Search results

Longitudinal investigations are often suggested but rarely used in operations and supply chain management (OSCM), mainly due to the difficulty of obtaining data. There is a silver lining in the form of existing large-scale and planned repeated cross-sectional (RCS) data sets, an approach commonly used in sociology and political sciences. This study aims to review all relevant RCS surveys with a focus on OSCM, as well as data and methods to motivate longitudinal research and to study trends at the plant, industry and geographic levels.

A comparison of RCS, panel and hybrid surveys is presented. Existing RCS data sets in the OSCM discipline and their features are discussed. In total, 30 years of Global Manufacturing Research Group data are used to explore the applicability of analytical methods at the plant and aggregate level and in the form of multilevel modeling.

RCS analysis is a viable alternative to overcome the confines associated with panel data. The structure of the existing data sets restricts quantitative analysis due to survey and sampling issues. Opportunities surrounding RCS analysis are illustrated, and survey design recommendations are provided.

The longitudinal aspect of RCS surveys can answer new and untested research questions through repeated random sampling in focused topic areas. Planned RCS surveys can benefit from the provided recommendations.

RCS research designs are generally overlooked in OSCM. This study provides an analysis of RCS data sets and future survey recommendations.

Undergraduate students are known to be a high-risk group for mental health problems. The purpose of this paper is to constitute a repeated cross-sectional study on the trend of depression over the years and factors associated with depression among undergraduates.

Cross-sectional data from five surveys between 2013 and 2020 (N = 1,578) among the undergraduates of Universiti Tunku Abdul Rahman, a private university in Kampar Malaysia, were combined. The Depression Anxiety and Stress Scale-21 was used to screen for depression. Cochran’s Armitage test was used to detect trend in depression. Logistic regression, random forest regression and extra gradient boosting regression were used to identify risk factors and classification.

The prevalence of depressive symptoms was found to be between 26.4% and 36.8% between the years with an average of 29.9%. There was no significant time trend in the prevalence. The risk of depressive symptoms was higher among female students, those who were dependent on family for financial support and those who were stressed.

Periodical screening for depression is warranted for the identification of students at risk for depression. Professional cognitive-behavioral therapies, peer support and consulting services should be made available to the students in need.

Depression among students had been studied widely, but the trend over years remains unexplored, especially in developing countries.

The purpose of this paper is to provide a systematic overview with guidelines how to use partial least squares (PLS) path modeling in longitudinal studies. Practical examples from a study of the acceptance of battery electric vehicles (BEVs) in corporate fleets are used for demonstration purposes.

In this study, data at three points in time were collected: before the initial use of a BEV, after three and after six months of extensive usage of BEVs.

Three different models are identified depending on the research objective and on the data basis. Multigroup analyses are suggested to test the difference between the path coefficients of latent variables at different points in time. Limitations for the use of repeated cross-sectional data have to be observed.

Academics and practitioners will benefit from this paper by receiving an overview of the different PLS path models in longitudinal studies. A decision-tree enables them to make a choice regarding the most appropriate model and suggests a sequence of complementary analyses. So far, there is a lack of a tutorial type paper delivering such guidance.

The authors propose a framework to estimate the probability of being poor in a dynamic setting based on a large information set that includes individual characteristics and macro-economic variables. The joint inclusion of personal characteristics along with contextual factors allows separation of idiosyncratic shocks from aggregate shocks affecting poverty. The authors combine data from different cross-sectional surveys and fit a dynamic logistic hierarchical model within a Bayesian framework using standard Markov chain Monte Carlo techniques. The authors’ approach is exemplified by estimating household poverty status in Kyrgyz Republic as a function of time, regions, country, regional level variables and household level socio-demographic characteristics.

This study aims to examine workers' distributive and interactional justice perceptions at three different moments in time over a period of eight years, assess their degree of stability and identify their most stable antecedents and outcomes.

Data was collected through an overlapping repeated cross-sectional design. Of the participants involved, 334 were surveyed in 2000, 259 participated in 2004, and 285 participated in 2008.

Distributive justice is more stable than interactional justice. Organizational support is the most stable predictor of distributive justice, and the quality of supervisor practices is the most stable predictor of interactional justice. Contrary to expected, interactional justice has a stronger relationship to workers' attitudes directed both at the organization and supervisor, and at the immediate work context.

This study adopts a long-term perspective covering an eight-year period. Furthermore, it focuses on two dimensions of justice that have been less studied over time.

The purpose of this article is to present an empirical analysis of complex sample data with regard to the biasing effect of non‐independence of observations on standard error parameter estimates. Using field data structured in the form of repeated measurements it is to be shown, in a two‐factor confirmatory factor analysis model, how the bias in SE can be derived when the non‐independence is ignored.

Three estimation procedures are compared: normal asymptotic theory (maximum likelihood); non‐parametric standard error estimation (naïve bootstrap); and sandwich (robust covariance matrix) estimation (pseudo‐maximum likelihood).

The study reveals that, when using either normal asymptotic theory or non‐parametric standard error estimation, the SE bias produced by the non‐independence of observations can be noteworthy.

Considering the methodological constraints in employing field data, the three analyses examined must be interpreted independently and as a result taxonomic generalisations are limited. However, the study still provides “case study” evidence suggesting the existence of the relationship between non‐independence of observations and standard error bias estimates.

Given the increasing popularity of structural equation models in the social sciences and in particular in the marketing discipline, the paper provides a theoretical and practical insight into how to treat repeated measures and clustered data in general, adding to previous methodological research. Some conclusions and suggestions for researchers who make use of partial least squares modelling are also drawn.

In this study, we attempt to estimate the disability costs of households employing the Standard of Livings (SoL) approach and evaluate the impact of the Universal health system reform implemented in Turkey in 2008.

We apply a Structural Equation Modelling (SEM), which simultaneously estimates the disability and living standard equations, including unobserved latent variables. Moreover, we apply a difference-in-differences (DiD) framework to investigate the impact of the universal health insurance (UHI) system and the Green Card programme on living standards. The empirical analysis relies on data derived from the cross-sectional Household Budget Surveys (HBS) during the period 2002–2013.

Our findings suggest a negative and significant impact of disability on SoL, where disability costs reach the 23% of the household income, which is equivalent almost to $2,600 (USD). Furthermore, the disability costs are reduced from $4,450 to $2,260 due to the UHI and the Green Card programme.

A major limitation of the study is the data structure, which is based on repeated cross-sectional surveys. By using panel data, it is possible to follow the same individual across time and to implement panel data models to control for unobserved heterogeneity and omitted-variable bias.

Disability has adverse effects on living standards. The estimation of the disability-related costs may provide a useful guide on policy planning and the design of social benefits.

The contribution of this paper is that it is the first study estimating the disability-related costs in Turkey. Furthermore, the contribution lies in the investigation of the 2008 health reform and the Green Card programme and its impact on disability costs.

The purpose of this paper is to investigate the effects of tax reform on entrepreneurship in South Africa using repeated cross-sectional data from the World Bank.

The paper adopts a difference-in-difference estimation technique as well as contrasting periods before and after the tax reform. This contrast is achieved by examining individuals in the formal and informal sector and measuring the effectiveness of the reform on self-employment.

The results indicate that the tax reform had a positive and significant effect on the probability of becoming self-employed in South Africa and is robust across different econometric specifications.

The authors use individual-level data to measure the effectiveness of a tax reform policy on entrepreneurship. Utilizing the South African post-Apartheid tax reform as a natural experiment allows the authors to identify the effects of taxes on the choice of becoming self-employed.

The purpose of this paper is to explore the determinants of life satisfaction in Turkey. Moreover, this study explores the effects of air pollution and crime problems on well-being.

The estimates are based on cross-sectional data from the health survey in Turkey during the years 2010 and 2012. Various econometric models are applied such as the ordered logit and the random-effects generalized latent class ordered logit. Moreover, using pseudo panel data created based on age and region cohorts adapted probit fixed effects and the “blow-up and cluster” estimators are applied. In addition, various estimates by sex, age group, urban and rural areas as well as between individuals with good and poor health status are followed.

The results show that the individuals who self-reported who are exposed to air pollution and crimes present on average 0.2-0.5 less satisfaction scores than those who are not exposed to air pollution and crimes. In terms of monetary values, they are willing to pay more than those who are not exposed to air pollution and crimes by 13-19 Turkish Liras per month. Moreover, the generalized latent class ordered logit shows that there is considerable heterogeneity among the most satisfied and least satisfied individuals.

The originality of the paper lies in the fact that this is the first study to provide an analysis of life satisfaction using micro-level data from Turkey. Moreover, various econometric approaches are applied to compare the results. In addition, examining the heterogeneous effects among individuals with different life satisfaction rankings, it is possible to examine the effects of various factors on well-being and how they differ among individuals. Finally, by examining exposure to air pollution and crimes in the neighbourhood and their effects on well-being, it is possible to control for characteristics of the deprived areas.

The purpose of this study is to evaluate the impact of research and development (R&D) subsidies and tax credits on the innovative processes of Chilean firms.

Probit and tobit models for pseudo-panel with instrumental variables are estimated using data from different versions of the Innovation Survey covering the period 2007–2016.

The results show that R&D subsidies and tax credits have a statistically significant and positive effect on the probability of performing internal and external R&D, but do not affect the intensity of R&D spending, reflecting a crowding-out effect on private funds of both instruments. On the other hand, firms that simultaneously receive R&D subsidies and tax credits have a lower percentage of innovative sales. Furthermore, there are not effects statistically significant of the R&D subsidies and/or tax credits on the number of intellectual property rights applications.

It is concluded that both instruments have not been effective to encourage innovative outputs in Chilean firms.

Este estudio evalúa el impacto de los subsidios e incentivos tributarios para investigación y desarrollo (I&D) sobre los procesos innovativos de las firmas chilenas.

Se estiman modelos Probit y Tobit con variables instrumentales para pseudo-panel a partir de datos provenientes de diversas versiones de la Encuesta de Innovación que cubren el periodo 2007–2016.

Los resultados muestran que los subsidios e incentivos tributarios para I&D tienen un efecto positivo y estadísticamente significativo sobre la probabilidad de realizar I&D interna y externa, pero no afectan la intensidad del gasto en I&D lo que refleja un efecto expulsión sobre los fondos privados de ambos instrumentos. Por otro lado, las firmas que reciben simultáneamente subsidios e incentivos tributarios para I&D tienen menor porcentaje de ventas innovativas. Además, no se detecta un impacto significativo de los subsidios y/o incentivos tributarios sobre los derechos de propiedad intelectual solicitados por las firmas.

Así, se concluye que ambos instrumentos no han sido efectivos para incentivar los outputs innovativos en las firmas chilenas.