Does the BFI-2 Provide Official Norms for Percentiles?

Does the BFI-2 come with an official percentile norm?

The Berkeley Personality Lab makes a clear distinction on its BFI-2 page: scoring instructions are available, but the lab says there is no official BFI-2 manual with published norms. It points instead to Srivastava and colleagues’ 2003 paper, which reports age- and gender-related means from an internet sample using the earlier BFI; scores were converted to percentage of maximum possible and graphed from ages 20 to 60. Those averages are not an official BFI-2 percentile table.

Scoring and norming answer different questions. Scoring applies an item key to responses to produce a domain or facet score. A percentile ranks that score in a reference group’s score distribution. A person can therefore receive a calculated BFI-2 score without having an official population norm for interpreting an individual percentile. Berkeley’s statement describes the official materials it lists; it does not show that independent researchers have never collected BFI-2 comparison data.

The label ‘BFI-2 percentile’ alone does not reveal whether the comparison is an official norm, a research sample, a local dataset, or something else. If a report does not disclose its basis, that basis is unclear; missing disclosure does not establish who created the comparison or how. Ask for the data source, sample, and ranking method before treating the number as a meaningful relative position.

Sources: Big Five Inventory - Berkeley Personality Lab; Development of Personality in Early and Middle Adulthood: Set Like Plaster or Persistent Change? - University of Texas at Austin author-hosted paper; Glossary - National Council on Measurement in Education

What does the BFI-2 research establish instead?

Soto and John’s 2017 BFI-2 paper reports three development and validation studies. The first defined a 15-facet hierarchy and built a 110-item pool; the second refined it to the 60-item inventory using conceptual and empirical criteria. In Study 3, the authors examined measurement properties and relations to criteria in two independent samples: 1,000 adult visitors to a personality website and a student sample of 470, of whom 438 completed additional measures. They used self-ratings, peer ratings, established personality measures, and self- and peer-reported criteria. Results supported a hierarchical domain-and-facet structure, reliability and self–peer agreement, and associations with related measures; adding BFI-2 facets improved prediction for 14 of 20 criteria beyond domain scores.

That is evidence about measurement and construct-related relationships, not evidence that the study produced an official, representative norm table for individual percentiles. The authors also note limits: most participants were young or middle-aged adults with at least high-school education, limiting generalization to older adults and people with less education; the online and student validation samples do not amount to a representative population sample. The paper’s findings support bounded research uses, but cannot establish that a particular provider’s reference group fits a reader.

A researcher can still compare scores within a clearly described BFI-2 sample to answer a bounded research question. That comparison should be labeled and interpreted within its sample; validation evidence does not turn it into an official, broadly applicable norm.

Sources: The Next Big Five Inventory (BFI-2): Developing and Assessing a Hierarchical Model With 15 Facets

What does the older BFI comparison actually tell us?

The APA publisher’s full text for Srivastava and colleagues’ 2003 study describes a cross-sectional analysis of 132,515 internet respondents from the United States and Canada, aged 21–60, who completed a Big Five measure. Comparing different age groups at one point in time, the authors reported that mean Conscientiousness and Agreeableness generally increased through early and middle adulthood; mean Neuroticism declined among women and remained stable among men. These are group mean patterns from the earlier BFI research, not an individual BFI-2 percentile or a distribution for ranking one person. The internet respondents were not a representative norm sample, and the cross-sectional design cannot show that individuals changed as they aged.

The National Council on Measurement in Education defines norm-referenced interpretation as comparing performance with a specified reference-population distribution. A mean is an average; by itself it does not show the full distribution needed to calculate a percentile. The APA study supports statements about its sampled internet respondents, but its design, earlier instrument, and sample do not automatically fit a BFI-2 reader.

When a report cites ‘published norms,’ follow the citation. Check whether it concerns the BFI-2 version used, supplies a score distribution rather than only means, describes how the sample was recruited, and explains why that population is relevant. A citation to real research does not by itself establish that its results support an individual rank in another setting.

Sources: Development of Personality in Early and Middle Adulthood: Set Like Plaster or Persistent Change? - University of Texas at Austin author-hosted paper; Glossary - National Council on Measurement in Education

How should I read a BFI-2 percentile in a report?

Treat a percentile as a conditional comparison, not as the amount of a trait you possess. NCME defines percentile rank by a score’s position in a specified distribution. It does not mean that someone has that percentage of a trait or predict how they will act in a particular situation. Without the distribution, the number lacks the comparison needed for interpretation.

Use a short audit. Confirm whether the report used the full BFI-2 or a short form and how it calculated the score. Identify the reference group, its source and date, and relevant features such as language and age. Check that the cited material provides individual score distributions and a percentile method, rather than only group means. Then ask whether that group fits the question you want the score to answer. When these details are missing, describe the percentile’s basis as unknown and request clarification; do not assume it was created by the provider.

For low-stakes reflection, a documented comparison can be one prompt for noticing behavior in concrete situations. It does not settle what a tendency means at work or elsewhere. Request the source and calculation method; if the provider cannot supply them, set the percentile aside and use the underlying score cautiously. The Work Pattern Report at /assessment offers a separate, non-validated self-reflection on work tendencies. It does not provide BFI-2 norms or determine job fit.

Sources: Glossary - National Council on Measurement in Education; Big Five Inventory - Berkeley Personality Lab

Questions readers ask