The Hawthorne effect is a psychological phenomenon characterized by individuals altering their conduct due to the perception of being under scrutiny. This effect was identified during studies performed at the Hawthorne Western Electric facility; however, its foundational accounts are regarded as apocryphal by certain academics.
The Hawthorne effect is a type of behavioral reactivity in which individuals modify an aspect of their behavior in response to their awareness of being observed. The effect was discovered in the context of research conducted at the Hawthorne Western Electric plant; however, some scholars think the descriptions are fictitious.
Initial investigations focused on employees manufacturing electrical relays at the Hawthorne Works, a Western Electric plant located in Cicero, Illinois. A lighting study, conducted from 1924 to 1927, involved workers undergoing various illumination adjustments, which were purportedly linked to enhanced productivity. This assertion was subsequently disproven. Subsequently, an Elton Mayo study (1927–1928) introduced structural modifications, such as altered rest periods, for a cohort of six female workers. Nevertheless, this particular study suffered from methodological deficiencies and a lack of control, precluding the derivation of definitive conclusions. Mayo subsequently undertook two further experiments to investigate the phenomenon: the mass interviewing experiment (1928–1930) and the bank wiring observation experiment (1931–1932).
Henry Landsberger, a sociology professor at UNC-Chapel Hill, subsequently proposed that the temporary elevation in worker productivity might stem from the novelty of participating in research and the heightened scrutiny received. This specific interpretation was subsequently termed "the Hawthorne effect."
History
The term "Hawthorne effect" was introduced in 1953 by John R. P. French, following the Hawthorne studies conducted from 1924 to 1932 at the Hawthorne Works, a Western Electric manufacturing facility situated in Cicero, near Chicago. The Hawthorne Works had initiated research to ascertain whether employee productivity varied with different levels of illumination. Employee output appeared to rise during periods of modification but reverted to baseline levels upon the study's conclusion. An alternative hypothesis posits that the observed increase in worker productivity resulted from the motivational impact of receiving focused attention.
This effect was noted even with minimal increments in illumination. In these lighting studies, illumination levels were manipulated to investigate their impact on employee output. When discussing the Hawthorne effect, the majority of industrial and organizational psychology textbooks predominantly reference the illumination experiments, rather than the broader spectrum of investigations undertaken.
Initial investigations primarily concentrated on modifying workplace lighting. However, other interventions, including workstation cleanliness, obstacle removal from floors, and workstation relocation, have similarly demonstrated temporary productivity gains. Consequently, the Hawthorne effect is not exclusively attributable to alterations in illumination but can encompass a range of causal factors.
Illumination experiment
The illumination experiment was carried out between 1924 and 1927. Its primary objective was to ascertain the influence of varying light conditions on employee output. The study utilized two distinct environments: an experimental room, where employees performed their duties under manipulated lighting, and a control room, where tasks were completed under standard conditions. The prevailing hypothesis posited that an increase in illumination within the experimental room would correlate with enhanced productivity.
Nevertheless, upon elevating light intensity in the experimental room, researchers observed productivity improvements in both environments. Subsequently, when illumination in the experimental room was reduced, an identical outcome—heightened productivity in both rooms—was recorded. A decline in productivity within the experimental room only occurred when light levels diminished to an extent comparable to moonlight, thereby impeding visibility.
The ultimate conclusion was that illumination itself exerted no discernible impact on productivity. Instead, it was inferred that an unidentified variable was responsible for the observed productivity gains across both settings. Further experimental phases were deemed necessary to isolate the underlying cause.
Relay assembly experiments
In 1927, an experiment commenced wherein researchers selected two female employees as initial subjects, who were then tasked with choosing four additional women to form a test group. This cohort of women subsequently operated in a distinct room until 1928, engaged in the assembly of telephone relays.
The mechanical measurement of output involved counting the number of completed relays each worker deposited into a chute. To establish a baseline productivity level, this measurement commenced covertly two weeks prior to the women's relocation to the experimental room and persisted throughout the entire study. Within the experimental setting, a supervisor engaged in discussions with the workers regarding fluctuations in their productivity.
Key variables investigated included:
- Initially, workers were provided with two 5-minute breaks, a preference they had previously expressed, which was subsequently altered to two 10-minute breaks. This adjustment led to an increase in productivity. However, when the break structure was changed to six 5-minute intervals, productivity declined, as the frequent interruptions disrupted the workers' operational rhythm.
- The provision of morning meals, such as soup or coffee with a sandwich, and evening snacks also contributed to enhanced productivity.
- Adjusting the workday's conclusion from 5:00 PM to 4:30 PM and eliminating Saturday work shifts similarly resulted in increased productivity.
Modifying any variable typically led to an increase in productivity, even when the change involved reverting to the original conditions. This phenomenon is often attributed to a natural adaptation to the environment, independent of the participants' awareness of the experiment's objectives. The researchers ultimately concluded that the heightened effort from workers stemmed from their perception of individual monitoring.
Researchers posited that the actual drivers of increased productivity included the autonomy to select coworkers, collaborative group work, a sense of special treatment (indicated by their placement in a distinct experimental room), and the presence of an empathetic supervisor. A prominent interpretation, largely derived from Elton Mayo's research, suggested that "the six individuals became a team and the team gave itself wholeheartedly and spontaneously to cooperation in the experiment". Additionally, a subsequent relay assembly test room study was conducted, though its findings did not exhibit the same level of significance as those from the initial experiment.
Mass Interviewing Program
This program, implemented between 1928 and 1930, encompassed 20,000 interviews. Initially, direct questioning was employed, focusing on inquiries concerning company supervision and policies. However, the limitation of this approach was that responses were often restricted to simple "yes" or "no" answers, proving inadequate for identifying underlying issues. Consequently, researchers adopted an indirect questioning method, where interviewers primarily listened. This shift yielded significant insights into worker behavior, particularly highlighting that individual conduct is profoundly influenced by group dynamics.
Bank Wiring Room Experiments
The subsequent study aimed to investigate the impact of payment incentives and small group dynamics on productivity. Surprisingly, productivity actually declined. It appeared that workers had grown suspicious, fearing that increased productivity might be used to rationalize future layoffs. Conducted by Elton Mayo and W. Lloyd Warner between 1931 and 1932, this study involved a group of fourteen men assembling telephone switching equipment. Despite workers being compensated based on individual productivity, the researchers observed a decrease in output, attributed to the men's apprehension that the company might reduce their base wage. Detailed observation further uncovered the presence of informal groups, or "cliques," operating within the formal organizational structure. These cliques established unwritten behavioral rules and enforcement mechanisms. Their function was to regulate group members and manage interactions with supervisors; for instance, when supervisors posed questions, clique members provided consistent responses, even if inaccurate. These findings underscore that workers were more influenced by the social pressures of their peer groups than by managerial control and incentives.
Interpretation and Criticism
Richard Nisbett characterized the Hawthorne effect as "a glorified anecdote," asserting that empirical data becomes superfluous once the anecdotal account is established. Various researchers have proposed alternative interpretations to elucidate these phenomena. J. G. Adair cautioned against significant factual inaccuracies prevalent in most secondary literature concerning the Hawthorne effect, noting that numerous studies failed to replicate it. Adair posited that the Hawthorne effect should be conceptualized as a manifestation of Orne's (1973) experimental demand effect, contingent upon participants' situational interpretations. This perspective underscores the importance of manipulation checks in social science experimentation. He further contended that the effect stemmed not from the mere awareness of observation, but from the participants' critical interpretation of their circumstances, prompting inquiry into how these interpretations interacted with their individual objectives.
Potential explanations for the Hawthorne effect encompass the influence of performance feedback and participants' motivation concerning the experimenter. Initial provision of performance feedback within an experimental context can potentially enhance participants' skills. Furthermore, research on the demand effect indicates that individuals may be inclined to satisfy the experimenter's expectations, provided this inclination does not conflict with other motivations. Conversely, participants might also harbor skepticism regarding the experimenter's objectives. Consequently, the Hawthorne effect may manifest exclusively in scenarios involving actionable feedback or a shift in motivational states.
Parsons delineated the Hawthorne effect as "the confounding that occurs if experimenters fail to realize how the consequences of subjects' performance affect what subjects do," encompassing learning effects such as permanent skill enhancement and feedback-driven adjustments to align with immediate objectives. His primary assertion was that in studies where laborers deposited completed items into chutes, participants could monitor their own production rates via accessible counters.
Mayo posited that the observed effect stemmed from workers' responses to the empathy and engagement demonstrated by the observers. He interpreted the study as illustrating an experimenter effect, which he reframed as a management effect: the capacity of management to alter worker performance by influencing their emotional states. Mayo suggested that a significant component of the Hawthorne effect involved workers experiencing a sense of autonomy and control as a collective, rather than merely being supervised. The experimental interventions were crucial in persuading workers that the conditions within their specialized five-person work group genuinely differed from those on the general shop floor. Subsequent replication of the study yielded comparable effects among mica-splitting laborers.
In a review of educational research, Clark and Sugrue documented that uncontrolled novelty effects typically induce an average increase of 30% of a standard deviation (SD), equating to a 50–63% score improvement, with this elevation diminishing significantly after eight weeks. More specifically, they observed a 50% SD increase for durations up to four weeks, a 30% SD increase for five to eight weeks, and a 20% SD increase for periods exceeding eight weeks, which constitutes less than 1% of the total variance.
Harry Braverman highlighted that the Hawthorne experiments originated from industrial psychology, with researchers primarily examining the predictability of worker performance through pre-employment assessments. The Hawthorne study revealed "that the performance of workers had little relation to their ability and in fact often bore an inverse relation to test scores ." Braverman contended that these studies actually demonstrated that the workplace constituted not "a system of bureaucratic formal organisation on the Weberian model, nor a system of informal group relations, as in the interpretation of Mayo and his followers but rather a system of power, of class antagonisms." This finding significantly undermined efforts to leverage behavioral sciences for the manipulation of workers in service of managerial interests.
Economists Steven Levitt and John A. List extensively sought the foundational data from the original illumination studies, which some authors inaccurately termed "experiments," eventually locating it on microfilm at the University of Wisconsin in Milwaukee in 2011. Their subsequent re-analysis revealed only marginal long-term evidence for the Hawthorne effect, significantly less pronounced than initially proposed. This outcome corroborated the findings of S. R. G. Jones's 1992 article, which scrutinized the relay experiments. Notwithstanding the lack of empirical support for the Hawthorne effect within the initial research, List has maintained his conviction in the effect's authenticity.
Gustav Wickström and Tom Bendix (2000) contend that the purported "Hawthorne effect" lacks clarity and is subject to debate. They propose that, for a robust evaluation of intervention efficacy, researchers ought to incorporate specific psychological and social variables that could have influenced the observed outcomes.
Alternatively, the illumination experiments might be attributable to a longitudinal learning effect. Parsons has refrained from analyzing these experiments, citing their inadequate publication and consequent lack of detailed information, in contrast to his extensive personal communications with Roethlisberger and Dickson.
Contemporary research continues to evaluate the Hawthorne effect. Notwithstanding the critiques, this phenomenon is frequently considered during the design of studies and the formulation of their conclusions. Furthermore, strategies have been devised to mitigate its influence, such as conducting observations in field studies remotely, from behind a barrier like a two-way mirror, or by employing unobtrusive measurement techniques.
Greenwood, Bolton, and Greenwood (1983) conducted interviews with several experiment participants, revealing that these individuals received substantially higher compensation. Bolton's archival materials pertaining to his research on the Hawthorne effect are preserved at West Virginia University.
Trial Effect
Numerous medical scientists have investigated the potential "trial effect" (or "clinical trial effect") within clinical research. Some researchers hypothesize that factors beyond mere attention and observation contribute to this effect, including marginally improved care, enhanced patient compliance or adherence, and selection bias. The latter mechanism can manifest through several pathways: (1) Clinicians may preferentially enroll patients perceived to have higher adherence potential and a reduced probability of subsequent loss to follow-up. (2) Trial inclusion and exclusion criteria frequently preclude individuals with certain comorbidities; while essential for confounding control, this practice often results in studies involving healthier patient subpopulations.
Secondary Observer Effect
Although the observer effect, as widely recognized from the Hawthorne experiments, may have been inaccurately characterized (refer to the preceding discussion), its theoretical prevalence and credibility have prompted researchers to hypothesize its occurrence at a secondary level. Consequently, the concept of a "secondary observer effect" has been proposed, suggesting that researchers analyzing secondary data, such as survey responses or various indicators, can inadvertently influence their scientific findings. Unlike the primary observer effect, which impacts research subjects, the secondary observer effect stems from the researchers' own methodological idiosyncrasies, which shape their data handling and even the specific data acquired from secondary sources. For instance, researchers might implement seemingly minor statistical analysis steps—such as weighting strategies, factor analytic techniques, or estimation choices—that ultimately yield substantially different results from the same dataset. Furthermore, variations in default settings across different software packages can introduce minor yet significant fluctuations. Lastly, the secondary data utilized by researchers may not be entirely identical, despite appearing so. For example, the OECD compiles and disseminates diverse socio-economic data; however, these datasets evolve over time, meaning a researcher downloading Australian GDP data for 2000 might obtain slightly different values than another researcher accessing the same data several years later. Nate Breznau initially introduced the concept of the secondary observer effect in a relatively obscure publication.
Despite limited scholarly focus on this phenomenon, its scientific implications are substantial. Recent investigations illustrate this effect by tasking multiple researchers or teams with a specific problem, wherein they independently analyze identical datasets to derive a solution. This methodology, termed crowdsourcing data analysis, was notably employed in a seminal 2015 study by Silberzahn, Rafael, Eric Uhlmann, Dan Martin, and Brian Nosek et al., which examined red cards and player race in football (soccer).
References
References
- Ciment, Shoshy. "Costco Is Offering an Additional $2 an Hour to Its Hourly Employees across the US as the Coronavirus Outbreak Causes Massive Shopping Surges." *Business Insider*, Business Insider, March 23, 2020.
- Miller, Katherine, and Joshua Barbour. *Organizational Communication: Approaches and Processes*, 7th Edition. Cengage Learning, 2014.
The Hawthorne, Pygmalion, Placebo, and Other Expectancy Effects: Some Notes, authored by Stephen W. Draper, Department of Psychology, University of Glasgow.
- The Hawthorne, Pygmalion, placebo and other expectancy effects: some notes, by Stephen W. Draper, Department of Psychology, University of Glasgow.
- BBC Radio 4: Mind Changers: The Hawthorne Effect.
- Harvard Business School and the Hawthorne Experiments (1924–1933), from Harvard Business School.