of research within the discipline during the last two decades. These trends are: (a) the philosophical analysis of economic modelling and economic explanation; (b) the epistemology of causal inference, evidence diversity and evidence-based policy and (c) the investigation of the methodological underpinnings and public policy implications of behavioural economics. The final output is inevitably not exhaustive, yet it aims at offering a fair taste of some of the most representative questions in the field on which many philosophers, methodologists and social scientists have recently been placing a great deal of intellectual effort. The topics and references compiled in this review should serve at least as safe introductions to some of the central research questions in the philosophy and methodology of economics.

Much of what we learn from empirical research is based on a specific empirical model(s) presented in the literature. However, the range of plausible models given the data is potentially larger, thus creating an additional source of uncertainty termed: model uncertainty. The purpose of this paper is to examine the effect of model uncertainty on empirical research in HRM and suggest potential solutions to deal with the same.

Using a sample of call center employees from India, the authors test the robustness of predictors of intention to leave based on the unfolding model proposed by Harman et.al. (2007). Methodologically, the authors use Bayesian Model Averaging (BMA) to identify the specific variables within the unfolding model that have a robust relationship with turnover intentions after accounting for model uncertainty.

The findings show that indeed model uncertainty can impact what we learn from empirical studies. More specifically, in the context of the sample, using four plausible model specifications, the authors show that the conclusions can vary depending on which model the authors choose to interpret. Furthermore, using BMA, the authors find that only two variables, job satisfaction and perceived organizational support, are model specification independent robust predictors of intention to leave.

The research has specific implications for the development of HR analytics and informs managers on which are the most robust elements affecting attrition.

While empirical research typically acknowledges and corrects for the presence of sampling uncertainty through p-values, rarely does it acknowledge the presence of model uncertainty (which variables to include in a model). To the best of the authors’ knowledge, it is the first study to show the effect and offer a solution to studying total uncertainty (sampling uncertainty + model uncertainty) on empirical research in HRM. The work should open more doors toward more studies evaluating the robustness of key HRM constructs in explaining important work-related outcomes.

This paper is about the effect of human capital on foreign direct investment (FDI). The purpose of this paper is to find out if developing countries with high levels of human capital (educated people and well-trained labour force) are more successful in attracting FDI. The underlying hypothesis has been tested repeatedly without reaching a consensus view or providing an answer to the basic question. This is to be expected because FDI is determined by a large number of factors, making the results sensitive to the selected set of explanatory variables, which forms the basis of the Leamer (1983) critique of the use of multiple regression to derive inference. Furthermore, confirmation bias and publication bias entice researchers to be selective in choosing the set of results they report.

The technique of extreme bounds analysis, as originally suggested by Leamer (1983) and modified by Sala-i-Martin (1997), is used to determine the importance of human capital for the ability of developing countries to attract FDI. The authors use a cross-sectional sample covering 103 developing and transition countries.

The results show no contradiction between firms seeking human capital and cheap labour. No matter what proxy is used to represent human capital, it turns out that the most important factor for attracting FDI is the variable “employee compensation”, which is the wage bill, implying that multinational firms look for cheap and also skilled labour in the host country.

In this paper, the authors follow the procedure prescribed by Leamer (1983), and modified by Sala-i-Martin (1997), using extreme bounds analysis to distinguish between robust and fragile determinants of FDI, with particular emphasis on human capital. Instead of deriving inference from one regression equation by determining the statistical significance of the coefficient on the variable of interest, the extreme bounds or the distribution of estimated coefficients are used to distinguish between robust and fragile variables. This means that emphasis is shifted from significance, as implied by a single regression equation, to robustness, which is based on a large number of equations. The authors conduct tests on three proxies for human capital to find out if they are robust determinants of FDI and also judge the degree of robustness relative to other determinants.

The overall goal of science is to build a valid and reliable body of knowledge about the functioning of the world and how applying that knowledge can change it. As personnel and human resources management researchers, we aim to contribute to the respective bodies of knowledge to provide both employers and employees with a workable foundation to help with those problems they are confronted with. However, what research on research has consistently demonstrated is that the scientific endeavor possesses existential issues including a substantial lack of (a) solid theory, (b) replicability, (c) reproducibility, (d) proper and generalizable samples, (e) sufficient quality control (i.e., peer review), (f) robust and trustworthy statistical results, (g) availability of research, and (h) sufficient practical implications. In this chapter, we first sing a song of sorrow regarding the current state of the social sciences in general and personnel and human resources management specifically. Then, we investigate potential grievances that might have led to it (i.e., questionable research practices, misplaced incentives), only to end with a verse of hope by outlining an avenue for betterment (i.e., open science and policy changes at multiple levels).

Prior research has documented inconclusive and/or mixed empirical evidence on the timing performance of hybrid funds. Their performance inferences generally do not efficiently control for fixed-income exposure, conditioning information, and cross-correlations in fund returns. This study examines the stock and bond timing performances of hybrid funds while controlling and accounting for these important issues. It also discusses the inferential implications of using alternative bootstrap resampling approaches.

We examine the stock and bond timing performances of hybrid funds using (un)conditional multi-factor benchmark models with robust estimation inferences. We also rely on the block bootstrap method to account for cross-correlations in fund returns and to separate the effects of luck or sampling variation from manager skill.

We find that the timing performance of portfolios of funds is neutral and sensitive to controlling for fixed-income exposures and choice of the timing measurement model. The block-bootstrap analyses of funds in the tails of the distributions of stock timing performances suggest that sampling variation explains the underperformance of extreme left tail funds and confirms the good and bad luck in the bond timing management of tail funds. We report inference changes based on whether the Kosowski et al. or the Fama and French bootstrap approach is used.

This study provides extensive and robust evidence on the stock and bond timing performances of hybrid funds and their sensitivity based on (un)conditional linear multi-factor benchmark models. It examines the timing performances in the extreme tails funds using the block bootstrap method to efficiently identify (un)skilled fund managers. It also highlights the sensitivity of inferences to the choice of testing methodology.

This article investigates the robustness of impulse response estimators to near unit roots and near cointegration in vector autoregressive (VAR) models. We compare estimators based on VAR specifications determined by pretests for unit roots and cointegration as well as unrestricted VAR specifications in levels. Our main finding is that the impulse response estimators obtained from the levels specification tend to be most robust when the magnitude of the roots is not known. The pretest specification works well only when the restrictions imposed by the model are satisfied. Its performance deteriorates even for small deviations from the exact unit root for one or more model variables. We illustrate the practical relevance of our results through simulation examples and an empirical application.

The purpose of this paper is to investigate the relationship between the share price volatility of Mediterranean banks and their dividend policies, with particular emphasis on the variation of results across sub-samples and the outcomes when omitting outlier observations.

The authors use dividend yield and dividend payout as proxies of dividend policy, and regress these ratios together with other control variables to model volatility. The robustness of the results is assessed by re-using a data set which omits the outliers relating to the aftermath of the 2007 financial crisis and by forming sub-samples using cluster analysis.

The results show that the elimination of outliers and the setting up of sub-samples lead to different inferences about the underlying relationship between dividend policy and volatility. In addition traditional indicators of statistical significance may give the impression of a robust relationship, when this may not be the case.

The paper offers insights to stock traders and corporate managers in terms of better understanding the effect of dividend policies on share price volatility and its related risks and opportunities.

The study presents noteworthy empirical evidence in terms of its rigorous approach towards checking the robustness of results.