What the BFAA Study Adds to Understanding Daily Behavior

What does the BFAA study add to understanding changing behavior?

When behavior looks different from one day to the next, a person may have changed a stable tendency, or may report a different experience during a particular window. The Big Five Aspect Adjectives (BFAA) study measures traits and states at five Big Five domains and ten aspects beneath them. A state is a report tied to a recent window; a trait is a broader tendency summarized across time. A momentary report describes what was present then without establishing what is typical. “Measuring Personality States Below the Big Five: Development and Validation of the Big Five Aspect Adjectives for Traits and States” reports four studies. The first supported a 60-item inventory in a US sample of 352. A second Australian sample of 144 was used to compare it with the Big Five Aspect Scales (BFAS) and examine four-week test-retest reliability. A third US study of 549 tested trait-state correspondence and links with affect. In the fourth, 265 UK participants completed up to 24 surveys each, yielding 3,999 observations. The authors report aspect-level trait-state correspondence in this repeated design. Together, these studies add evidence that researchers can study aspect-level state reports with a purpose-built measure, rather than assuming trait questions work unchanged for a recent window. This offers a finer lens on changing self-reported experience. It is a measurement advance, not an explanation of why someone acted differently on a particular day. The studies relied on adjective ratings, which do not replace observation of actions; repeated reports also do not make each episode a stable trait. BFAA supports more finely structured questions about momentary reports. The cause and personal meaning of any one change remain open.

Sources: Measuring Personality States Below the Big Five: Development and Validation of the Big Five Aspect Adjectives for Traits and States

Why measure aspects beneath the five broad domains?

A broad trait label can be useful and still leave an important question unanswered: which part of that tendency is showing up in a particular situation? The Big Five Aspect Scales (BFAS) offer an intermediate level for asking that question. An aspect is a model-based grouping of related narrower traits within a broad domain. It sits between a domain such as Extraversion and more specific facets or behaviors. In “Between Facets and Domains: 10 Aspects of the Big Five,” the researchers analyzed 75 facet scales from two Big Five inventories in a community sample of 481 adults. Their analyses indicated two distinct but correlated aspects within each of the five domains. They then constructed a 100-item BFAS and validated it in a second sample of 480. That middle level matters because a domain can gather together tendencies that are related without being interchangeable. The BFAS paper names Assertiveness and Enthusiasm as two aspects of Extraversion. Assertiveness concerns a more agentic side of the domain, including taking initiative or speaking up; Enthusiasm captures more of the energetic, positive-engagement side, including enjoying social exchange. Someone may readily direct a discussion yet find prolonged mingling tiring, or enjoy a lively conversation without wanting to lead it. The examples show why one behavior should not stand in for every part of a broad domain; they do not describe any particular person. For research on changing behavior, the distinction suggests a more precise question. If a person reports being unusually quiet in one recent window, a broad Extraversion measure may not tell us whether the difference concerns willingness to take the floor, enjoyment of company, or something else in the situation. An aspect measure gives researchers a way to ask whether reports associated with these more specific patterns move together or separately over time. The Big Five Aspect Adjectives for Traits and States (BFAA) takes this aspect framework and tests adjective reports at both broader and aspect levels, including state reports tied to a recent period. The BFAA study adds evidence for measuring state reports at the aspect level. It does not follow that the measure identifies the cause of a quiet moment or observes behavior independently of what participants report. The distinction keeps interpretation proportional. A domain remains useful when the question is broad, such as whether sociable engagement tends to be more common across many settings. An aspect may help when a recurring difference within that domain matters, such as initiative varying separately from enjoyment. More detail is useful only when a question depends on the distinction. The aspects remain correlated. In the original BFAS research, they were described as distinct but related dimensions, and the authors’ labels summarize patterns found in particular inventories and samples. Measures or questions may organize traits differently. This is why the aspect level should be treated as a lens for forming sharper observations, not as a hidden “true self” revealed by one day. If a person speaks less during one meeting, the event alone cannot tell us whether assertiveness shifted, enthusiasm shifted, the setting changed, or several things occurred together. Repeated observations may show a pattern, but an aspect label still summarizes evidence rather than explaining its cause. Ask: what did I do, in which part of the situation, and is the recurring difference about taking initiative, enjoying interaction, or neither? The answer remains open to evidence. The practical addition, then, is resolution. Domains give a compact map; aspects distinguish related tendencies a label can blur. BFAA applies that distinction to time-bounded self-reports and tests whether aspect-level state structure is supported. It offers a more specific vocabulary for investigating day-to-day reports. It does not make every variation meaningful, prove that all ten distinctions fit every person, or turn a report about a recent window into a stable identity. Use the finer level when a repeated difference calls for it; otherwise, the broad description may be enough.

Sources: Between Facets and Domains: 10 Aspects of the Big Five; Measuring Personality States Below the Big Five: Development and Validation of the Big Five Aspect Adjectives for Traits and States

What did the BFAA researchers actually test?

The BFAA paper tested more than whether people describe themselves differently from one moment to another. Across four studies, it developed an adjective measure for Big Five aspects in both broad trait and time-bounded state forms, then tested how those forms related to an established aspect measure, to affect, and to repeated reports. The sequence matters: evidence for the measure’s structure and trait use came before the authors tested state correspondence, and a repeated-sampling study extended that test across many reporting windows. The result is a research instrument with initial evidence for aspect-level state measurement, not an explanation of why any one person behaved differently on a given day. In Study 1, 352 participants in the United States provided evidence supporting a 60-item inventory. The authors examined its structure and trait properties, including whether it reproduced the complex structure found in the Big Five Aspect Scales (BFAS) and showed similar patterns of association with criterion measures. The BFAS is the established comparison measure that organizes the Big Five into ten aspects. This first study therefore asks whether the proposed adjective inventory can represent the intended trait structure and behave in broadly comparable ways. It does not, by itself, establish that a short-window state report has the same structure. That distinction is central to the paper’s question: a list that works for “in general” descriptions still needs evidence when respondents are asked about a particular recent period. The publisher’s account reports support for the inventory, while the conclusions remain bounded by the samples and measures studied. Study 2 used an Australian sample of 144 to compare BFAA with the BFAS and assess four-week test-retest reliability. In ordinary language, test-retest reliability asks whether scores from the same measure show consistency when the measure is repeated after an interval. This provides a separate check on the trait form; it should not be mistaken for proof that a person’s state remains unchanged over four weeks. Study 3 then shifted the question. In a US sample of 549, the researchers modeled correspondence between BFAS trait reports and BFAA state reports, and tested whether BFAA state reports were associated with positive and negative affect in theoretically expected directions. Participants rated adjectives for a recent window; the paper gives “the last three hours” as the Study 3 state timeframe example. Affect here means reported feeling states such as positive or negative mood. The reported associations support the measure’s expected relations with momentary experience, but they do not show that affect caused a behavior, or that a rating is an independent observation of what someone did. Study 4 repeated the correspondence test with experience sampling in a UK sample of 265 participants. Experience sampling means collecting reports at multiple points in everyday life rather than asking for one retrospective summary. The study used up to 24 surveys per participant and reported 3,999 observations. Those figures describe repeated reports nested within people: 3,999 is not the number of participants, and repeated entries do not turn this into a sample of thousands of independent individuals. The authors used multilevel multivariate modeling, a way to analyze relations across repeated observations while accounting for their grouping within participants. They report that this study replicated the trait-state correspondence found in Study 3, providing evidence that aspect-level correspondence also appears across sampled windows. The study still asked participants to rate adjectives about themselves; it did not independently code actions, manipulate situations, or identify the cause of a particular shift. Taken together, the four studies provide a stronger basis than simply changing a trait questionnaire’s instructions: they test item structure, comparison with BFAS, trait retest, theoretically expected affect relations, and repeated state correspondence. Their contribution is validation under these tested conditions. Readers should take the repeated design as evidence that researchers can examine aspect-level patterns over time, while keeping self-report, sample limits, and causal questions distinct.

Sources: Measuring Personality States Below the Big Five: Development and Validation of the Big Five Aspect Adjectives for Traits and States

What had daily-state research established before BFAA?

Daily variation was not BFAA’s discovery. Earlier experience-sampling research had already asked how people’s reported personality expression shifts across ordinary moments, and how those reports relate to broader trait descriptions. The narrower contribution of the Big Five Aspect Adjectives for Traits and States (BFAA) study is whether researchers can measure such time-bounded reports with structure at a finer level: ten aspects nested within the five broad domains. A new way to measure a familiar phenomenon can add resolution without being the first evidence that the phenomenon exists. Experience sampling collects repeated reports in natural settings, close to when experiences occur. Experience Sampling Methods: A Modern Idiographic Approach to Personality Research describes how repeated reports reveal patterns within a person over time, beyond comparisons of one person’s average with another’s. Reports may be gathered at set or variable times, or after an event. Because observations are nested within people, many reports from one person do not become many separate people. Researchers can examine both a typical level and variation around it. A single personality number can hide the difference between an average and a range. A mean state report describes a person’s typical reported expression over the sampled period. Variability describes how much the reports spread around that mean. Similar averages can conceal different ranges of fluctuation, and one person may report different states at different moments. They answer different questions about the reports. The methods review treats tendencies as patterns across states, rather than assuming a trait appears at one fixed level on every occasion. The review summarizes an earlier experience-sampling study by Fleeson in which within-person variability in Big Five states was high. Its account says variability within people exceeded variability between people in the traits they were expressing at a moment. The review uses this finding to explain why a person’s distribution of states, including its typical level and spread, can be informative. This historical result is not a direct BFAA validation, nor evidence that everyone varies equally or that reports capture outward action. It shows that momentary variation in broad trait expression was already under study. A separate domain-level comparison comes from The Implications of Big Five Standing for the Distribution of Trait Manifestation in Behavior: Fifteen Experience-Sampling Studies and a Meta-Analysis. The accessible PubMed abstract describes a meta-analysis of 15 experience-sampling studies conducted over eight years, with more than 20,000 reports of trait manifestation in behavior. Participants first completed typical trait questionnaires, then described current behavior repeatedly over several days. The analysis found that trait standing predicted individual differences in average reported trait manifestation, with correlations from .42 to .56, and approaching .60 in more stringently restricted studies. This concerns broad trait standing and average reported manifestation across people; it does not explain a particular person’s action at a particular moment. Together, these earlier sources set a boundary around what is new. Researchers had evidence that momentary Big Five expression varies within individuals and that broad trait standing relates to patterns in repeatedly reported behavior. Repeated observations supported examining averages, ranges, and within-person patterns. BFAA asks a more fine-grained measurement question: whether state reports can also be organized and validated at the aspect level below each broad domain. The meta-analysis is a comparison point, not evidence for BFAA’s aspect structure; the methods review does not test the newer measure. The practical distinction is useful. A change on one day is evidence about a report or behavior in a particular window; it is not, by itself, evidence that a broad tendency has changed. A repeated pattern may be more informative, especially if the same kind of difference appears across occasions. BFAA’s added resolution may help researchers ask whether that recurring difference sits within one broad domain’s aspects. It does not supply an explanation for an individual episode. The basic finding that expression varies came first; the study’s contribution is to test a more detailed vocabulary for describing those state reports.

Sources: The Implications of Big Five Standing for the Distribution of Trait Manifestation in Behavior: Fifteen Experience-Sampling Studies and a Meta-Analysis; Experience Sampling Methods: A Modern Idiographic Approach to Personality Research

Why is explicit state validation a meaningful addition?

A time-bounded personality report needs evidence for the question it asks. Asking how someone tends to be and asking how they were over the last few hours are different tasks. The second answer concerns a particular window and may reflect tendencies, current feelings, the situation, memory, and questionnaire wording. A trait measure cannot be assumed to work unchanged as a state measure because the same adjectives appear in both. Explicit validation checks whether momentary reports relate to other measures and outcomes as the model predicts, rather than treating a changed instruction as sufficient evidence. The open methods review “Assessing Personality States: What to Consider when Constructing Personality State Measures” explains the distinction. Reviewing 24 studies from major personality journals, it found wide variation in how state measures were built and evaluated. A common approach averaged repeated state scores within each person, then reported internal consistency or correlations with one-time trait reports. The authors point out that these analyses describe the average-state score, not a single moment. A mean across windows may be estimated consistently even when one window is noisy or shaped by immediate circumstances. A relation between average state and broad trait alone does not show that each momentary answer is a sound reading of that state. This qualifies earlier state research; it does not dismiss it. Repeated reports can describe typical levels and variation, and test whether a person’s average corresponds with a broader tendency. But evidence at the average level should not transfer silently to one report. If a questionnaire asks, “How were you in the last three hours?”, its interpretation concerns a bounded, self-reported window. To support that use, researchers need evidence involving state-form responses and clarity about what they represent. “Reliable” means scores show appropriate consistency for their intended interpretation; “valid” means evidence supports the meaning and use assigned to them. Neither means a report captures an experience without error. “Measuring Personality States Below the Big Five: Development and Validation of the Big Five Aspect Adjectives for Traits and States” addresses this with four studies. They included US samples of 352 and 549 people, an Australian sample of 144, and a UK sample of 265. The first two examined the adjective inventory in trait form, including comparison with the Big Five Aspect Scales and a four-week retest. The third tested trait-state correspondence and links between state reports and momentary positive and negative affect. The fourth repeated state assessments: 265 UK participants were invited to complete up to 24 surveys, yielding 3,999 observations. The authors report correspondence between trait aspects and matching state aspects. The 3,999 figure counts observations, not people; repeated reports from one participant are related, not independent. This design offers more evidence than changing questionnaire instructions. It supports the BFAA’s intended use for self-reported aspects in trait and state forms, including repeated windows. Affect associations add a theoretically relevant comparison. They do not show that adjective ratings equal independently observed actions, that affect caused an answer, or that the measure works equally well for every interval or population. Findings support the tested instrument and procedures, not every adaptation or personal interpretation. For a reader, the distinction is useful. A state questionnaire can describe a recent period; read its result as a report about that window, not proof of stable identity or a complete account of events. If an interpretation matters, ask what the item measured, which period it covered, and whether evidence concerns single reports, averages, or both. The BFAA study adds evidence at the aspect level under specified research conditions. It does not make a brief answer the whole truth about a moment.

Sources: Assessing Personality States: What to Consider when Constructing Personality State Measures; Measuring Personality States Below the Big Five: Development and Validation of the Big Five Aspect Adjectives for Traits and States

How can context explain a shift without turning it into a label?

A change in a state report can coincide with a change in setting, task, current feeling, or chance to act. The report alone cannot tell which of these mattered, or whether the shift had one cause. That distinction separates two questions: what a person reported during a bounded period, and what conditions accompanied that report. The Big Five Aspect Adjectives measure may give future researchers finer language for describing the first question. It does not turn the second into a solved explanation. An aspect label is not a decoder for a single scene. Situation-contingency research offers evidence that context can matter systematically. In “Situation-Based Contingencies Underlying Trait-Content Manifestation in Behavior,” participants reported several times a day for two or five weeks on current Big Five states and characteristics of the concurrent situation. Multilevel analyses found that some situation characteristics covaried with state reports, that these patterns differed by trait, and that individuals varied reliably in their contingencies. The paper describes two studies, not a universal rule for how everyone responds. Its result supports asking whether a state shifts with a feature such as structure or familiarity; it does not establish why one person spoke less in a particular meeting. The assessments also asked participants to characterize their current state, rather than independently coding each physical action. The Stanford Media and Personality Lab record for “Personality-Place Transactions: Mapping the Relationships between Big Five Personality Traits, States, and Daily Places” describes three two-week experience-sampling studies: two exploratory and one confirmatory, with 2,350 participants and more than 63,000 momentary assessments. The record reports associations between stable Big Five traits and how often people visited different places. It also reports that participants described themselves as more extraverted in the moment in some social places than at home. This is a useful comparison because it links daily settings with both recurring place patterns and momentary reports. It cannot explain a particular action: people select places, settings combine features, and self-reports are not observations. These findings concern broad Big Five traits and states, not the BFAA's ten aspects. Consider an ordinary illustration, not a research case: someone contributes little in a rushed meeting with unfamiliar colleagues, then speaks readily in a prepared discussion with familiar ones. “Quiet” is the observable contrast only if it is defined concretely, such as fewer spoken turns. “Low assertiveness” or “low extraversion” is an interpretation layered onto it. The meetings differ in several ways at once: pace, preparation, familiarity, status, topic, and perhaps the person's energy or interest. The comparison cannot isolate any one cause. Nor does speaking more in the second meeting prove that the person has a different stable personality there. The illustration shows why recording the action and conditions separately is more informative than turning one moment into a label. For reflection, keep three notes distinct: what happened, what the setting demanded or allowed, and what interpretation seems plausible. For example, record “asked no questions during the ten-minute update,” then note whether there was time to respond, whether questions were invited, and how familiar the topic felt. A later contrast may make one explanation more plausible, but a few observations remain personal evidence, not a validated assessment. If a difference recurs across occasions, an aspect distinction might help pose a more specific question than a broad domain does. That is an inference about how finer measurement could guide inquiry; the BFAA findings do not show that an aspect explains the situational mechanism. Keep alternatives open until repeated observations distinguish them. The practical gain is not a more confident label, but a better description of what changed and what else changed with it.

Sources: Situation-Based Contingencies Underlying Trait-Content Manifestation in Behavior; Personality-Place Transactions: Mapping the Relationships between Big Five Personality Traits, States, and Daily Places

What can one changing day tell you about a broader tendency?

One changing day can tell you what you reported about a particular window; it cannot, by itself, tell you what kind of person you are or why you acted that way. Repeated observations can show what recurs, how much it varies, and which situations coincide with it. The BFAA study makes a finer account of self-reported states available for research. For personal reflection, the useful starting point is a concrete action observed across occasions, not a label applied to one moment. Experience-sampling research treats repeated reports as snapshots rather than competing verdicts. In “Experience Sampling Methods: A Modern Idiographic Approach to Personality Research,” the authors explain how observations gathered in daily settings can be examined within the same person over time. A person's average report describes a typical level; variation shows how widely reports moved. Earlier work reviewed there found substantial variation in momentary Big Five state reports. This makes fluctuation compatible with personality research; it does not mean every change is meaningful, or that a short personal record can establish a stable tendency. A practical reflection aid is to keep four parts of an episode separate. First, write the observable action in ordinary terms: “I waited until the meeting ended to raise my concern,” rather than “I was withdrawn.” Second, note the time and setting, what the task required, how familiar the situation was, and whether there was a real opportunity to act differently. Third, revisit the question across occasions, including instances that do not fit the first impression. Finally, consider whether an aspect distinction clarifies a repeated difference. This is an editorial aid, not a validated scoring method or BFAA protocol. Suppose someone notices they spoke little in one meeting. The observation is limited: they spoke little in that meeting. Note whether the discussion was rushed, the topic familiar, or a speaking turn available. Elsewhere, they might speak early, ask questions, or lead. Those contrasts do not prove a cause. They identify what else to examine before turning one episode into a trait description. The question is not “Which scene is the real me?” but “What repeats, and under what conditions?” “Situation-Based Contingencies in Personality” reports research examining how situation features relate to Big Five state expression, including reliable differences between people in those situation–behavior patterns. This supports taking context seriously when interpreting variation. It does not establish why a particular reader behaved a certain way on a particular day, or show that an aspect label explains the situation. A deadline, a familiar colleague, low energy, or several factors together could matter; one episode cannot sort those possibilities out. Keep the observed action, circumstances, and trait interpretation distinct. Repeated notes can sharpen a question without producing certainty. A behavior recurring across settings may suggest a broad tendency; appearing mainly under narrower conditions makes context central. A mixed pattern is information too: people need not express every tendency the same way on every occasion. Research designs use repeated sampling suited to their questions; a casual log lacks a study's controls and analysis. Its purpose is to help someone notice patterns and ask better questions, not to make a diagnosis or prediction. BFAA's aspect-level language may help researchers ask whether state reports differ more specifically than broad domains alone capture. It does not turn self-description into independently observed behavior. A reader considering a recurring pattern can ask: What did I do, and over what period? In which settings did it happen, and where did it not? Does an aspect distinction sharpen the description, or am I reaching for a label too soon? A day is meaningful but remains one bounded observation.

Sources: Experience Sampling Methods: A Modern Idiographic Approach to Personality Research; Situation-Based Contingencies Underlying Trait-Content Manifestation in Behavior

How should you use the finding next?

Use the Big Five Aspect Adjectives for Traits and States (BFAA) as evidence that researchers can study self-reported personality states at a finer level than the five broad domains. Its four-study paper supports this measure under the conditions examined; it does not give you a personal score, verify your actions independently, or explain why you behaved differently on a particular day. The open methods review in PubMed Central distinguishes a person’s average across reports from the reliability of a single momentary report. One episode describes a window, not your usual pattern.

For a recurring behavior, record the action, timing, and situation across occasions. Ask whether it appears in different settings, whether exceptions cluster in particular conditions, and whether an aspect distinction clarifies what a broad label misses. This is a reflection aid, not a validated scoring method or BFAA protocol. Interpretation may change if the behavior occurs only in one narrow situation.

For a broader view of how tendencies combine, the private Context Profile offers an optional reflection exercise across several continuums. It is non-validated; it assigns no diagnosis or fixed type and does not explain career outcomes or why an event occurred. You can also explore the personality profile library at /topics.

Sources: Measuring Personality States Below the Big Five: Development and Validation of the Big Five Aspect Adjectives for Traits and States; Experience Sampling Methods: A Modern Idiographic Approach to Personality Research

Questions readers ask