Mobility Inventory (MI)

Abstract

The Mobility Inventory for Agoraphobia (MI), developed by Dianne L. Chambless and colleagues in 1985, is one of the most widely utilized and psychometrically validated self-report instruments for assessing agoraphobic avoidance behavior. Designed to measure behavioral avoidance across a spectrum of situations and public spaces, the instrument uniquely evaluates two distinct ecological contexts: avoidance When Accompanied by a trusted companion and avoidance When Alone. The primary section comprises 27 situational items (encompassing public places, open expanses, enclosed environments, public transit, driving, and perceived confinement) evaluated on a 5-point Likert scale ranging from 1 (“never avoid”) to 5 (“always avoid”). Beyond these contextual ratings, the instrument incorporates supplemental clinical indices, including the identification of the respondent’s five most debilitating situations, past-week panic attack frequency and intensity, and the qualitative boundaries of the individual’s safe perimeter (“safety zone”). Decades of empirical investigation have demonstrated that the Mobility Inventory exhibits exceptional internal consistency (Cronbach’s $\alpha$ frequently exceeding .95 for the Alone subscale and .90 for the Accompanied subscale), robust test-retest reliability, and pronounced sensitivity to treatment effects in cognitive-behavioral therapy and pharmacotherapy. Furthermore, the MI possesses strong convergent, discriminant, and criterion-related validity, reliably distinguishing patients with agoraphobia from healthy individuals and clinical controls diagnosed with other anxiety disorders. This comprehensive review examines the psychometric foundation, structural dimensions, theoretical underpinnings, diagnostic utility, and practical administration of the Mobility Inventory.

Keywords

Mobility Inventory, Mobility Inventory for Agoraphobia, MIA, agoraphobic avoidance, panic disorder, behavioral assessment, fear of fear, exposure therapy, clinical psychometrics, avoidance behavior

Authors

The Mobility Inventory was originated and psychometrically established by a distinguished team of clinical psychologists and biostatisticians:

• Dianne L. Chambless, Ph.D. (1948–2023) — Professor Emerita of Psychology at the University of Pennsylvania. A pioneering authority in clinical psychology, Dr. Chambless was internationally renowned for her empirical investigations into panic disorder, agoraphobia, evidence-based psychological interventions, and the “fear-of-fear” conceptualization.
• G. Craig Caputo, Ph.D. — Clinical psychologist specializing in cognitive-behavioral interventions and anxiety disorder management.
• Susan E. Jasin, Ph.D. — Clinical psychologist and researcher contributing extensively to the clinical characterization and behavioral tracking of phobic avoidance.
• Edward J. Gracely, Ph.D. — Associate Professor of Family, Community & Preventive Medicine and Biostatistics at Drexel University College of Medicine, contributing foundational statistical modeling, metric construction, and psychometric validation.
• Christine Williams, M.A. — Researcher in experimental and clinical psychology who collaborated on the scale’s initial clinical validation cohorts at Temple University and the Medical College of Pennsylvania.

Purpose

The Mobility Inventory for Agoraphobia was engineered to address a pervasive diagnostic and assessment deficit in the clinical literature of the early 1980s. Existing global anxiety scales and general phobia inventories (such as the Fear Questionnaire) often obscured the precise nature of agoraphobic avoidance by failing to distinguish between independent mobility and accompanied mobility. Clinicians and behavioral scientists recognized that individuals with agoraphobia rarely exhibit absolute, invariant immobility; rather, their avoidance operates as a dynamic continuum governed by interpersonal proximity, perceived escape routes, and safety cues.

The primary clinical and empirical purposes of the MI encompass:

• Ecological Dissection of Avoidance: Systematically quantifying the functional differential between an individual’s behavioral capacity when operating independently (When Alone) versus when accompanied by a trusted individual (When Accompanied). This dual-condition structure provides insight into the patient’s reliance on interpersonal safety signals.
• Treatment Baseline Formulation and Target Selection: Identifying specific environmental cues (e.g., supermarkets, bridges, expressways, open courtyards) that provoke escape responses, enabling the formulation of hierarchical in vivo exposure therapy protocols.
• Sensitivity to Therapeutic Change: Serving as a primary outcome measure in randomized controlled trials (RCTs) evaluating cognitive-behavioral interventions, interoceptive exposure, acceptance-based paradigms, and psychopharmacological agents. Because the MI uses a graded 5-point avoidance continuum across common public environments, it captures micro-level clinical reductions in avoidance before an individual achieves complete spatial freedom.
• Multi-Dimensional Symptom Profiling: Supplementing behavioral avoidance data with continuous parameters regarding panic frequency, perceived panic attack severity, and spatial restriction (the “safety zone” radius).

Psychological Construct

The core psychological construct measured by the Mobility Inventory is agoraphobic avoidance, operationalized as the deliberate evasion of spaces, settings, or situations where escape might be psychologically or physically difficult, or where prompt medical assistance would be unobtainable in the event of incapacitating autonomic arousal or a panic attack. The MI evaluates this construct not as a homogeneous or static phobic state, but as a multi-determined behavioral coping mechanism structured across distinct contextual domains:

1. Avoidance When Alone (MI-Alone / AAL)

The Avoidance When Alone dimension measures the respondent’s behavioral constraint when bereft of external safety cues or human anchors. This scale captures the unmitigated operational severity of agoraphobia. In the unaccompanied state, the individual’s cognitive appraisals of catastrophic internal sensations (such as tachycardia, lightheadedness, derealization, or hyperventilation) are met without the compensatory distress-buffering provided by a companion. Consequently, scores on the MI-Alone subscale are virtually always higher than those on the Accompanied subscale, reflecting baseline catastrophic appraisal and anticipatory dread in everyday travel.

2. Avoidance When Accompanied (MI-Accompanied / AAC)

The Avoidance When Accompanied dimension measures the persistence of avoidance even in the presence of a trusted friend, relative, or partner. This construct captures extreme agoraphobic severity. When an individual avoids environmental spaces (such as large auditoriums, wide open fields, or subterranean transit) despite having an attachment figure present, it signals severe fear conditioning and behavioral inhibition that overwhelms interpersonal safety signals.

3. Environmental Sub-domains

Empirical analyses delineate several ecological sub-domains within the 27 items:

• Enclosed and Constricted Spaces: Items such as elevators, tunnels, and small interior chambers tap claustrophobic fears interlocked with escape obstruction.
• Public and Commercial Venues: Supermarkets, department stores, shopping malls, restaurants, and classrooms reflect sensory overload, crowd congestion, and fear of social embarrassment or behavioral incapacitation.
• Open Spaces and Unanchored Terrains: Outside fields, wide streets, and large interior courtyards measure fear of open expanses without physical handrails, walls, or immediate shelter.
• Transportation and Distance Barriers: Expressways, buses, subways, trains, airplanes, and boats capture entrapment between geographic transit stops, speed constraints, and distance from home or hospital centers.

4. Ancillary Phenomenological Constructs

The MI contextualizes avoidance through discrete evaluations of recent autonomic surges (panic attacks over the prior 7 days and their qualitative severity) and the “Safety Zone.” The safety zone measures the geographical radius around the domestic perimeter beyond which anxiety escalates past manageable thresholds, providing a spatial representation of the patient’s functional confinement.

Theoretical Framework

The conceptual architecture of the Mobility Inventory rests primarily on the behavioral and cognitive theories of agoraphobia formulated during the late 20th century, particularly the “Fear of Fear” Model developed by Goldstein and Chambless (1978) and further substantiated by Chambless and Gracely (1989).

Under this theoretical framework, agoraphobia develops as a secondary, learned behavioral adaptation to unexpected panic attacks or intense autonomic surges. According to Mowrer’s Two-Factor Theory of avoidance learning, the initial conditioning event involves classical conditioning: harmless physiological fluctuations (interoceptive cues) and specific physical environments (exteroceptive cues) become paired with catastrophic terror and sympathetic arousal. Subsequently, the maintenance of the condition is sustained via operant conditioning: escaping or actively avoiding these settings immediately terminates anticipatory dread, thereby strongly reinforcing avoidant behavior through negative reinforcement.

The inclusion of the Accompanied versus Alone contrast directly reflects Stanley Rachman‘s concepts regarding safety signals and attachment-based threat mitigation. Rachman postulated that the presence of a trusted companion serves as an inhibitory safety stimulus, attenuating subjective fear and autonomic arousal by dampening perceived vulnerability. When a safety signal is present, the individual predicts a lower probability of catastrophic outcomes (such as fainting, dying, or public humiliation without rescue). Therefore, measuring both states allows researchers and clinicians to quantify the precise magnitude of the safety-signal effect.

Furthermore, the MI aligns with cognitive models of panic articulated by David M. Clark and David H. Barlow (Clark’s Cognitive Model and Barlow’s Triple Vulnerability Model). In these frameworks, the catastrophic misinterpretation of bodily sensations leads the patient to engage in avoidance behaviors designed to evade situations where panic is deemed likely or dangerous. The MI provides an external, behavioral readout of these underlying cognitive threats across spatial settings.

Validity

The Mobility Inventory has been evaluated across clinical cohorts, community samples, and diverse cultural populations, demonstrating excellent validity parameters:

Convergent Validity

The MI displays strong, statistically significant correlations with other gold-standard measures of agoraphobia and phobic avoidance. In the foundational studies by Chambless et al. (1985), the MI-Alone and MI-Accompanied scales correlated robustly with the Agoraphobia subscale of the Fear Questionnaire (FQ-Agoraphobia; $r = .65$ to $.78$). Subsequent psychometric appraisals (Chambless et al., 2011; Rodriguez et al., 2011) revealed strong convergent associations with the Panic Disorder Severity Scale (PDSS; $r = .55$ to $.68$), the Agoraphobic Cognitions Questionnaire (ACQ; $r = .48$ to $.62$), and the Body Sensations Questionnaire (BSQ; $r = .42$ to $.56$).

Discriminant Validity

The instrument clearly differentiates agoraphobic avoidance from general depressive symptomatology and circumscribed social fears. While moderate correlations exist between the MI and the Beck Depression Inventory (BDI; typically $r = .30$ to $.40$, reflecting the common secondary demoralization seen in housebound patients), these values are significantly lower than its associations with specific panic and agoraphobia metrics. Furthermore, correlations between the MI and measures of simple specific phobias or generalized social anxiety (such as the Social Phobia Scale) remain modest, confirming that the scale assesses agoraphobic fear of entrapment and panic rather than generic social evaluation apprehension.

Criterion-Related and Known-Groups Validity

The MI reliably discriminates between individuals diagnosed with panic disorder with agoraphobia (PDA), panic disorder without agoraphobia, social anxiety disorder, obsessive-compulsive disorder, and non-clinical controls. Agoraphobic cohorts routinely score multiple standard deviations above normative samples ($p < .001$).

Sensitivity to Treatment Change

The MI is highly sensitive to therapeutic interventions. In clinical trials of cognitive-behavioral therapy, virtual reality exposure, and selective serotonin reuptake inhibitors (SSRIs), both MI-Alone and MI-Accompanied scores demonstrate substantial, statistically significant decreases corresponding to large effect sizes (Cohen’s $d > 1.2$). Decreases in the Alone subscale frequently mirror clinical recovery and freedom of mobility.

Reliability

The psychometric reliability of the Mobility Inventory is exceptionally strong across independent international investigations:

Internal Consistency

Across both initial validation and replication cohorts, the items forming the Avoidance When Alone and Avoidance When Accompanied subscales demonstrate remarkable homogeneity:

• Avoidance When Alone (MI-Alone): Cronbach’s $\alpha$ coefficients consistently range from .94 to .97 across both clinical and non-clinical populations.
• Avoidance When Accompanied (MI-Accompanied): Cronbach’s $\alpha$ coefficients regularly span .90 to .95.
• Item-total correlations for situational items predominantly fall between .50 and .82, demonstrating that each included scenario contributes substantively to the overall measurement of the avoidance construct.

Test-Retest Reliability

The temporal stability of the MI has been confirmed over various time frames under non-treatment conditions:

• Over a short interval of one to two weeks, test-retest reliability coefficients ($r$) exceed .85 to .90 for both subscales.
• Over extended baseline assessment intervals (e.g., four to six weeks preceding waitlist control conditions), the stability coefficients remain high ($r = .75$ to $.86$), confirming that situational avoidance ratings do not show major spontaneous fluctuations in the absence of targeted psychological or pharmacological treatment.

Factor Analysis

Extensive exploratory factor analyses (EFA) and confirmatory factor analyses (CFA) conducted over the past three decades have yielded important insights into the internal architecture of the Mobility Inventory:

Unidimensional vs. Multi-Factorial Paradigms

In their initial 1985 psychometric study, Chambless and colleagues conducted principal components analysis with varimax rotation. Although a single primary general factor accounted for the vast majority of common variance (supporting the use of the overarching composite mean scores for MI-Alone and MI-Accompanied), their structural analysis identified meaningful sub-factors:

• Factor I: Public Places and Public Enclosures (loadings ranging from .58 to .83 on items such as supermarkets, shopping malls, department stores, auditoriums, and restaurants).
• Factor II: Open Spaces and Unanchored Outdoor Environments (loadings between .60 and .79 on items such as wide streets, courtyards, fields, and walking on the street).
• Factor III: Enclosed Spaces and Travel Obstacles (loadings between .52 and .77 on elevators, high places, tunnels, subways, and bridges).

Confirmatory Structural Modeling

Subsequent psychometric evaluations (e.g., Arrindell, 1993; Rodriguez et al., 2011) evaluated higher-order structural configurations. Confirmatory factor analysis has demonstrated that a hierarchical model—featuring a broad higher-order Agoraphobic Avoidance factor overarching three or four correlated lower-order contextual factors—demonstrates acceptable goodness-of-fit indices (Comparative Fit Index [CFI] $ge .92$, Tucker-Lewis Index [TLI] $ge .91$, and Root Mean Square Error of Approximation [RMSEA] $le .06$). Despite the theoretical validity of these sub-domains, psychometric authorities universally recommend retaining the global Alone and Accompanied composite scores for clinical and evaluative decision-making, as individual situational items share substantial inter-item covariance driven by core panic-related apprehensions.

Instrument / Measurement Tool

• Instrument Name: Mobility Inventory (MI) / Mobility Inventory for Agoraphobia (MIA)
• Instrument Type: Self-administered clinical questionnaire / Behavioral rating scale
• Target Population: Adults and adolescents presenting with agoraphobia, panic disorder, or situational avoidance
• Administration Time: Approximately 10 to 15 minutes
• Scale Structure:
  • Part 1: Situational Avoidance (Items 1–27): 26 standardized environmental locations and situations plus 1 self-defined situation (“Other”), each rated independently across two parallel environmental conditions: When Accompanied and When Alone.
  • Part 2: Priority Target Delineation: Qualitative or categorical selection of the 5 situations where avoidance or anxiety most negatively impacts daily functioning.
  • Part 3: Panic Attack Metrics: Standardized diagnostic definition of panic attack, past 7-day numerical frequency count, and a 5-point panic severity rating (1 = Very mild to 5 = Extremely).
  • Part 4: Safety Zone Assessment: Qualitative description of the presence, specific location, and geographic radius of the respondent’s perceived home safety zone.
• Response Format (Part 1): 5-point Likert-type scale:
  • 1 = Never avoid
  • 2 = Rarely avoid
  • 3 = Avoid about half of the time
  • 4 = Avoid most of the time
  • 5 = Always avoid
  • (Note: Respondents are instructed to leave blank any situations that do not apply to them.)
• Scoring Rules:
  • Avoidance When Accompanied (MI-Acc): Calculated by summing the ratings for all answered items in the “When Accompanied” column and dividing by the total number of items endorsed in that column (excluding blank or non-applicable items). Yields a mean score ranging from 1.00 to 5.00.
  • Avoidance When Alone (MI-Alone): Calculated by summing the ratings for all answered items in the “When Alone” column and dividing by the total number of items endorsed in that column (excluding blank or non-applicable items). Yields a mean score ranging from 1.00 to 5.00.
  • Handling Missing/Inapplicable Items: If an individual never travels by boat or airplane for reasons unrelated to anxiety, the item is left blank and omitted from both the numerator and denominator, preserving metric validity.

Permissions & Fee and Test Year

The Mobility Inventory for Agoraphobia was initially published in 1985 by Dianne L. Chambless, G. Craig Caputo, Susan E. Jasin, Edward J. Gracely, and Christine Williams in the journal Behaviour Research and Therapy.

The instrument is in the public domain for academic, clinical, research, and non-commercial educational purposes. The authors made the scale freely accessible to foster clinical assessment and outcome research in anxiety disorders. Clinicians and clinical researchers may administer, reproduce, and score the Mobility Inventory without the payment of royalties or licensing fees, provided that appropriate scholarly attribution is accorded to the original 1985 publication. Commercial publishers or electronic health record (EHR) platforms seeking proprietary integration should verify standard permissions via the copyright holder of the journal article (Elsevier).

References

• Arrindell, W. A. (1993). The fear of fear concept: Stability, internal structural validity and predictive value in agoraphobic outpatients. Behaviour Research and Therapy, 31(2), 161–178. https://doi.org/10.1016/0005-7967(93)90069-B
• Chambless, D. L., Caputo, G. C., Jasin, S. E., Gracely, E. J., & Williams, C. (1985). The Mobility Inventory for Agoraphobia. Behaviour Research and Therapy, 23(1), 35–44. https://doi.org/10.1016/0005-7967(85)90140-8
• Chambless, D. L., & Gracely, E. J. (1989). Fear of fear and the anxiety disorders. Cognitive Therapy and Research, 13(1), 9–20. https://doi.org/10.1007/BF01178486
• Chambless, D. L., Sharpless, B. A., Rodriguez, D., McCarthy, K. S., Milrod, B. L., Khalsa, S. R., & Barber, J. P. (2011). Psychometric properties of the Mobility Inventory for Agoraphobia: Convergent, discriminant, and criterion-related validity. Behavior Therapy, 42(4), 689–599. https://doi.org/10.1016/j.beth.2011.03.001
• Goldstein, A. J., & Chambless, D. L. (1978). A analysis of agoraphobia. Behavior Therapy, 9(1), 47–59. https://doi.org/10.1016/S0005-7894(78)80053-9
• Rodriguez, D., Chambless, D. L., & Sharpless, B. A. (2011). Factor analysis of the Mobility Inventory for Agoraphobia. Journal of Psychopathology and Behavioral Assessment, 33(4), 433–444. https://doi.org/10.1007/s10862-011-9243-7

Items of the Scale