Algorithmic Management Questionnaire

Abstract

The Algorithmic Management Questionnaire (AMQ; Parent-Rocheleau et al., 2024) is a psychometrically validated self-report instrument designed to quantify workers’ subjective exposure to algorithmic management across contemporary organizational environments. As artificial intelligence, autonomous optimization algorithms, and digital surveillance platforms increasingly mediate managerial duties historically conducted by human supervisors, psychometrically sound measurement of algorithmic control has become essential. Developed via a six-phase deductive scale development methodology following established psychometric protocols (Hinkin, 1998), the AMQ assesses five distinct functional dimensions: Monitoring, Goal Setting, Scheduling, Performance Rating, and Compensation. The finalized instrument comprises 20 items evaluated on a standardized Likert response format. Confirmatory factor analyses conducted across multiple independent worker samples demonstrate that the five-factor model achieves exceptional construct validity and goodness-of-fit (e.g., χ²[162] = 346.8, Comparative Fit Index [CFI] = .975, Tucker-Lewis Index [TLI] = .971, Root Mean Square Error of Approximation [RMSEA] = .048, Standardized Root Mean Square Residual [SRMR] = .044). Internal consistency reliability across all subscales is robust, with Cronbach’s alpha coefficients consistently ranging from .810 to .907. Predictive validity testing confirms that elevated perceived exposure to algorithmic management prospectively predicts significant decreases in experienced job autonomy (β = −.24, p < .01) and job complexity (β = −.26, p < .01), corroborating prominent job design theories. The AMQ provides organizational psychologists, human resource management researchers, and workplace sociologists with a rigorous, standardized tool to evaluate the technological restructuring of labor, algorithmic governance, and employee psychological well-being.

Keywords

Algorithmic management, algorithmic control, artificial intelligence in HRM, work design, electronic performance monitoring, automated scheduling, automated performance evaluation, dynamic compensation, job autonomy, psychometrics.

Authors

The Algorithmic Management Questionnaire was conceptualized, operationalized, and psychometrically validated by an international team of organizational behavior and human resource management scholars:

• Xavier Parent-Rocheleau — Department of Human Resources Management, HEC Montréal, Montreal, Quebec, Canada. ORCID: 0000-0001-5015-3214. Corresponding Author: [email protected].
• Sharon K. Parker — Centre for Transformative Work Design, Future of Work Institute, Curtin University, Perth, Western Australia, Australia. ORCID: 0000-0002-0978-1873.
• Antoine Bujold — Department of Human Resources Management, HEC Montréal, Montreal, Quebec, Canada. ORCID: 0000-0002-6784-4895.
• Marie-Claude Gaudet — Department of Human Resources Management, HEC Montréal, Montreal, Quebec, Canada.

Purpose

The rapid integration of machine learning, automated decision-making systems (ADMS), and workforce analytics has substantially reconfigured the operational dynamics of contemporary workplaces. Originating within digital labor platforms (e.g., ride-hailing, on-demand food delivery, freelance micro-tasking), algorithmic management practices have progressively migrated into conventional organizational environments, including supply-chain logistics, retail, financial services, healthcare, and remote knowledge work. Algorithmic management involves the use of computer-programmed procedures, predictive modeling, and continuous data streams to coordinate, direct, evaluate, and discipline human labor with limited direct human managerial oversight.

Prior to the establishment of the AMQ, empirical research exploring algorithmic management was largely fragmented, relying on qualitative case studies, single-item proxies, or ad hoc questionnaires lacking formal psychometric validation. This empirical gap hindered the systematic comparison of findings across industries and obscured the specific psychological mechanisms through which automated oversight impacts worker strain, motivation, engagement, and retention. The AMQ was developed to fill this void by providing a standardized, multidimensional measurement model capable of capturing both the breadth and granularity of automated management techniques.

In occupational health psychology and organizational behavior research, the AMQ serves several critical functions:

• Quantifying Multidimensional Exposure: The instrument enables researchers to determine the precise degree and specific areas in which employees are subject to automated decision-making, acknowledging that algorithmic intervention is rarely uniform across all organizational functions.
• Evaluating Work Design Deterioration: By linking empirical AMQ scores to validated work characteristics (e.g., the Job Characteristics Model or the SMART work design framework), investigators can identify how algorithmic governance encroaches on psychological empowerment, decision latitude, skill variety, and relational architecture.
• Investigating Employee Well-Being and Stress: The scale allows researchers to examine how continuous automated monitoring, algorithmic pacing, and performance metrics correlate with technostress, work-related exhaustion, burnout, turnover intentions, and counterproductive work behaviors.
• Diagnostics for Organizational Auditing: Human resource professionals and organizational consultants can use the AMQ to benchmark algorithmic exposure, audit automated workforce systems for excessive control, and design interventions that rebalance computational efficiency with human dignity and psychological safety.

Psychological Construct

The AMQ measures workers’ perceived exposure to Algorithmic Management (AM), operationalized as an overarching higher-order construct encompassing the computational automation of core human resource and operational supervisory functions. Drawing from foundational organizational control literature and recent socio-technical conceptualizations (Kellogg et al., 2020; Parent-Rocheleau et al., 2024), the construct reflects technological systems that execute classical managerial imperatives: monitoring behaviors, formulating objectives, organizing workflows, conducting appraisals, and allocating financial incentives.

The construct is structured into five distinct, interrelated subdimensions:

1. Algorithmic Monitoring

Algorithmic monitoring captures the extent to which digital tools, enterprise software, telemetry, or algorithmic apparatuses continuously and unobtrusively track employees’ operational activities, temporal expenditures, physical movements, or digital footprints in real time. Rather than relying on periodic supervisory observation, algorithmic monitoring represents persistent, granular, automated surveillance. Examples include automated keystroke loggers, GPS telematics in fleet operations, active screen-capture software, and badge-based location tracking within physical distribution centers.

2. Algorithmic Goal Setting

This dimension operationalizes the degree to which production targets, performance benchmarks, delivery windows, or output quotas are calculated and assigned directly by automated algorithmic systems rather than through collaborative human negotiation. Automated goal setting frequently relies on dynamic historical datasets and real-time optimization models, establishing rigorous performance expectations with minimal accommodation for subjective contextual contingencies (e.g., automated parcel-per-hour expectations in fulfillment warehouses or platform-calculated trip completion deadlines).

3. Algorithmic Scheduling (Task Allocation)

Algorithmic scheduling evaluates the degree to which algorithmic protocols govern the dispatching, sequencing, routing, and temporal distribution of labor. This subscale measures automated task distribution, dynamic shift assignment based on predictive labor-demand algorithms, and computational determination of the precise order in which tasks must be executed. Employees governed by algorithmic scheduling experience computerized dispatching with minimal discretion over their daily workflow or schedule boundaries.

4. Algorithmic Performance Rating

Algorithmic performance rating reflects the automated computational processing of operational performance metrics to evaluate worker competence, adherence to operational standards, and relative ranking. This includes the automated generation of digital scorecards, predictive efficiency ratings, and algorithm-driven compliance tracking that determines whether an individual meets retention thresholds. In these environments, evaluation occurs without human managerial calibration, contextual advocacy, or qualitative dialogue.

5. Algorithmic Compensation

The compensation dimension captures workers’ perceptions that their direct financial remuneration, bonuses, task-based pay rates, or financial penalties are determined by computational algorithms. This manifests prominently in dynamic pricing, automated surge rates, algorithmic deduction of wages for minor non-compliance or delivery delays, and computational commission systems that compute financial outcomes based on algorithmic scorecards.

Theoretical Framework

The theoretical architecture of the Algorithmic Management Questionnaire is anchored in the integration of classic organizational sociology, cybernetic control models, and modern work design theory.

1. Labor Process Theory and Technical Control

Sociological models of the workplace (Edwards, 1979) identify an evolution of control mechanisms: from simple direct supervisory control to structural bureaucratic control, and subsequently to technical control. In technical control, the physical and digital apparatus of production itself directs and paces labor. Algorithmic management represents the digital evolution of technical control into what scholars term algorithmic control (Kellogg et al., 2020; Wood et al., 2019). Algorithmic control operates through three mechanisms: direction (automated task allocation and scheduling), evaluation (real-time algorithmic tracking and scorecards), and discipline (automated pay adjustments, dispatch de-prioritization, or algorithmic deactivation).

2. Work Design Theory and the SMART Model

The scale is closely tied to contemporary work design theory, particularly the extended formulations of the Job Characteristics Model (Hackman & Oldham, 1976) and the SMART work design framework (Parker et al., 2021). The SMART model highlights five essential characteristics of high-quality work: Stimulating (job complexity, skill variety), Mastery (role clarity, feedback), Autonomous (decision-making freedom, operational control), Relational (social support, social contact), and Tolerable (manageable demands, reasonable pacing).

Parent-Rocheleau et al. (2024) hypothesized that high algorithmic exposure fundamentally disrupts these work characteristics. Automated pacing, algorithmic allocation, and automated goal setting constrain employee decision latitude, reducing experienced job autonomy. Concurrently, the decomposition of tasks into computationally optimized steps can degrade task variety and cognitive problem-solving, reducing job complexity. The AMQ provides the empirical measurement needed to test these structural impacts.

3. Self-Determination Theory (SDT)

Deci and Ryan’s (2000) Self-Determination Theory posits that psychological health, intrinsic motivation, and vitality require the satisfaction of three basic psychological needs: autonomy, competence, and relatedness. Continuous algorithmic monitoring and automated evaluation often shift an individual’s perceived locus of causality from internal to external, undermining autonomous motivation. When performance is audited by algorithms that lack contextual empathy, workers may feel reduced competence and experience isolation due to the absence of human managerial connection.

Validity

The AMQ underwent comprehensive validation following standard six-phase psychometric scale development protocols (Hinkin, 1998; Parent-Rocheleau et al., 2024):

Content Validity

An initial pool of 36 candidate items (six items per postulated functional dimension) was generated through an exhaustive systematic literature review covering empirical and qualitative studies of digital platform work, algorithmic governance, and modern electronic performance monitoring (e.g., Lehdonvirta, 2018; Möhlmann et al., 2021; Rahman, 2021; Wood et al., 2019). Content validity was formally evaluated by a panel of subject-matter experts in organizational behavior, human resource management, and work design. Experts assessed each item for conceptual clarity, construct representativeness, domain coverage, and linguistic precision. Items demonstrating cognitive redundancy, semantic ambiguity, or low domain specificity were systematically refined or removed, resulting in the finalized 20-item instrument (four items per dimension).

Convergent Validity

Convergent validity was established by examining the statistical association between the AMQ subscales and conceptually related technological constructs drawn from the Technology Acceptance Model (TAM; Davis, 1989). As theoretically anticipated, exposure to algorithmic management exhibited moderate to strong positive correlations with perceived system usefulness (correlations ranging from r = .43 to .53, p < .001) and perceived ease of use (correlations ranging from r = .35 to .43, p < .001). Furthermore, average variance extracted (AVE) values across all five latent factors exceeded the recommended .50 threshold, confirming that the latent constructs account for a substantial majority of the variance in their respective indicator items.

Discriminant Validity

Discriminant validity was verified using both the Fornell and Larcker (1981) criterion and heterotrait-monotrait (HTMT) ratio evaluations. The square root of the AVE for each dimension was greater than any inter-factor correlation between that factor and any other construct. Additionally, the AMQ demonstrated low correlations (ranging from r = .03 to .22) with theoretically unrelated external criterion variables, confirming that the questionnaire captures algorithmic exposure rather than general organizational climate, negative affectivity, or broader technophobia.

Predictive and Criterion-Related Validity

Criterion and predictive validity were evaluated through longitudinal and cross-lagged structural equation modeling. Perceived exposure to algorithmic management at Time 1 significantly and negatively predicted employee-reported job autonomy at Time 2 (β = −.24, p < .01) and job complexity at Time 2 (β = −.26, p < .01), controlling for baseline organizational and demographic variables. These findings provide empirical support for the theoretical hypothesis that algorithmic control mechanisms can erode key work design characteristics.

Reliability

The AMQ displays high internal consistency reliability across diverse occupational sectors and participant cohorts (Parent-Rocheleau et al., 2024):

• Cronbach’s Alpha (α): Across development and validation samples, internal consistency coefficients for all five dimensions consistently exceeded the standard .70 and .80 thresholds:
  • Monitoring: α = .87 to .91
  • Goal Setting: α = .85 to .89
  • Scheduling: α = .81 to .86
  • Performance Rating: α = .86 to .90
  • Compensation: α = .84 to .88

  Overall scale reliability (all 20 items combined) yielded an alpha of α > .92.

• Composite Reliability (CR): Raykov’s composite reliability coefficients for each latent factor ranged from .821 to .912, demonstrating that the indicators are reliable across samples.
• Test-Retest Stability: Sub-samples assessed across multi-wave temporal intervals demonstrated stable test-retest coefficients (intraclass correlation coefficients ranging from .72 to .81 over a four-week span), indicating that the AMQ reliably captures sustained organizational working conditions rather than temporary work fluctuations.

Factor Analysis

The structural dimensionality of the AMQ was investigated using both Exploratory Factor Analysis (EFA) during initial item winnowing and Confirmatory Factor Analysis (CFA) across multiple validation samples.

Confirmatory Factor Analysis (CFA) Model Fit

In Sample 1 (comprising platform and gig-economy workers, N = 398), the hypothesized five-factor oblique measurement model demonstrated strong fit indices:

• χ² (160, N = 398) = 346.80, p < .001
• Comparative Fit Index (CFI) = .975
• Tucker-Lewis Index (TLI) = .971
• Root Mean Square Error of Approximation (RMSEA) = .048 (90% CI [.041, .055])
• Standardized Root Mean Square Residual (SRMR) = .044

To establish cross-sample generalizability, the five-factor model was cross-validated in Sample 2 (comprising workers across traditional corporate sectors exposed to automated HR tools, N = 352):

• χ² (161, N = 352) = 300.29, p < .001
• Comparative Fit Index (CFI) = .968
• Tucker-Lewis Index (TLI) = .962
• Root Mean Square Error of Approximation (RMSEA) = .049 (90% CI [.041, .057])
• Standardized Root Mean Square Residual (SRMR) = .052

Alternative Model Comparisons

Competing factor structures were tested against the hypothesized five-factor model using chi-square difference testing (Δχ²):

• Single-Factor Model (Harmon’s One-Factor test for common method bias): Collapsing all 20 items onto a single latent AM factor resulted in poor fit (χ²[170] = 1842.15, CFI = .642, TLI = .600, RMSEA = .158, SRMR = .135). The five-factor model showed superior fit (Δχ²[10] = 1495.35, p < .001).
• Second-Order Hierarchical Model: A second-order model (where the five first-order factors load onto a general algorithmic management superordinate construct) demonstrated acceptable fit (χ²[165] = 378.12, CFI = .969, TLI = .964, RMSEA = .051, SRMR = .049). This justifies calculating both individual subscale scores and an aggregate composite AM exposure score.

Standardized factor loadings across all 20 retained items were uniform, ranging from .71 to .89, with no significant cross-loadings, demonstrating clean factorial validity.

Instrument / Measurement Tool

• Instrument Name: Algorithmic Management Questionnaire (AMQ)
• Authors: Xavier Parent-Rocheleau, Sharon K. Parker, Antoine Bujold, and Marie-Claude Gaudet (2024)
• Target Population: Adult workforce populations (ages 18+) operating in platform-mediated, hybrid, automated, or digitally tracked employment settings.
• Administration Format: Self-report questionnaire; administered via paper-and-pencil or secure electronic survey platforms.
• Completion Time: Approximately 4 to 6 minutes.
• Total Number of Items: 20 items
• Subscale Structure (4 items per dimension):
  • Monitoring: Items 1, 2, 3, 4
  • Goal Setting: Items 5, 6, 7, 8
  • Scheduling: Items 9, 10, 11, 12
  • Performance Rating: Items 13, 14, 15, 16
  • Compensation: Items 17, 18, 19, 20
• Response Format: Standard 5-point Likert scale:
  • 1 = Strongly disagree
  • 2 = Disagree
  • 3 = Neither agree nor disagree
  • 4 = Agree
  • 5 = Strongly agree

  (Alternative validated response anchor: 1 = To a very small extent, to 5 = To a very large extent)

• Scoring and Interpretation Procedures:
  • Subscale Scores: Calculated by summing and averaging the 4 items corresponding to each dimension (Score range: 1.00 to 5.00).
  • Overall Composite Score: Calculated by averaging all 20 items (Score range: 1.00 to 5.00). Higher scores indicate greater exposure to algorithmic control.
  • Reverse-Scoring: None. All items are positively keyed to reflect the presence of algorithmic management.

Permissions & Fee and Test Year

The Algorithmic Management Questionnaire was published in 2024 in Human Resource Management (Parent-Rocheleau et al., 2024). The scale is copyrighted by John Wiley & Sons, Ltd., and the authors. The authors permit academic, scientific, and non-commercial research use of the scale without royalty fees, provided appropriate bibliographic citation is maintained. Commercial applications, organizational auditing tools, enterprise software integration, or fee-for-service consulting deployments require written permission and formal licensing from the primary author or publisher. For inquiries regarding licensing or academic adaptations, contact Dr. Xavier Parent-Rocheleau at HEC Montréal ([email protected]).

References

• Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319–340. https://doi.org/10.2307/249008
• Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268. https://doi.org/10.1207/S15327965PLI1104_01
• Edwards, R. (1979). Contested terrain: The transformation of the workplace in the twentieth century. Basic Books.
• Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39–50. https://doi.org/10.1177/002224378101800104
• Hackman, J. R., & Oldham, G. R. (1976). Motivation through the design of work: Test of a theory. Organizational Behavior and Human Performance, 16(2), 250–279. https://doi.org/10.1016/0030-5073(76)90016-7
• Hinkin, T. R. (1998). A brief tutorial on the development of measures for use in survey questionnaires. Organizational Research Methods, 1(1), 104–121. https://doi.org/10.1177/109442819800100106
• Kellogg, K. C., Valentine, M. A., & Christin, A. (2020). Algorithms at work: The new contested terrain of control. Academy of Management Annals, 14(1), 366–410. https://doi.org/10.5465/annals.2018.0174
• Lehdonvirta, V. (2018). Flexibility in the gig economy: Managing time on three online piecework platforms. New Technology, Work and Employment, 33(1), 13–29. https://doi.org/10.1111/ntwe.12102
• Möhlmann, M., Zalmanson, L., Henfridsson, O., & Feldberg, N. (2021). Algorithmic management of work on online labor platforms: When matching meets control. MIS Quarterly, 45(4), 1999–2022. https://doi.org/10.25300/MISQ/2021/15333
• Parent-Rocheleau, X., Parker, S. K., Bujold, A., & Gaudet, M.-C. (2024). Creation of the Algorithmic Management Questionnaire: A six-phase scale development process. Human Resource Management, 63(1), 25–44. https://doi.org/10.1002/hrm.22185
• Parker, S. K., Knight, C., & Keller, A. (2021). Work design: A return to fundamentals and an agenda for the future. Annual Review of Organizational Psychology and Organizational Behavior, 8, 287–308. https://doi.org/10.1146/annurev-orgpsych-012420-091455
• Rahman, H. A. (2021). The invisible cage: Workers’ reactivity to opaque algorithmic evaluations in the gig economy. Administrative Science Quarterly, 66(4), 945–988. https://doi.org/10.1177/00018392211010118
• Wood, A. J., Graham, M., Lehdonvirta, V., & Hjorth, I. (2019). Good gig, bad gig: Autonomy and algorithmic control in the global gig economy. Work, Employment and Society, 33(1), 56–75. https://doi.org/10.1177/0018726717718504

Items of the Scale