Frost perfectionism scale: 35 statements, six dimensions
Perfectionism gets treated as one trait. Frost and colleagues split it into six in 1990: concern over mistakes, personal standards, parental expectations, parental criticism, doubts about actions, and organization. Thirty-five statements, each rated 1 (strongly disagree) to 5 (strongly agree).
Nothing is reverse scored. The total runs 35 to 175 with organization included, and 29 to 145 without it. Which of those two numbers you should use depends on what you are asking, and researchers disagree about it.
| Dimension | Items | Item numbers | Alpha | What it covers |
|---|---|---|---|---|
| Concern over mistakes | 9 | 9, 10, 13, 14, 18, 21, 23, 25, 34 | 0.88 | Reading errors as failure and expecting others to judge you for them |
| Personal standards | 7 | 4, 6, 12, 16, 19, 24, 30 | 0.83 | Setting very high goals for yourself |
| Parental expectations | 5 | 1, 11, 15, 20, 26 | 0.84 | Believing your parents expected excellence |
| Parental criticism | 4 | 3, 5, 22, 35 | 0.84 | Believing they were hard to satisfy |
| Doubts about actions | 4 | 17, 28, 32, 33 | 0.77 | Repeating and rechecking because it still feels wrong |
| Organization | 6 | 2, 7, 8, 27, 29, 31 | 0.93 | Preference for order and neatness |
Why organization sits apart
Organization correlates weakly with the clinical parts of perfectionism. Someone can be highly organized without being harsh with themselves, and the reverse happens too. That is why some researchers leave those 6 items out of the total, and why a high organization score on its own should not be read as a problem. Stöber (1998) went further and argued for four factors instead of six, merging the parental items and folding doubts about actions into concern over mistakes.
Reliability
Subscale alphas from the original study range from 0.77 to 0.93. A Spanish validation with 582 female students reported a total alpha of 0.93, subscales between 0.74 and 0.91, and a retest intraclass correlation of 0.89 over an interval with a confidence interval from 0.80 to 0.94. An Australian community sample of 206 reported 0.93 for the total, 0.85 for personal standards, and 0.92 for the combined evaluation worries score. Over ten weeks in adolescents, subscale retest correlations ranged from 0.63 to 0.88, which is a reminder that these dimensions move.
What samples average
| Measure | Mean (SD) |
|---|---|
| FMPS total, Australian community sample | 95.91 (21.42) |
| Personal standards | 22.50 (5.89) |
| Concern over mistakes | 20.53 (7.83) |
| Doubts about actions | 9.84 (3.77) |
| Evaluation worries (concern over mistakes plus doubts) | 30.37 (10.65) |
That sample held 206 people, 74% women, aged 18 to 86 with a mean age of 30.56. It is a community sample rather than a clinical one, so these are ordinary numbers, not patient numbers.
Reading a score
| Total | Reading |
|---|---|
| 35-75 | Low perfectionism |
| 76-95 | Moderate |
| 96-115 | High |
| 116-145 | Very high |
| 146-175 | Extremely high |
These are descriptive bands for this page, not clinical thresholds. The FMPS does not have an agreed cutoff. What carries more information is the shape of your six dimension scores: high concern over mistakes with high doubts about actions points at self-critical perfectionism, while high personal standards with low concern over mistakes looks more like conscientious ambition.
Chinese version
A Chinese revision kept all 35 items and reported a total alpha of 0.84, with subscales of 0.79 for concern over mistakes, 0.71 for personal standards, 0.61 for parental criticism, 0.60 for doubts about actions, and 0.87 for organization, plus a retest correlation of 0.79. Two of those subscale numbers are low enough to be cautious with. Exploratory analysis produced eight factors explaining 55.8% of the variance, and forcing the data back into the original six-factor model gave a comparative fit index of 0.88, which is acceptable rather than good. Scores correlated with the symptom checklist factors and went the other way with self-esteem and life satisfaction.
What this page is not
Perfectionism is not a diagnosis, and high standards are not a disorder. The part that tends to cause trouble is the self-criticism attached to them, which is what the concern over mistakes and doubts about actions items are asking about. If that part is interfering with sleep, work or relationships, it responds well to therapy. This page is not a diagnosis and does not replace one.