Jadwal Sholat

Memuat jadwal sholat…

Ekonomi & Bisnis editorial

Open AccessOA2023

Inference of Factors for Labor Productivity Growth Used Randomized Experiment and Statistical Causality

This paper proposes a hybrid technology integrating causal Bayesian networks, randomized controlled experiments, and statistical methods to infer causal relations between company programs and labor productivity, even under nonlinearity, dynamism, stochasticity, and non-stationarity. Tested on a food service company, the approach identifies synergistic effects of programs on productivity in the presence of hidden factors.
E. Orlova· Mathematics· 2023· DOI 10.3390/math11040863

The core problem

The study of causal dependencies in economics is challenging because it requires considering not only the object's structure but also a multitude of factors that are either unknown or difficult to measure. This paper attempts to overcome this problem by applying the theory of statistical causality to labor productivity management. The author proposes a new technology for inferring causal relations between special programs implemented in a company and employee labor productivity. The novelty lies in a hybrid object model that combines two models: (1) a structural object model about its functioning and development to provide causal inference and predict the effect of explicit factors; and (2) a model based on observed data to clarify causality and test it empirically. The technology integrates causal Bayesian networks, randomized controlled experiments, and statistical methods, enabling evaluation of program effects on labor efficiency under nonlinearity, dynamism, stochasticity, and non-stationarity of initial data. Unlike other approaches, it determines the synergistic effect of a cause (program) on the effect—labor productivity—in the presence of hidden factors. The practical significance is demonstrated by testing on actual data from a food service company, contributing to labor productivity growth under uncertainty and supporting sustainable development and profitability growth.

Innovation

The proposed technology is based on a hybrid object model that integrates two complementary models. The first is a structural object model that describes the company's functioning and development, enabling causal inference and prediction of the effects of explicit factors. The second is a data-driven model that uses observed data to clarify and empirically test causality. The technology combines causal Bayesian networks, randomized controlled experiments, and statistical methods. This integration allows the evaluation of program effects on labor efficiency under conditions of nonlinearity, dynamism, stochasticity, and non-stationarity. A key feature is the ability to determine the synergistic effect of a cause (program) on the effect (labor productivity) when hidden factors are present. The methodology is designed to handle the complexity of economic systems where many factors are unknown or hard to measure, providing a robust framework for causal inference in labor productivity management.
Introduction
The study of causal dependencies in economics is challenging because it requires considering not only the object's structure but also a multitude of factors that are either unknown or difficult to measure. This paper attempts to overcome this problem by applying the theory of statistical causality to labor productivity management. The author proposes a new technology for inferring causal relations between special programs implemented in a company and employee labor productivity. The novelty lies in a hybrid object model that combines two models: (1) a structural object model about its functioning and development to provide causal inference and predict the effect of explicit factors; and (2) a model based on observed data to clarify causality and test it empirically. The technology integrates causal Bayesian networks, randomized controlled experiments, and statistical methods, enabling evaluation of program effects on labor efficiency under nonlinearity, dynamism, stochasticity, and non-stationarity of initial data. Unlike other approaches, it determines the synergistic effect of a cause (program) on the effect—labor productivity—in the presence of hidden factors. The practical significance is demonstrated by testing on actual data from a food service company, contributing to labor productivity growth under uncertainty and supporting sustainable development and profitability growth.

Why it matters

The paper addresses the difficulty of causal inference in economics due to the multitude of unknown or hard-to-measure factors. By combining structural modeling with data-driven causal discovery and randomized experiments, the proposed technology offers a novel way to overcome these challenges. The hybrid approach allows for both prediction of explicit factor effects and empirical testing of causality, which is particularly valuable under nonlinearity, dynamism, stochasticity, and non-stationarity. The ability to determine synergistic effects of programs on labor productivity in the presence of hidden factors distinguishes this technology from others. The practical significance is underscored by the successful application to a food service company, showing that the results can aid in labor productivity growth under uncertainty and contribute to sustainable development and profitability. The author suggests that the technology could be adapted to other companies and contexts, although further validation may be needed.

Who should read this

Pembaca Research Digest dan praktisi rumpun ini.

Opening member content…