Frost Multidimensional Perfectionism Scale (FMPS)

Abstract

The Frost Multidimensional Perfectionism Scale (FMPS), developed by Randy O. Frost and colleagues in 1990, represents a landmark psychometric instrument designed to assess perfectionism as a multifaceted, cognitive-behavioral construct. Prior to the introduction of the FMPS, clinical and personality psychology predominantly conceptualized perfectionism as a unidimensional, inherently pathological trait centered on the setting of excessively high standards. Frost and his co-investigators challenged this paradigm by positing that perfectionism encompasses both personal achievement orientations and critical evaluative tendencies, deeply rooted in familial and cognitive dynamics. The original instrument consists of 35 self-report items evaluated on a 5-point Likert scale, organized into six distinct subscales: Concern over Mistakes (CM), Personal Standards (PS), Parental Expectations (PE), Parental Criticism (PC), Doubts about Actions (D), and Organization (O). Psychometric evaluations across diverse clinical and non-clinical populations demonstrate robust internal consistency, with overall scale Cronbach’s alpha coefficients typically exceeding .90, and subscale coefficients ranging from .72 to .93. Factor analytic studies have supported both the original six-factor model and an empirically refined four-factor structure proposed by Joachim Stöber, alongside higher-order bifactor models that distinguish between Evaluative Concerns Perfectionism (often maladaptive or neurotic) and Personal Strivings Perfectionism (often adaptive or conscientious). The FMPS remains one of the most widely cited and psychometrically validated instruments for evaluating perfectionism across clinical psychology, psychiatry, behavioral medicine, and educational research, particularly in relation to obsessive-compulsive disorder, eating disorders, major depressive disorder, and academic burnout.

Keywords

Frost Multidimensional Perfectionism Scale, FMPS, perfectionism, psychometrics, evaluative concerns, personal standards, concern over mistakes, doubts about actions, factor analysis, construct validity

Authors

The Frost Multidimensional Perfectionism Scale was developed by a team of clinical psychologists and psychometric researchers at Smith College in Northampton, Massachusetts:

• Randy O. Frost, Ph.D. — Professor Emeritus of Psychology at Smith College. Dr. Frost is an internationally recognized authority on the cognitive-behavioral phenomenology of perfectionism, hoarding disorder, and obsessive-compulsive phenomena.
• Patricia Marten, B.A. — Department of Psychology, Smith College. Co-investigator on the foundational empirical validation studies establishing the multidimensional structure of perfectionism.
• Cathleen Lahart, B.A. — Department of Psychology, Smith College. Contributed to item generation, developmental conceptualization, and initial clinical sample validation.
• Robin Rosenblate, B.A. — Department of Psychology, Smith College. Co-author of the seminal 1990 validation study, focusing on the psychometric evaluation of familial antecedents in perfectionistic pathology.

Subsequent psychometric refinements, structural factor evaluations, and shortened iterations were spearheaded by leading international researchers, notably Dr. Joachim Stöber at the University of Kent, alongside Australian psychometricians Nigar G. Khawaja, Kerry A. Armstrong, Colleen C. Hawkins, Helen M. G. Watt, and Kenneth E. Sinclair.

Purpose

The primary purpose of the Frost Multidimensional Perfectionism Scale is to provide an empirically sound, nuanced, and structurally differentiated measurement of perfectionism. Historically, psychological literature—influenced heavily by earlier diagnostic frameworks and measures such as the Burns Perfectionism Scale (Burns, 1980)—treated perfectionism as a singular, maladaptive personality trait characterized by rigid cognitive schemas and an irrational demand for flawlessness. Frost, Marten, Lahart, and Rosenblate (1990) argued that such unidimensional models failed to capture the complex architecture of perfectionism, notably the critical distinction between setting ambitious personal standards and the debilitating, self-critical fear of failing to meet them.

From a theoretical standpoint, the FMPS was constructed to disentangle the cognitive, behavioral, and developmental facets of perfectionism. The authors recognized that while high personal standards may accompany healthy striving, high achievement, and professional conscientiousness, the psychological burden of perfectionism primarily stems from evaluative vulnerabilities: excessive concern over making errors, perpetual uncertainty regarding the quality of one’s actions, and the internalized belief that parental affection, acceptance, and approval are strictly conditional upon flawless execution.

In clinical practice, the FMPS serves as an indispensable diagnostic and therapeutic assessment tool. Clinicians utilize the scale to:

• Identify cognitive vulnerabilities that precipitate and perpetuate cognitive-behavioral pathology in patients presenting with anorexia nervosa, bulimia nervosa, social anxiety disorder, and chronic mood disorders.
• Differentiate between clinical perfectionism—which warrants intensive cognitive restructuring and exposure interventions—and benign, goal-directed conscientiousness that supports academic or occupational productivity.
• Track therapeutic progress across longitudinal interventions by monitoring reductions in maladaptive evaluative concerns (e.g., Concern over Mistakes and Doubts about Actions) while supporting resilient, realistic self-standards.

In research environments, the FMPS provides investigators with the granularity required to conduct structural equation modeling, behavioral genetics studies, and cross-cultural analyses regarding how perfectionism intersects with self-esteem, procrastination, perceived stress, athletic performance, and psychological well-being.

Psychological Construct

Perfectionism, within the framework of the FMPS, is conceptualized not as a monolithic construct, but as a multidimensional disposition encompassing six distinct yet interconnected components. Each dimension captures a unique psychological, cognitive, or developmental facet:

1. Concern over Mistakes (CM)

Comprising 9 items, Concern over Mistakes represents the core pathological engine of perfectionism. This dimension reflects a respondent’s propensity to interpret mistakes as catastrophic failures and to believe that any deficit in performance inevitably leads to a loss of respect, social rejection, and personal worthlessness. Individuals scoring high on CM exhibit rigid, dichotomous thinking (“all-or-nothing” cognition), equating a partial failure with complete, irreversible failure (e.g., Item 14: “If I fail partly, it is as bad as being a complete failure”). CM is the dimension most robustly correlated with clinical psychopathology, including depressive symptoms, generalized anxiety, and suicidal ideation.

2. Personal Standards (PS)

Comprising 7 items, Personal Standards assesses the setting of exceptionally high self-expectations and goals, alongside an excessive emphasis on those standards for self-evaluation (e.g., Item 12: “I set higher goals than most people”). While elevated PS scores can drive adaptive achievement and occupational competence, they become pathological when an individual’s sense of self-esteem is exclusively contingent upon relentlessly meeting these demanding benchmarks.

3. Parental Expectations (PE)

Comprising 5 items, Parental Expectations evaluates the individual’s perception that their parents established unrealistically high standards and demanded nothing short of perfection (e.g., Item 1: “My parents set very high standards for me”; Item 15: “Only outstanding performance is good enough in my family”). This developmental dimension reflects the systemic and familial pressures that foster introjected conditions of worth.

4. Parental Criticism (PC)

Comprising 4 items, Parental Criticism measures the respondent’s historical perception that their parents were excessively critical, punitive, or withholding of affection when their performance was perceived as imperfect (e.g., Item 3: “As a child, I was punished for doing things less than perfect”; Item 5: “My parents never tried to understand my mistakes”). Elevated PC scores frequently co-occur with insecure attachment styles and core beliefs centered on unworthiness.

5. Doubts about Actions (D)

Comprising 4 items, Doubts about Actions reflects an enduring sense of uncertainty regarding the quality, completeness, or adequacy of one’s performance (e.g., Item 17: “Even when I do something very carefully, I often feel that it is not quite right”). Individuals with elevated D scores demonstrate cognitive hesitation, persistent checking behaviors, and chronic task repetition, exhibiting striking phenomenology shared with obsessive-compulsive checking compulsions.

6. Organization (O)

Comprising 6 items, Organization reflects an individual’s preference for orderliness, neatness, structured routines, and methodical precision (e.g., Item 2: “Organization is very important to me”; Item 7: “I am a neat person”). In their foundational work, Frost et al. (1990) noted that although organization is an observable behavioral correlate of perfectionism, it demonstrates negligible or even negative correlations with the other five subscales and psychopathology. Consequently, psychometricians frequently treat Organization as a separate stylistic index and exclude it when calculating the overall composite perfectionism score.

Theoretical Framework

The development of the FMPS rests upon foundational psychological paradigms established across psychodynamic, developmental, cognitive-behavioral, and social learning traditions.

Hamachek’s Typology of Perfectionism

A central theoretical catalyst for the FMPS was Don E. Hamachek’s seminal 1978 paper differentiating between normal and neurotic perfectionism. Hamachek posited that normal perfectionists set high goals but derive genuine pleasure from their strivings, maintain flexibility when circumstances demand, and recognize their human limitations. In contrast, neurotic perfectionists set demands beyond their reach, can never feel satisfied with their achievements, are driven by an intense fear of failure rather than a pursuit of success, and live in perpetual dread of making mistakes. Frost et al. operationalized this distinction psychometrically by designing items that isolate neurotic evaluative concerns (CM, D, PC) from achievement-oriented strivings (PS, O).

Beck’s Cognitive Model and Dysfunctional Schemas

The cognitive underpinnings of the FMPS align directly with Aaron T. Beck‘s cognitive theory of depression and anxiety. According to Beck, psychological distress arises from deeply entrenched core cognitive schemas characterized by cognitive distortions, such as dichotomous (black-and-white) thinking, selective abstraction, and catastrophizing. The FMPS explicitly captures these cognitive distortions. For instance, the CM subscale operationalizes the cognitive vulnerability wherein any error is interpreted as proof of global personal unworthiness, demonstrating how maladaptive perfectionists filter their lived experience through rigid conditional operational rules (“If I make a mistake, I am an utter failure”).

Social Learning and Familial Socialization

The incorporation of Parental Expectations (PE) and Parental Criticism (PC) reflects Albert Bandura‘s social learning theory and developmental attachment frameworks. Perfectionism does not emerge in a vacuum; it is heavily reinforced through direct conditioning, vicarious modeling, and parental socialization practices. Children raised by highly critical or demanding caregivers internalize the belief that love, warmth, and validation are contingent upon external performance. Over time, these external parental demands are converted into internalized, autonomous cognitive structures that self-punish any deviation from perfection.

Comparative Framework: Frost et al. vs. Hewitt & Flett

Contemporaneously with Frost et al., Canadian researchers Paul L. Hewitt and Gordon L. Flett (1991) developed the Multidimensional Perfectionism Scale (HF-MPS), which operationalized perfectionism through interpersonal and directional orientations: Self-Oriented Perfectionism (SOP), Other-Oriented Perfectionism (OOP), and Socially Prescribed Perfectionism (SPP). While the HF-MPS emphasizes the relational direction of perfectionistic demands (toward self, toward others, or perceived from society), the FMPS provides a deeper cognitive-phenomenological assessment of the individual’s inner evaluative processes (mistake concern, action doubts) and their developmental family antecedents.

Validity

The psychometric validity of the FMPS has been established across hundreds of empirical studies involving clinical cohorts, university students, adolescents, and competitive athletic populations worldwide.

Construct and Factorial Validity

Construct validity is evidenced by the scale’s coherent structural alignment with theoretical expectations. Exploratory and confirmatory factor analyses consistently validate the distinctiveness of the six underlying dimensions. Furthermore, higher-order structural modeling confirms that the subscales load systematically onto two overarching latent constructs: Evaluative Concerns Perfectionism (ECP; comprising Concern over Mistakes, Doubts about Actions, Parental Criticism, and Parental Expectations) and Personal Strivings Perfectionism (PSP; comprising Personal Standards and Organization).

Convergent Validity

The convergent validity of the FMPS is supported by strong, statistically significant correlations with alternative perfectionism measures and theoretically aligned psychological constructs:

• Hewitt & Flett MPS: The FMPS Concern over Mistakes and Parental Expectations subscales correlate strongly with Socially Prescribed Perfectionism (r = .50 to .65), while the Personal Standards subscale exhibits high convergence with Self-Oriented Perfectionism (r = .60 to .75).
• Eating Disorder Inventory (EDI): Frost et al. (1990) and subsequent researchers demonstrated substantial positive correlations between the FMPS total score, CM, and the Perfectionism subscale of the EDI (r > .60).
• Affective Pathology: Concern over Mistakes and Doubts about Actions consistently correlate moderately to strongly with the Beck Depression Inventory (BDI; r = .40 to .55) and the State-Trait Anxiety Inventory (STAI; r = .45 to .60).
• Obsessive-Compulsive Phenomenology: Doubts about Actions and Concern over Mistakes demonstrate robust associations with obsessive checking, contamination fears, and symmetry impulses on the Padua Inventory and the Maudsley Obsessional Compulsive Inventory (MOCI; r = .35 to .52).

Discriminant and Predictive Validity

Discriminant validity has been demonstrated by showing that while ECP subscales correlate heavily with neuroticism, negative affectivity, and shame-proneness, the Personal Standards and Organization subscales correlate positively with conscientiousness, positive affect, mastery goals, and academic self-efficacy. Importantly, the FMPS demonstrates clinical predictive validity: elevated baseline scores on Concern over Mistakes and Doubts about Actions prospectively predict poor response to standard cognitive-behavioral treatment in clinical cohorts suffering from major depressive disorder and bulimia nervosa.

Reliability

Extensive psychometric investigations confirm that the FMPS possesses high internal consistency reliability and substantial temporal stability across diverse populations.

Internal Consistency

In the original validation study by Frost et al. (1990), which examined 106 female undergraduate students, the total scale demonstrated an exceptional Cronbach’s alpha coefficient of .90. Subscale reliability coefficients were robust across both the original study and subsequent large-scale psychometric evaluations (e.g., Stöber, 1998; Hawkins et al., 2006; Khawaja & Armstrong, 2005):

• Concern over Mistakes (CM): α = .87 to .91
• Personal Standards (PS): α = .78 to .83
• Parental Expectations (PE): α = .84 to .88
• Parental Criticism (PC): α = .77 to .84
• Doubts about Actions (D): α = .72 to .77
• Organization (O): α = .86 to .93

Stöber (1998) examined a sample of 243 German university students using the translated FMPS and reported virtually identical internal consistency metrics: Total Scale α = .90, CM α = .87, PS α = .78, PE α = .88, PC α = .81, D α = .73, and O α = .86. In adolescent cohorts, Hawkins, Watt, and Sinclair (2006) confirmed that these coefficients remain stable across age groups, reporting alpha coefficients between .76 and .92 among Australian secondary school girls.

Test-Retest Reliability

Studies evaluating the temporal stability of the FMPS over intervals ranging from four weeks to six months demonstrate test-retest correlation coefficients consistently between r = .75 and r = .88 across the subscales. The developmental subscales (Parental Expectations and Parental Criticism) exhibit the highest temporal stability (r > .85), reflecting their retrospective, historically anchored nature. Even the cognitively reactive subscales, such as Concern over Mistakes, maintain remarkable trait-like stability in the absence of targeted psychological intervention.

Factor Analysis

The underlying factor structure of the FMPS has been the subject of extensive empirical scrutiny, leading to valuable structural refinements and higher-order conceptual models.

Original Exploratory Factor Analysis

Frost et al. (1990) conducted an exploratory factor analysis (EFA) using principal components extraction followed by oblique rotation on their 35-item pool. This analysis cleanly extracted six factors with eigenvalues exceeding 1.0, accounting for significant total variance. Items loaded cleanly onto their conceptual targets, with factor loadings generally exceeding .45, and minimal cross-loadings, establishing the canonical six-factor model.

Stöber’s Four-Factor Model Refinement

In a seminal re-evaluation, Joachim Stöber (1998) utilized confirmatory factor analysis (CFA) to address psychometric redundancies and cross-loadings observed within the original scale. Stöber demonstrated that a condensed four-factor model yielded superior parsimony and model fit indices compared to the six-factor solution:

• Concern over Mistakes and Doubts about Actions (CMDA): Combining the items of CM and D into a unified construct reflecting cognitive fear of failure and performance hesitation.
• Parental Expectations and Criticism (PEPC): Merging the familial items of PE and PC into a single developmental factor representing adverse parental evaluative pressure.
• Personal Standards (PS): Retaining the independent achievement striving items.
• Organization (O / ORG): Retaining the distinct orderliness items.

Stöber’s four-factor structure demonstrated improved goodness-of-fit parameters (χ²/df ratio < 2.0, Comparative Fit Index [CFI] > .92, Root Mean Square Error of Approximation [RMSEA] < .06) and has been replicated widely, particularly in international validation studies (e.g., Khawaja & Armstrong, 2005).

Higher-Order Two-Factor Model

Subsequent psychometric consensus supports a higher-order structural framework bifurcating perfectionism into:

1. Maladaptive Evaluative Concerns Perfectionism (ECP): Subsumes Concern over Mistakes, Doubts about Actions, Parental Criticism, and Parental Expectations. This overarching factor is consistently linked to distress, neuroticism, perfectionistic rumination, and clinical disorders.
2. Adaptive Personal Strivings Perfectionism (PSP): Subsumes Personal Standards and Organization. This factor demonstrates strong loadings on conscientiousness, intrinsic motivation, and athletic or academic excellence.

Instrument / Measurement Tool

The complete technical attributes and administration standards of the FMPS are structured as follows:

• Test Type: Self-report psychological inventory; multidimensional personality assessment.
• Target Population: Adolescents (ages 13 and older) and adults across both clinical and non-clinical settings.
• Number of Items: 35 items.
• Administration Time: Approximately 5 to 10 minutes.
• Response Scale: 5-point Likert scale ranging from 1 = Strongly Disagree (or “not at all true”) to 5 = Strongly Agree (or “very true”).
• Subscale Item Distribution:
  • Concern over Mistakes (CM): 9 items (Items 9, 10, 13, 14, 18, 21, 23, 25, 34).
  • Personal Standards (PS): 7 items (Items 4, 6, 12, 16, 19, 24, 30).
  • Parental Expectations (PE): 5 items (Items 1, 11, 15, 20, 26).
  • Parental Criticism (PC): 4 items (Items 3, 5, 22, 35).
  • Doubts about Actions (D): 4 items (Items 17, 28, 32, 33).
  • Organization (O): 6 items (Items 2, 7, 8, 27, 29, 31).
• Scoring Procedures:
  • All items are scored in a direct positive direction (1 to 5). There are no reverse-coded items.
  • Subscale Scores: Calculated by summing the item ratings within each designated dimension. Mean item ratings (sum divided by item count) can also be used to facilitate direct cross-subscale comparisons.
  • Total Perfectionism Score (FMPS-Total): Calculated by summing the scores of the 29 items across CM, PS, PE, PC, and D (score range: 29 to 145). In alignment with Frost et al. (1990) and subsequent psychometric guidelines, the 6 items of the Organization (O) subscale are excluded from the total composite score because Organization does not correlate with core perfectionistic pathology.
  • Evaluative Concerns Composite: Often calculated in clinical research by summing CM, D, PC, and PE (22 items; score range: 22 to 110).
• Interpretation Guidelines: Higher scores denote more intense manifestations of the respective perfectionistic dimension. Scores on the composite FMPS-Total or CM subscale exceeding 1.5 standard deviations above non-clinical normative means indicate significant clinical perfectionism warranting therapeutic evaluation.

Permissions & Fee and Test Year

The Frost Multidimensional Perfectionism Scale was originally published in 1990 in the peer-reviewed journal Cognitive Therapy and Research (Springer Nature). As an academic psychological instrument, the FMPS is widely accessible for educational, non-commercial, and clinical research purposes under standard academic fair-use conventions, provided full formal citation is given to Dr. Randy O. Frost and colleagues.

There are no licensing fees or royalties mandated for individual academic researchers or clinicians administering the questionnaire in routine non-commercial research or standard psychological practice. However, commercial distribution, integration into fee-for-service digital assessment platforms, or reproduction within published testing batteries requires explicit written permission from the copyright holder (Springer Nature / original author Dr. Randy O. Frost). For permissions, inquiries should be directed to the permissions department of Springer Science+Business Media or the Department of Psychology at Smith College.

References

• Burns, D. D. (1980). The perfectionist’s script for self-defeat. Psychology Today, 14(6), 34–52.
• Frost, R. O., Marten, P., Lahart, C., & Rosenblate, R. (1990). The dimensions of perfectionism. Cognitive Therapy and Research, 14(5), 449–468. https://doi.org/10.1007/BF01172967
• Hamachek, D. E. (1978). Psychodynamics of normal and neurotic perfectionism. Psychology: A Journal of Human Behavior, 15(1), 27–33.
• Hawkins, C. C., Watt, H. M. G., & Sinclair, K. E. (2006). Psychometric properties of the Frost Multidimensional Perfectionism Scale with Australian adolescent girls. Educational and Psychological Measurement, 66(6), 1001–1021. https://doi.org/10.1177/0013164406288164
• Hewitt, P. L., & Flett, G. L. (1991). Perfectionism in deficits in self-regulation and psychological distress. Journal of Personality and Social Psychology, 60(3), 456–470. https://doi.org/10.1037/0022-3514.60.3.456
• Khawaja, N. G., & Armstrong, K. A. (2005). Factor structure and psychometric properties of the Frost Multidimensional Perfectionism Scale: Developing shorter versions using an Australian sample. Australian Journal of Psychology, 57(2), 129–138. https://doi.org/10.1080/00049530500048704
• Stöber, J. (1998). The Frost Multidimensional Perfectionism Scale: More perfect with four (instead of six) dimensions. Personality and Individual Differences, 24(4), 481–491. https://doi.org/10.1016/S0191-8869(97)00207-9

Items of the Scale