Physical Symptoms Inventory (PSI)

Abstract

The Physical Symptoms Inventory (PSI) is an established psychometric assessment developed by Paul E. Spector and Steve M. Jex in 1998 to evaluate somatic strain and functional physical health complaints within occupational and health psychology frameworks. Originating from the foundational need to quantify physiological stress manifestations without confounding them with affective or psychiatric symptomatology, the PSI measures the frequency and healthcare-seeking behavioral correlates of common physical complaints experienced over a specified recall period (typically the past 30 days). The full instrument comprises 18 items, while abbreviated 13-item and 12-item variants have been extensively deployed in organizational research. Items encompass a diverse spectrum of somatic complaints, including gastrointestinal disturbances, musculoskeletal pain, sleep disruption, cardiovascular reactivity, neurological symptoms, and general fatigue. Respondents rate items on a 5-point frequency scale ranging from 1 (Not at all) to 5 (Every day), often coupled with a dichotomous secondary probe indicating whether a medical consultation occurred for each symptom. Psychometrically, the instrument exhibits acceptable to high internal consistency (Cronbach’s alpha coefficients typically ranging from .72 to .85 across diverse occupational samples), despite the theoretical classification of somatic symptoms as causal indicators (formative measurement) rather than traditional reflective latent constructs. Criterion and construct validity are robustly evidenced through moderate-to-strong associations with workplace stressors (such as interpersonal conflict, quantitative workload, and organizational constraints), negative affectivity, and physiological indices of allostatic load. This article provides an exhaustive examination of the PSI’s conceptual origins, theoretical paradigms, psychometric properties, factor structure, scoring algorithms, and academic applications.

Keywords

Physical Symptoms Inventory, PSI, somatic strain, occupational stress, job stressor-strain model, psychosomatic symptoms, psychometrics, physical health complaints, Paul Spector, Steve Jex, occupational health psychology

Authors

The Physical Symptoms Inventory was developed by:

• Paul E. Spector, Ph.D. — Distinguished Professor Emeritus of Industrial and Organizational Psychology at the University of South Florida, Tampa, Florida, United States. Dr. Spector is an internationally renowned scholar in occupational health psychology, job stress, workplace aggression, and psychometric measurement methodologies.
• Steve M. Jex, Ph.D. — Professor of Industrial and Organizational Psychology, previously at Central Michigan University and Bowling Green State University, and currently affiliated with the University of Central Florida. Dr. Jex is widely recognized for his research on organizational stress, coping mechanisms, workplace mistreatment, and employee well-being.

The primary validation of the instrument was published in the landmark 1998 paper titled “Development of Four Self-Report Measures of Job Stressors and Strain: Interpersonal Conflict at Work Scale, Organizational Constraints Scale, Quantitative Workload Inventory, and Physical Symptoms Inventory” in the Journal of Occupational Health Psychology.

Purpose

The primary purpose of the Physical Symptoms Inventory (PSI) is to provide a brief, psychometrically sound, and reliable measurement tool for assessing somatic manifestations of psychological strain. In occupational health psychology, organizational behavior, and behavioral medicine, researchers and practitioners require instruments capable of disentangling physical strain from purely psychological distress (such as anxiety, depression, or emotional exhaustion). Many legacy instruments utilized in stress research suffered from significant limitations: they were either excessively lengthy (e.g., standard clinical symptom checklists exceeding 90 items), proprietary and cost-prohibitive, heavily confounded with psychiatric diagnosis, or saturated with items measuring subjective emotional states rather than verifiable bodily symptoms.

Spector and Jex (1998) addressed these limitations by constructing the PSI to fulfill three distinct objectives:

1. Non-Redundancy with Affective Strain: The inventory deliberately excludes explicit psychological or affective terms (e.g., “felt nervous,” “felt hopeless”), focusing exclusively on functional physical symptoms (e.g., headache, backache, upset stomach). This architectural separation prevents artificial inflation of correlations between perceived environmental stressors and somatic outcomes caused by shared affective variance or negative affectivity (Watson & Clark, 1984).
2. Dual-Assessment Framework: The scale permits investigators to measure not only the subjective occurrence frequency of physical symptoms but also an objective behavioral marker of symptom severity: whether or not the respondent sought medical intervention (“Did you see a doctor for this?”). This dual-metric approach offers clinicians and organizational diagnosticians insight into healthcare utilization patterns and direct organizational healthcare costs.
3. Applicability Across Occupational and Non-Occupational Cohorts: Designed with neutral, universally understood physiological descriptors, the PSI is equally applicable to manufacturing workers, healthcare professionals, corporate executives, military personnel, and university students. It serves as an effective screening instrument in occupational health surveillance, ergonomics assessments, and longitudinal stress-intervention trials.

Psychological Construct

The construct assessed by the Physical Symptoms Inventory is somatic strain, specifically manifested as functional, non-specific somatic symptoms resulting from sustained psychophysiological arousal. Somatic strain represents the physiological component of the stress response, wherein prolonged exposure to psychosocial stressors triggers sustained autonomic, endocrine, and immunological dysregulation.

Rather than reflecting a singular organic disease state, the PSI operationalizes somatic strain as a constellation of bodily symptoms across several interrelated physiological domains:

• Gastrointestinal Disturbances: Items evaluating nausea/upset stomach, acid indigestion/heartburn, stomach cramps, diarrhea, and constipation reflect the direct sensitivity of the enteric nervous system to stress. Psychosocial stress disrupts the brain-gut axis, altering gastrointestinal motility, mucosal blood flow, and visceral sensitivity through corticotropin-releasing factor (CRF) signaling.
• Musculoskeletal Discomfort: The backache item targets musculoskeletal tension. Sustained sympathetic activation and continuous low-level motor unit recruitment under cognitive or emotional load lead to localized muscle ischemia, tissue hypoxia, and myofascial pain syndromes.
• Sleep Architecture Disruption: The “trouble sleeping” item captures insomnia, delayed sleep onset, and non-restorative sleep patterns induced by nocturnal hyperarousal and elevated evening cortisol levels.
• Neurological and Sensory Symptoms: Headaches (primarily tension-type and stress-triggered migraines), dizziness, and eye strain reflect autonomic vascular dysregulation, sustained ocular convergence during computer-based tasks, and central nervous system sensory amplification.
• Cardiorespiratory Reactivity: Symptoms such as chest pain, shortness of breath, and heart pounding when not exercising index acute autonomic reactivity, specifically hyperventilation patterns and sympathetic dominance over parasympathetic vagal tone.
• Systemic and Immunological Vulnerability: Items assessing skin rashes, infections, fever, loss of appetite, and chronic tiredness or fatigue gauge generalized systemic wear-and-tear, neuroendocrine exhaustion, and secondary immunosuppression caused by chronic glucocorticoid resistance.

Theoretical Framework

The Physical Symptoms Inventory is anchored within multiple prominent conceptual paradigms across stress physiology and industrial-organizational psychology:

1. The Job Stressor-Strain Model

Formulated by occupational stress researchers (Beehr & Newman, 1978; Spector, 1998), the Job Stressor-Strain Model posits an environmental causal sequence. Workplace stressors (objective environmental conditions such as workload, ambiguity, and conflict) are cognitively appraised by the employee, eliciting short-term psychological, behavioral, and physiological reactions. When stressors persist without adequate recovery, these acute reactions crystallize into chronic physical strains, which the PSI directly captures.

2. The Allostatic Load Paradigm

Developed by Bruce McEwen and Eliot Stellar (1993), the theory of allostatic load explains how chronic psychological stress causes multi-system physiological exhaustion. Under chronic threat, the sympathetic-adrenal-medullary (SAM) axis and the hypothalamic-pituitary-adrenal (HPA) axis are repeatedly activated. Over time, excessive release of catecholamines and glucocorticoids damages cardiovascular tissue, dysregulates immune mediators, and alters visceral sensitivity, translating into the exact somatic complaints cataloged in the PSI.

3. The Formative (Causal Indicator) Measurement Theory

A critical psychometric foundation emphasized by Spector and Jex (1998), citing Kenneth Bollen and Lennox (1991), is that somatic symptoms operate conceptually as causal (formative) indicators rather than traditional reflective (effect) indicators. In classical test theory, items of a scale are assumed to be reflections of an underlying latent construct; thus, high inter-item correlations are mathematically required. However, physical symptoms do not necessarily co-occur due to an underlying common factor: a worker under severe stress may develop severe back pain and insomnia without experiencing skin rashes or diarrhea. Therefore, somatic strain is a composite index of physiological strain markers, where each individual symptom contributes additively to the overall strain burden.

Validity

Empirical studies conducted over more than two decades have established extensive construct, convergent, discriminant, and criterion-related validity for the Physical Symptoms Inventory across varied organizational and international cohorts:

Construct and Convergent Validity

In their initial validation across multiple heterogeneous samples (totaling over 1,000 employees), Spector and Jex (1998) reported statistically significant, robust positive correlations between the PSI and several concurrently validated stressor scales:

• Interpersonal Conflict at Work Scale (ICAWS): Correlations with PSI frequency scores ranged from r = .23 to .38 (p < .001).
• Organizational Constraints Scale (OCS): Correlations ranged from r = .28 to .41 (p < .001).
• Quantitative Workload Inventory (QWI): Correlations ranged from r = .24 to .35 (p < .001).
• Role Ambiguity and Role Conflict: Significant positive associations (r values typically .20 to .35).

Furthermore, convergent validity is documented through substantial correlations with other validated somatic measures, such as the Pennebaker Inventory of Limbic Languidness (PILL; r > .60) and the Somatization subscale of the Brief Symptom Inventory (BSI-18).

Discriminant Validity

Discriminant validity has been confirmed by evaluating the PSI alongside measures of trait affectivity and job attitudes. Although negative affectivity (NA) correlates with symptom reporting (common across all self-report methodologies, typically r = .25 to .40), partial correlation analyses show that the relationships between organizational stressors and the PSI remain statistically significant after controlling for trait negative affect. Moreover, while job dissatisfaction and intention to turnover demonstrate modest correlations with the PSI (r = .20 to .30), exploratory and confirmatory factor analyses demonstrate that somatic strain items load separately from attitudinal and cognitive constructs.

Criterion-Related and Predictive Validity

The behavioral medical consultation component of the PSI demonstrates direct criterion-related validity. Spector and Jex (1998) found that the doctor-visit score correlated positively with objective sick leave, absenteeism records, and prescription medication usage. In longitudinal occupational health studies, baseline PSI scores have predicted subsequent long-term sickness absence, early retirement intentions, and cardiovascular clinical events over multi-year follow-up periods.

Reliability

The reliability of the Physical Symptoms Inventory has been scrutinized across numerous field studies and laboratory experiments:

Internal Consistency

In the seminal 1998 investigation, Spector and Jex reported internal consistency reliability estimates (Cronbach’s alpha) for the frequency scoring method ranging from α = .72 to .85 across various working samples, with a mean alpha of approximately .80 across the studies examined. For the doctor-visit scoring method, internal consistency is slightly lower (typically .65 to .75), attributable to the low base rate of seeking professional medical care for minor, self-limiting symptoms (e.g., eye strain or transient heartburn).

Importantly, psychometric theorists (Bollen & Lennox, 1991; Spector & Jex, 1998) note that Cronbach’s alpha should not be used as the sole benchmark for evaluating the PSI. Because physical symptoms can be independent biological reactions to stress, a lower internal consistency coefficient does not indicate poor instrument validity; rather, it reflects the heterogeneous physiological expression of human stress responses.

Test-Retest Stability

Temporal stability estimates depend heavily on the recall window utilized. When administered with the standard 30-day recall timeframe, test-retest correlations over a 2- to 4-week interval typically range from r = .68 to .78, indicating adequate temporal stability while remaining sensitive to fluctuations in environmental stressors. Daily diary and ecological momentary assessment (EMA) adaptations utilizing adapted 24-hour recall windows demonstrate appropriate within-person daily variance responsive to day-level workplace conflict and acute workload peaks.

Factor Analysis

The underlying dimensionality of the Physical Symptoms Inventory has been investigated through both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA):

Exploratory Factor Structure

When unconstrained exploratory factor analyses are conducted on the 18 items, the PSI typically yields either a dominant single general factor or a multi-factor solution comprising three to five correlated subdimensions:

1. Gastrointestinal Factor: High loadings (> .55) from upset stomach, acid indigestion, stomach cramps, diarrhea, and constipation.
2. Fatigue / Musculoskeletal / Sleep Factor: High loadings from backache, trouble sleeping, and chronic tiredness/fatigue.
3. Cardiorespiratory and Autonomic Factor: High loadings from chest pain, heart pounding, shortness of breath, and dizziness.
4. Head / Sensory Factor: Loadings from headache and eye strain.
5. Immune / Infectious Factor: Loadings from fever, infection, and skin rash.

Confirmatory Factor Analysis and Structural Models

In structural equation modeling (SEM), researchers frequently treat the PSI as a unidimensional composite strain indicator. Fit indices for a strict single-factor reflective CFA model often yield modest fit (e.g., RMSEA ≈ .07–.09, CFI ≈ .85–.90) due to localized item dependencies (e.g., strong residual covariance between “diarrhea” and “constipation,” or “acid indigestion” and “upset stomach”). When modeled with correlated error terms between anatomically related symptom pairs or structured as a bifactor model (one general somatic strain factor alongside domain-specific group factors), model fit improves substantially (CFI > .94, TLI > .92, RMSEA < .05).

However, from a formative index perspective, many psychometricians advocate modeling the PSI as a formative construct or composite variable, wherein individual symptom indicators directly cause the somatic strain composite without the assumption that items must correlate highly or share equivalent factor loadings.

Instrument / Measurement Tool

• Full Instrument Name: Physical Symptoms Inventory
• Acronym: PSI
• Primary Developers: Paul E. Spector, Ph.D., and Steve M. Jex, Ph.D. (1998)
• Construct Measured: Somatic strain, functional physical health complaints, and associated medical consultation behavior.
• Target Population: Working adults, clinical outpatients, university students, and general community populations.
• Administration Modality: Self-report questionnaire administered via paper-and-pencil, online survey platforms, or ecological momentary assessment (EMA) mobile applications.
• Administration Time: Approximately 3 to 5 minutes for the 18-item full scale; 2 to 3 minutes for abbreviated versions.
• Available Versions:
  • 18-Item Version: Comprehensive inventory covering gastrointestinal, neurological, musculoskeletal, cardiorespiratory, and systemic symptoms.
  • 13-Item Version: Abbreviated scale developed for large-scale survey batteries, retaining high-incidence items (e.g., including ringing in the ears).
  • 12-Item Version: Streamlined version omitting redundant or low-frequency indicators to reduce survey fatigue.
• Response Format: Dual-response frequency and doctor-visit evaluation:
  • Frequency Response Scale (Mandatory 5-point format): 1 = Not at all, 2 = Once or twice, 3 = Once or twice per week, 4 = Several times per week, 5 = Every day.
  • Healthcare Utilization Probe (Optional dichotomous question per item): “Did you see a doctor for this?” (Scored: No = 0, Yes = 1; or integrated into categorical options: No, Yes, but I didn’t see doctor, Yes, and I saw doctor).
• Recall Period: Standard instructions specify “During the past 30 days” (can be modified to “past week” or “today” for longitudinal/diary investigations).
• Scoring Protocols:
  • Symptom Frequency Score (Sum or Mean): Summing the item ratings (yielding a range of 18 to 90 for the 18-item version) or computing the mean across completed items (ranging from 1.0 to 5.0). Higher scores denote greater somatic strain.
  • Symptom Count Score: Counting the total number of distinct symptoms experienced at least once during the recall window (dichotomizing responses: 1 = symptom absent [score of 1]; 1 = symptom present [scores of 2 through 5]). Range: 0 to 18.
  • Doctor Visit Score: Summing the total number of symptoms for which a medical doctor was consulted. Range: 0 to 18.
• Reverse-Scoring: None. All items are positively keyed toward somatic complaints.

Permissions & Fee and Test Year

The Physical Symptoms Inventory was first formally published in 1998 by Paul E. Spector and Steve M. Jex in the Journal of Occupational Health Psychology.

Under the conditions established by the copyright holder, Dr. Paul E. Spector, the PSI is classified as an open-access psychometric instrument for academic, non-commercial research, and educational purposes. Researchers may administer, reproduce, and translate the inventory without paying licensing fees or seeking explicit written permission, provided that:

1. The instrument is used solely for non-commercial research, institutional evaluation, or educational training.
2. Full and proper academic attribution is provided in all resulting papers, theses, and publications citing Spector and Jex (1998).
3. The items and scale structures are not resold, repackaged into commercial for-profit consulting tools, or copyrighted by third parties.

Commercial entities, corporate consulting firms, and clinical commercial testing vendors intending to use the scale for profit must obtain permission from the authors. Official instructions, conditions of use, and downloadable survey copies are maintained on Dr. Paul Spector’s official academic assessment repository hosted by the University of South Florida.

References

• Beehr, T. A., & Newman, J. E. (1978). Job stress, employee health, and organizational effectiveness: A facet analysis, model, and literature review. Personnel Psychology, 31(4), 665–699. https://doi.org/10.1111/j.1744-6570.1978.tb02118.x
• Bollen, K., & Lennox, R. (1991). Conventional wisdom on measurement: A structural equation perspective. Psychological Bulletin, 110(2), 305–314. https://doi.org/10.1037/0033-2909.110.2.305
• Demerouti, E., Bakker, A. B., Nachreiner, F., & Schaufeli, W. B. (2001). The job demands-resources model of burnout. Journal of Applied Psychology, 86(3), 499–512. https://doi.org/10.1037/0021-9010.86.3.499
• Keenan, A., & Newton, T. J. (1985). Stressful events, stressors, and psychological strains in young professional engineers. Journal of Occupational Behaviour, 6(2), 151–156. https://doi.org/10.1002/job.4030060207
• Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. Springer Publishing Company.
• McEwen, B. S. (1998). Protective and damaging effects of stress mediators. New England Journal of Medicine, 338(3), 171–179. https://doi.org/10.1056/NEJM199801153380307
• McEwen, B. S., & Stellar, E. (1993). Stress and the individual: Mechanisms leading to disease. Archives of Internal Medicine, 153(18), 2093–2101. https://doi.org/10.1001/archinte.1993.00410180039004
• Peters, L. H., & O’Connor, E. J. (1980). Situational constraints and work outcomes: The influences of a frequently overlooked construct. Academy of Management Review, 5(3), 391–397. https://doi.org/10.5465/amr.1980.4288856
• Spector, P. E. (1998). A control-theory of the job stress process. In C. L. Cooper (Ed.), Theories of organizational stress (pp. 153–169). Oxford University Press.
• Spector, P. E., & Jex, S. M. (1998). Development of four self-report measures of job stressors and strain: Interpersonal Conflict at Work Scale, Organizational Constraints Scale, Quantitative Workload Inventory, and Physical Symptoms Inventory. Journal of Occupational Health Psychology, 3(4), 356–367. https://doi.org/10.1037/1076-8998.3.4.356
• Watson, D., & Clark, L. A. (1984). Negative affectivity: The disposition to experience aversive emotional states. Psychological Bulletin, 96(3), 465–490. https://doi.org/10.1037/0033-2909.96.3.465

13. Items of the Scale (Questionnaire)