Does Workplace Environment Boost Performance? A Practical Mediation Analysis Walkthrough

A friend left his company because the office was cramped and crowded, which reduced his work efficiency.

Many factors affect work efficiency, and a common causal chain can be summarized as:

Work environment → Job satisfaction → Job performance

The question is whether the work environment (independent variable) influences employee performance (dependent variable) through job satisfaction (mediator).

What Is Mediation Effect?

In a typical mediation model, three main variables are involved:

Independent variable (X) : the primary factor of interest, e.g., work environment.

Mediator (M) : the variable that transmits the effect of X to Y, e.g., job satisfaction.

Dependent variable (Y) : the outcome influenced by X and M, e.g., job performance.

The core of mediation is understanding how X affects Y through M, which includes two paths:

Direct effect : X directly influences Y.

Indirect effect : X influences Y via M.

Mediation Effect Testing Methods

Common methods include the three‑step regression analysis, Sobel test, and Bootstrap method. This article focuses on the three‑step regression because it is relatively easy to understand.

Three‑Step Regression Analysis

Step 1: Total Effect

Test the overall effect of the independent variable (work environment) on the dependent variable (job performance) using a regression model.

Step 2: Mediation Effect

Test the effect of the independent variable on the mediator (job satisfaction) with another regression.

Step 3: Dependent Variable Regression

Include both the independent variable and the mediator in a regression predicting the dependent variable to assess the direct and indirect effects.

If the direct effect becomes non‑significant after adding the mediator, a full mediation is indicated; if it remains significant but reduced, a partial mediation exists.

Case Study

Assume a random sample of 100 employees was surveyed, collecting three scores:

Work environment rating (independent variable)

Job satisfaction rating (mediator)

Job performance rating (dependent variable)

All scores are on a 1‑10 scale. Regression results show:

The total effect is significant, indicating work environment impacts performance.

The mediation effect is significant, showing work environment affects job satisfaction.

When the mediator is added, the direct effect of work environment on performance remains significant but is reduced, and the mediator’s effect is also significant, indicating partial mediation.

Sobel Test and Bootstrap Method

The Sobel test calculates the product of the coefficient from X to M (a) and the coefficient from M to Y (b) and uses their standard errors to compute a Z‑statistic for the indirect effect.

The Bootstrap method repeatedly resamples the data with replacement to generate a distribution of the indirect effect (a × b). A confidence interval that does not include zero indicates a significant mediation effect.

Both approaches help verify the existence and significance of mediation. In this example, improving the workplace environment can increase job satisfaction, which in turn partially boosts job performance.