Abstract
This scoping literature review addresses the research question of measuring stress: What methodologies and associated metrics are employed for the in-situ assessment of stress among knowledge workers? Lab and field studies of any qualitative, quantitative, or mixed design reported in English from 2000 to 15 October 2024 were retrieved from Compendex, PubMed, and Web of Science. The 17 included studies covered office workers in several countries; all included qualitative elements while a minority (6) included quantitative methods. Future research would benefit from the development of unobtrusive data collection methods and analysis of different types of stressors on office workers. The review also informs which methodologies and metrics might be applied to hybrid office workers, a subset of knowledge/office workers, in future studies on stress.
Keywords
Introduction
Office workers, also known as knowledge workers or white-collar workers, make up about 58% of the working population in the United States (AFL-CIO, Department of Professional Employees, 2024; U.S. Bureau of Labor Statistics, 2023). The stress office workers experience can lead to poor organizational and personal outcomes, such as increased absenteeism and an increase in the prevalence of mental illness diagnoses (Bolliger et al., 2022). Limited research exists on the specific stressors impacting this group, and even less is known about the complex interplay between different stress types and their cumulative effects.
The descriptive words “office,” “knowledge,” “information,” and “white-collar” workers are all used interchangeably to describe the same group of people. These workers are employed in a variety of industries and roles (U.S. Bureau of Labor Statistics, 2023). They are clerical workers, administrative workers, managers, consultants, analysts, and professionals. Knowledge workers are often characterized by loosely defined roles, a high degree of autonomy, and predominantly sedentary work habits (Surawski, 2019). While some may have clearly delineated tasks, such as call center representatives or claims processors, the majority engage in complex cognitive activities requiring significant executive function, such as inhibition, working memory, and cognitive flexibility (Diamond, 2013; Surawski, 2019). A common theme in their work is use of the computer, whether a desktop or a laptop. Workers report spending anywhere from 4 to 8 hr each day on the computer (Griffiths et al., 2012; Hayes et al., 2007).
As the first step to investigate reliable quantitative and qualitative metrics for office worker stress, a scoping literature review was prepared on the following research question: What methodologies and associated metrics have been employed for the in-situ assessment of stress among knowledge workers? The impact of work stress on knowledge workers is an understudied area, potentially due to inconsistencies in measurement methodologies (Maxhuni et al., 2016). To address this, this review highlights only in situ measurements to better understand the immediate effects of stress on biometric signals and survey responses.
Approach
The Compendex, Web of Science and PubMed databases were selected. These databases are drawn from various sources in engineering, science, biomedicine, and life science. The inclusion criteria were: (1) studies that investigated office/knowledge workers or those completing office/knowledge workers’ tasks; (2) laboratory and field experiments; (3) studies that directly measured stress or estimated it using other method(s); (4) quantitative, qualitative, and mixed methods studies; (5) studies that measure stress while at work (location) or while performing work or tasks similar to those they would have performed at work (location independent); (6) papers written in English; and (7) papers published in or after 2,000 through 15 Oct 2024, taking into account the technological, social, and methodological changes that have shaped the modern workplace. The search logic used was: (((“stress at work” OR “work stress” AND measure*) AND (knowledge OR office work* OR office build*))). The review followed the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) process (Figure 1) (Page et al., 2021). The initial extraction resulted in 173 articles. After duplicates were removed and review by title, abstract, participants and methods were completed, 20 articles remained. Seventeen articles were included in the review; three were removed as the full text wasn’t available. This review did not include any distinctions between different kinds of stress (occupational, non-occupational, personal or psychosocial).
Flow diagram (based on PRISMA format).
Outcome
The extracted characteristics of the 17 studies, are listed in Table 1. The study population, qualitative measures, quantitative measures and statistical methods are summarized in the following sections.
Select Extracted Data from Reviewed Studies.
Asterisks denote the identified outcome measures.
Population of Reviewed Studies
The studies included in this review incorporated a diverse cohort of office workers, encompassed male and/or female workers; settings such as Japan, New Zealand, and Italy; and industries including healthcare, manufacturing, and municipalities. Additionally, the research included individuals in simulated office settings, showroom environments, and those working fully remotely or hybrid for at least 20 hr per week. The sample sizes across the studies ranged from 13 to 57 participants. All participants, whether in a lab or in the field, completed office type tasks. Some were controlled tasks, such as typing tests, and some were more reflective of an office worker’s tasks, including writing emails, creating a presentation, and report writing (Koldijk et al., 2018).
Qualitative Methods in Review
Qualitative methods for measuring stress in office workers primarily involved surveys and interviews in the 17 selected studies. These methods can identify individuals more prone to stress, such as those with Type A personalities, and evaluate the impact of work on overall well-being (Hagihara et al., 2001). Surveys can assess job-related stressors, such as workload, decision-making authority, and skill discretion, as well as coping mechanisms and recovery experiences as in Jun et al. (2019). Surveys used in these studies were delivered electronically, via paper, orally, and even through smart phone prompts as in Maxhuni et al. (2016). Surveys were delivered at various intervals— annually, semi-annually, monthly, daily, and several times throughout the day. Kerr et al. (2020) measured immediate stress levels during a specific situation. Measuring stress also took place at the beginning, during, or end of a workday. Delivery of surveys interrupted the worker performing tasks in all but three of the reviewed studies.
Office workers’ stress is multifaceted, and there are many links to stress in these studies. Hagihara et al. (2001) found that personality type and job demand were significant contributors to the amount of alcohol consumed by office workers, indicating that higher work stress led to higher alcohol consumption. Psychosocial and organizational factors were associated with musculoskeletal discomfort, vision issues, stress, and mental well-being (Seppälä, 2001). Igic et al. (2013) found that job control, or lack thereof, predicted physical strain through spinal compression. In addition, Troup and Dewe (2002) found the imbalance between desire for job control and actual level of job control was associated with higher work stress. Higher perceived work stress was related with compulsive eating behaviors in a study by Park and Sung (2021). Gender also played a role in stress moderation, indicating that women tend to seek social support for stressful work conditions more than men (Hulin et al., 2024).
Quantitative Methods in Review
Quantitative methods for measuring stress in office workers focus on physiological and biological measurements and assessments of the work environment. Physiological measurements, such as heart rate variability, and biological measurements, like salivary cortisol levels and electrodermal activity, can assess the possible manifestations of stress on the body. In the included studies, physiological measures of stress include heart rate variability, heart rate, and changes in blood volume (photoplethysmography). Biologic measures noted included posture, postural changes, and facial expressions. Work stress was found to be associated with significant changes in heart rate metrics and salivary cortisol (Kerr et al., 2020).
Assessments of the work environment, such as indoor air quality and computer interactions, also provided quantitative data related to stress. Pressure metrics, like keyboard pressure, mouse grip pressure, and seat pressure were measured using transducers either attached to a computer input device or the worker’s chair. Computer use data, such as number and speed of keystrokes and mouse use data (number of clicks, feet of mouse movement) were collected using computer logging software. Indoor air quality, temperature, humidity, and carbon dioxide levels were measured using traditional tools and revealed some correlations with work stress (Lindgren, 2010).
Observational and diary studies as well as structured interviews can provide quantitative data. These methods captured stress-related behaviors, interactions, and experiences in the workplace, and offered a comprehensive understanding of the factors contributing to stress in office workers. Koldijk et al. (2018) found that mental effort could be predicted by body postures, facial expressions, and computer use data in a laboratory environment.
Data Analysis Found in Review
The studies included in the review used common statistical tests and procedures like descriptive statistics and basic measures like means, standard deviations, and frequencies. These offered comparable ways to summarize data between studies. More advanced statistical tests were employed in Maina et al. (2009) including Tukey, chi-square, t-tests, and Mann–Whitney, to examine the relationship between work stress models and salivary cortisol. These types of tests compare two groups—either the means of those groups (t-test) or between groups/categories (chi-square). Mizoue et al. (2006) used chi-square and ANOVA to analyze the relationship between overtime work and cigarette consumption. Kerr et al. (2020) used ANOVA and Area Under the Curve (AUC) analysis for salivary cortisol to investigate the effects of experimentally induced work stress on physiological and psychological responses.
Seppälä (2001) used regression analysis to investigate the impact of psychosocial and organizational factors on stress, and Igic et al. (2013) used multilevel regression analysis to examine the impact of job control on spinal shrinkage. Park and Sung (2021) used correlation analysis to assess the strength and direction of the relationship between two variables, such as stress levels and compulsive eating behaviors.
As stress modeling is complex and multidimensional, complicated statistical measures and tests were often employed, using statistical modeling and machine learning techniques. For example, Koldijk et al. (2018) used Support Vector Machines (SVM), Bayesian classification trees, linear regression, and nearest neighbors to predict mental effort from various data sources. Principal Component Analysis (PCA) was used by Troup and Dewe (2002) to explore the multi-dimensional nature of job control, a significant factor in understanding work stress.
Conclusion
This literature review aimed to understand the current state of knowledge regarding in situ stress measurement among office and knowledge workers. The nature of data collection methods presented difficulties in correlating survey results or biometric data with specific work activities. Accurately matching data, particularly survey data, to specific aspects of a worker’s life, such as work, was challenging without a detailed record of daily activities. Koldijk et al. (2018) found that mental effort could be predicted by body postures, facial expressions, and computer use data; however, this study was conducted in a laboratory environment, giving pause for real world measurement activities for office workers. As noted in Schmid et al. (2024), while laboratory settings offer controlled environments for measuring stress, their generalizability to real-world work conditions, which are often dynamic and unpredictable, can be problematic.
Stress, particularly work-related stress, is a complex phenomenon that is difficult to measure comprehensively due to its multifaceted nature and individual variability. The review also identified gaps in existing research to guide future investigations. The review revealed that in situ measurements of office worker stress provide an understanding of its multifaceted aspects. However, obtaining such measurements, whether quantitative or qualitative, can be challenging in a work environment, necessitating careful consideration to minimize disruption and additional stress on workers. As a minority (6) of the studies employ both qualitative and quantitative methods, there is a need for additional mixed-methods research that integrates physiological and biological data with subjective experiences to provide a more holistic understanding of stress.
Future Directions
One noted research gap is minimizing the disruption to participants’ work experiences during qualitative survey delivery. Researchers can achieve this by using the findings of the reviewed studies to guide questionnaire development and delivery of the most relevant questions. This approach ensures that workers are not overwhelmed and that the collected information is both accurate and meaningful.
Future research should prioritize passive data collection methods and aim to link stress perception with quantitative physiological and biological measures to provide a more comprehensive picture of mental stress in office workers. The results and findings of the studies in this literature review can help to streamline data collection by focusing on necessary information. Investigating the complex interplay between different types of stressors, such as work-related, personal, and environmental, and their combined impact on office workers’ health and well-being is also crucial.
Footnotes
Acknowledgements
Many thanks to Dr. Farzan Sasangohar for his leadership, Dianna Morganti for her teaching, and Jacob Kolman for his reviewing.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
