Frost Multidimensional Perfectionism Scale (FMPS)

The FMPS was developed from the framework that perfectionism is not a single trait, but a multifaceted construct encompassing excessively high personal standards, concern over mistakes, doubts about the quality of one’s actions, perceptions of parental expectations and criticism, and a preference for order and organisation. Unlike unidimensional measures of perfectionism, the FMPS captures both the potentially adaptive and maladaptive dimensions of perfectionistic behaviour, making it well-suited for clinical formulation and treatment planning in clinical settings.

FMPS Subscales

In contemporary clinical use, the FMPS is commonly interpreted using four subscales (Stöber, 1998):

• Concern over Mistakes and Doubts about Actions (13 items: 9, 10, 13, 14, 17, 18, 21, 23, 25, 28, 32, 33, 34): items reflect the tendency to interpret mistakes as equivalent to failure, to believe that one will lose the respect of others following failure, and to experience persistent uncertainty about the quality of one’s actions. Example items include “If I fail partly, it is as bad as being a complete failure” and “I usually have doubts about the simple everyday things I do.” Elevations on this subscale are most consistently associated with psychological distress and impairment.
• Parental Expectations and Criticism (9 items: 1, 3, 5, 11, 15, 20, 22, 26, 35): items assess the perception that one’s parents set excessively high standards and were overly critical. Example items include “My parents set very high standards for me” and “I never felt like I could meet my parents’ expectations.” Elevations may reflect developmental or interpersonal contributors to perfectionistic beliefs.
• Personal Standards (7 items: 4, 6, 12, 16, 19, 24, 30): items capture the setting of excessively high personal goals and the importance placed on meeting those standards. Example items include “I have extremely high goals” and “I set higher goals than most people.” Elevations may reflect achievement striving. In isolation, this dimension may not indicate clinical concern; consider co-occurring evaluative concerns.
• Organisation (6 items: 2, 7, 8, 27, 29, 31): items measure the importance placed on order, neatness, and organisation. Example items include “Neatness is very important to me” and “I am an organised person.”

Clinical Applications of the FMPS

The FMPS is a useful tool when perfectionism may be contributing to psychological distress or maintaining symptoms. Perfectionism has been identified as a transdiagnostic process implicated in a range of mental health conditions, including depression, anxiety disorders, obsessive-compulsive disorder, and eating disorders (Egan et al., 2011).

The multidimensional structure of the FMPS allows clinicians to:

• Differentiate between self-critical perfectionism (Concern over Mistakes and Doubts about Actions) and high personal standards (Personal Standards). For example, a client presenting with procrastination and indecisiveness may show elevations on the Concern over Mistakes and Doubts about Actions subscale, reflecting a fear of making errors that leads to avoidance and checking behaviours. Conversely, a client with high Personal Standards but low Concern over Mistakes may be exhibiting a more adaptive form of perfectionism which is less strongly linked to psychological distress.
• Identify possible developmental influences (Parental Expectations and Criticism).
• Distinguish organisation preferences from clinically impairing perfectionism.

The subscale profile can also inform treatment targets. Cognitive-behavioural interventions for perfectionism (e.g., Shafran et al., 2018) typically target maladaptive cognitions captured by the Concern over Mistakes and Doubts subscale. When Parental Expectations and Criticism scores are elevated, this may suggest that perfectionism developed in response to early relational experience, and developmental or schema-focused approaches may be more appropriate.

FMPS Total and Subscale Scores

All items are scored on a 5-point Likert scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree). Higher scores indicate greater endorsement of perfectionistic beliefs and behaviours in a particular domain. The FMPS produces a Total Perfectionism score and four subscale scores.

The Total Perfectionism score is calculated as the sum of three of the four subscales (29 items; range 29–145):

• Concern over Mistakes and Doubts about Actions (13 items: 9, 10, 13, 14, 17, 18, 21, 23, 25, 28, 32, 33, 34; score range 13–65): reflects the tendency to equate mistakes with failure, fear loss of respect, and doubt the quality of one’s performance.
• Parental Expectations and Criticism (9 items: 1, 3, 5, 11, 15, 20, 22, 26, 35; score range 9–45): reflects perceived parental pressure to be perfect and criticism for falling short.
• Personal Standards (7 items: 4, 6, 12, 16, 19, 24, 30; score range 7–35): reflects the setting of excessively high personal standards and goals.
• The Organisation subscale (6 items: 2, 7, 8, 27, 29, 31; score range 6–30), which reflects the importance placed on order, neatness, and organisation, is excluded from the Total Perfectionism score. High organisation is not inherently problematic; however, high Organisation combined with elevated scores on the other subscales may exacerbate perfectionistic dysfunction.

FMPS Percentiles

All scores are presented as raw totals as well as percentiles, so clinicians can see both the client’s actual responses and how those scores compare with a normative sample. Percentile ranks are calculated based on Stöber’s (1998) normative sample of university students (N = 243, mean age 26.30). A percentile of 50 indicates an average level of perfectionism relative to this comparison group. Given that the normative sample consists of university students, percentiles should be interpreted with appropriate caution when applied to clinical populations, individuals from different age groups, or those from culturally diverse backgrounds. Percentiles provide a relative comparison point rather than absolute clinical thresholds.

FMPS Descriptors

The following percentile-anchored descriptors are used to aid interpretation of FMPS scores. These bands were developed by NovoPsych to assist interpretation, and should be treated as a practical guide rather than empirically validated clinical thresholds.

These descriptors are based on where a score falls in the normative distribution, rather than on the response scale labels (Strongly Disagree to Strongly Agree). This is because those labels reflect how strongly someone agrees with a statement, not how clinically elevated their perfectionism is.

• Very Low (≤5th percentile, score 29–44): Scores at this level are well below the normative average, suggesting minimal endorsement of perfectionistic beliefs or behaviours in this domain.
• Low (6th–24th percentile, score 45–60): Scores at this level are below the normative average, suggesting relatively low levels of perfectionism in this domain.
• Average (25th–75th percentile, score 61–83): Scores at this level fall within the typical range for the normative sample, suggesting perfectionism levels consistent with the general population.
• High (76th–94th percentile, score 84–99): Scores at this level are above the normative average, suggesting elevated perfectionism in this domain that may warrant clinical consideration.
• Very High (≥95th percentile, score 100–145): Scores at this level are well above the normative average, suggesting highly elevated perfectionism in this domain that may be clinically significant.

FMPS Tracking Progress

In the absence of a formally validated Reliable Change Index (RCI) or minimally important difference (MID) for the FMPS, changes of at least 0.5 standard deviations in scores may suggest potentially meaningful change (Norman et al., 2003; Turner et al., 2010).

Using the Stöber (1998) normative data, the following score changes may signal meaningful change:

• Total Perfectionism: approximately 9 points
• Concern over Mistakes and Doubts about Actions: approximately 6 points
• Parental Expectations and Criticism: approximately 5 points
• Personal Standards: approximately 3 points
• Organisation: approximately 3 points

These thresholds should be interpreted as approximate guides rather than definitive benchmarks, as they are derived from a non-clinical normative sample rather than treatment outcome data.

FMPS Graphs

Results are also presented in a table that lists raw scores, percentiles, and descriptors for Total Perfectionism and all four subscales.

On first administration, a percentile bar chart is presented, displaying the percentile scores for each subscale with descriptor bands (Very Low, Low, Average, Average, High, Very High) as background shading. Each bar reflects the client’s score on a 0–100 scale relative to a community sample.

Because all scores are converted to percentiles, each subscale is displayed on the same scale. This allows clinicians to quickly see which domains are most elevated, regardless of differences in the number of items per subscale. Higher bars indicate greater endorsement of perfectionistic traits in that domain relative to the normative sample.

FMPS Multiple Administration Graph

On subsequent administrations, a line graph displays subscale percentile scores over time, with descriptor bands as background shading, allowing clinicians to track changes across administrations.

FMPS Validity

The FMPS demonstrates good construct validity through its pattern of associations with theoretically related constructs. Frost et al. (1990) reported that overall perfectionism was significantly correlated with depression (r = .46), obsessive-compulsive symptoms (r = .36), and anxiety (r = .44) as measured by the Brief Symptom Inventory. The Concern over Mistakes subscale showed the strongest associations with psychopathology, particularly depression (r = .39) and general symptom distress severity (PSDI r = .48), while Doubts about Actions correlated with procrastination (r = .40; Stöber, 1998) and self-critical depression (r = .61; Frost et al., 1990). Conversely, Personal Standards showed positive associations with achievement orientation and goal-directed behaviour, supporting the distinction between maladaptive and more adaptive dimensions of perfectionism.

Discriminant validity was supported by the differential patterns of association across subscales. Organisation showed only non-significant associations with psychopathology, and Personal Standards demonstrated weak correlations with distress measures (Frost et al., 1990; Stöber, 1998). Howell et al. (2020) further supported this distinction through bifactor modelling, demonstrating that a general perfectionism factor was positively associated with depression (β = .447), anxiety (β = .376), and stress (β = .481), while a perfectionistic strivings group factor showed a negative association with depression (β = −.217).

Internal Consistency of the FMPS

The FMPS demonstrates good to excellent internal consistency across studies. Frost et al. (1990) reported Cronbach’s α values of .88 for Concern over Mistakes, .83 for Personal Standards, .84 for Parental Expectations, .84 for Parental Criticism, .77 for Doubts about Actions, and .93 for Organisation. Stöber (1998) reported α = .88 for the combined Concern over Mistakes and Doubts factor, α = .89 for the combined Parental Expectations and Criticism factor, α = .78 for Personal Standards, and α = .86 for Organisation, with total scale α = .88. In a large Australian university sample (N = 6,449), Stallman and Hurst (2011) reported α = .91 for a 29-item version, with subscale alphas ranging from .76 to .90. Rice & Dellwo, (2002) reported test-retest reliability over a 10-week period ranging from .63 to .88 across subscales in an Australian adolescent sample.

FMPS Factor Structure

The factor structure of the FMPS has been extensively examined. The original six-factor model (Frost et al., 1990) identified Concern over Mistakes, Personal Standards, Parental Expectations, Parental Criticism, Doubts about Actions, and Organisation as distinct dimensions. Subsequent research has consistently found that certain original factors are better combined. Stöber (1998) demonstrated that a four-factor solution, combining Concern over Mistakes with Doubts about Actions and Parental Expectations with Parental Criticism, provided a more parsimonious and replicable structure. This four-factor model has been widely adopted and is the scoring structure used on NovoPsych.

Normative Data

Normative data for the FMPS are derived from Stöber’s (1998) sample of German university students (N = 243, mean age 26.3 years, SD = 5.7). This dataset remains the only published normative sample providing means and standard deviations for the four-factor 35-item scoring structure used on NovoPsych. While subsequent studies have reported descriptive statistics for various populations, these have typically used modified versions or different factor structures, limiting their direct comparability to the original 35-item scoring. Note. CMD = Concern over Mistakes and Doubts about Actions; PEC = Parental Expectations and Criticism; PS = Personal Standards; O = Organisation. Stöber (1998) values represent the four-factor structure used on NovoPsych. Stallman and Hurst (2011) used a modified 29-item version with a five-factor structure. Dash indicates value not directly comparable or not reported for the same scoring model.

FMPS Descriptors

The following percentile-anchored descriptors are used to aid interpretation of FMPS scores. These bands were developed by NovoPsych to assist interpretation. They reflect standard percentile banding conventions commonly used in psychological assessment (e.g., Groth-Marnat & Wright, 2016) rather than FMPS-specific validation research, and should be treated as a practical guide rather than empirically validated clinical thresholds.

These descriptors are based on where a score falls in the normative distribution, rather than on the response scale labels (Strongly Disagree to Strongly Agree). This is because those labels reflect how strongly someone agrees with a statement, not how clinically elevated their perfectionism is.

• Very Low (≤5th percentile, score 29-44): Scores at this level are well below the normative average, suggesting minimal endorsement of perfectionistic beliefs or behaviours in this domain.
• Low (6th–24th percentile, score 45-60): Scores at this level are below the normative average, suggesting relatively low levels of perfectionism in this domain.
• Average (25th–75th percentile, score 61-83): Scores at this level fall within the typical range for the normative sample, suggesting perfectionism levels consistent with the general population.
• High (76th–94th percentile, score 84-99): Scores at this level are above the normative average, suggesting elevated perfectionism in this domain that may warrant clinical consideration.
• Very High (≥95th percentile, score 100-145): Scores at this level are well above the normative average, suggesting highly elevated perfectionism in this domain that may be clinically significant.

FMPS Sensitivity to Change

No known published study has directly examined the sensitivity to change of the FMPS using the Reliable Change Index (RCI) or minimally important difference (MID) methodology. However, the FMPS has been used as a primary or secondary outcome measure in clinical trials of cognitive-behavioural therapy for perfectionism (Egan et al., 2014; Rozental et al., 2017; Shafran et al., 2017), suggesting it is responsive to treatment-related changes. In the absence of a formally established change threshold, clinicians monitoring treatment progress can use the established recommendation that changes of at least 0.5 standard deviations in scores may suggest potentially meaningful change (Norman et al., 2003; Turner et al., 2010). Using the Stöber (1998) normative data, this corresponds to an approximate MID of 5.75 points for Concern over Mistakes and Doubts about Actions, 5.12 points for Parental Expectations and Criticism, 2.52 points for Personal Standards, and 2.92 points for Organisation.

Assess your patients with the Frost Multidimensional Perfectionism Scale (FMPS) using a NovoPsych account

• For Psychologists & Mental Health Clinicians
• Send Assessments to Patient’s Phone
• Receive Comprehensive Reports
• Access Over 100 Validated Assessments​
• Instant Psychometric Scoring
• Track Symptoms​
• Inform Treatment​