ChapterPDF Available

Big-Five Model

Big-Five Model
John A. Johnson
Pennsylvania State University
This is the manuscript version of the following publication:
Johnson, J. A. (2017). Big-Five model. In V. Zeigler-Hill, T.K. Shackelford (Eds.),
Encyclopedia of Personality and Individual Differences (1-16). New York: Springer.
DOI: 10.1007/978-3-319-28099-8_1212-1
Changes between this manuscript version and the final published version could have
Big-Five Model
Synonyms: B5M, Five-Factor Model, FFM, OCEAN
Definition/Abstract: The Big-Five Model (B5M) is a representation of the universe of
personality traits in terms of five broad personality dimensions: Extraversion,
Agreeableness, Conscientiousness, Emotional Stability, and Intellect or Imagination.
These five dimensions emerged reliably over decades of research factor-analyzing the
way people describe each other with ordinary-language traits such as aggressive,
introverted, and sociable. The B5M is nearly identical to the Five-Factor Model (FFM),
which emerged from research based on personality questionnaires. Although both the
B5M and FFM have common roots in the research of Raymond Cattell and although the
terms B5M and FFM are often used interchangeably today, the two models have
distinctive histories, are based on different methods and theoretical assumptions, and hold
slightly different conceptions of the factors, especially the fifth factor, labeled Openness
to Experience in the FFM. The FFM also uses the opposite of Emotional Stability,
Neuroticism, to label the fourth factor, and some individuals reorder the factors to spell
the acronym OCEAN. This encyclopedia entry reviews the history of the B5M and FFM
and discusses issues with the models that are being addressed by current research. This
research includes studies that address (1) the best way to conceptualize the factors; (2)
optimal ways of operationalizing the factors; (3) optimal division of each of the factors
into subfactors of personality; and (4) whether the factors might be combined into
History of the Big-Five Model
In their overview of the five major personality factors, McCrae and John (1992)
observed that the Big-Five Model and Five-Factor Model emerged from two distinct
historical pathways. The first, and older, historical origin of the B5M can be found in
analyses of personality-descriptive words from ordinary language. Because these words
were identified from a dictionary (lexicon), this line of research is often called the lexical
research program or lexical tradition. The second historical path to the five major
personality factors is found in questionnaires designed by professional psychologists to
represent theoretical constructs or for practical applications such as psychiatric diagnosis
or personnel selection. This second historical line of research is therefore called the
questionnaire research program or tradition.
The Lexical Research Program
John, Angleitner, and Ostendorf (1988) have provided a detailed account of the
history of the lexical research program. The lexical research program has been based on
what has been called "the fundamental lexical hypothesis—namely that the most
important individual differences in human transactions will come to be encoded as single
terms in some or all of the world's languages" (Goldberg, 1990, p. 1216). Various
researchers from England, Germany, and the United States were guided by the
fundamental lexical hypothesis, the most influential being Gordon W. Allport.
Gordon W. Allport. Harvard psychologist Gordon W. Allport and Dartmouth
psychologist Henry S. Odbert undertook a project to exhaustively catalog every word in
the English language that is "descriptive of personality or personal behavior (save those
that are obsolete) included in Webster's New International Dictionary" (Allport & Odbert,
1936, p. 24). After identifying 17,953 personality-descriptive words, Allport and Odbert
sorted these traits into several categories. Most of the words fell into categories that they
considered to be ill-suited for a scientific description of personality. The inappropriate
words included 4541 terms that described temporary states or activities (e.g., satisfied,
scared); 5226 terms that described social evaluations (e.g., fascinating, insignificant); and
3682 miscellaneous terms that described physical characteristics commonly associated
with psychological traits (e.g., lean, red-headed), capacities and talents (e.g., gifted), or
were metaphorical or obscure (e.g., hard, piggish).
The only category they considered to be appropriate for personality description
contained 4504 terms that referred to stable traits that objectively describe people (e.g.,
aggressive, introverted, sociable.). This subset of 4504 terms, which trait researchers
have called prime trait terms, became the starting point of future efforts to derive a set of
basic personality traits.
Raymond B. Cattell. The next step toward the Big-Five Model was undertaken by
Raymond B. Cattell. Cattell was responsible for advancing both the lexical and
questionnaire research tracks that initially went in different directions until they were
reunited in the 1970s. Cattell recognized that a list of 4,504 personality trait terms was
still impractically long for research studies and undertook a procedure for reducing this
list. Cattell's descriptions of his procedure were sometimes imprecise and inconsistent
across publications. The summary of that procedure below reconstructs as accurately as
possible what Cattell did.
Cattell first added to Allport and Odbert's (1936) list of 4504 prime trait terms
100 terms from their list of temporary states or moods. Next, he and a student of literature
sorted the terms into roughly 160 clusters of terms they judged to be synonyms or nearly
identical in meaning (e.g., effusive, gabby, loquacious, talkative). Most of these clusters
were bipolar, which means he paired each trait word with an opposite (e.g., talkative vs.
silent). He wanted each cluster to have roughly the same number of terms, so he selected
about 13 terms from each cluster to represent the cluster and discarded the remaining
terms. Not satisfied with limiting his study to words from ordinary language, Cattell
supplemented his clusters by adding new clusters based on technical terms from
psychological research, especially terms dealing with neurosis and psychosis. He also
added a cluster representing general intelligence, 9 special abilities, and 11 clusters
describing interests. In some reports the total number of clusters he reports is as high as
181 or 182, but most of his publications concern research on 171 clusters (John,
Angleitner, & Ostendorf, 1988).
Next, Cattell had knowledgeable acquaintances of 100 adults rate how
characteristic or uncharacteristic each of the 171 clusters was in describing the target
person's personality, and then Cattell computed Pearson correlation coefficients among
all 171 variables. Pearson correlation coefficients represent how strongly any two ratings
are related. For example, when acquaintances rated the "talkative" cluster to be
characteristic of a target person, they also considered the "enthusiastic" cluster (which
included the terms enthusiastic, exuberant, high-spirited, and lively) to be characteristic
of the same person, so ratings for these two trait clusters were highly correlated.
Mathematically, Pearson correlation coefficients can range from -1.00 to 1.00.
This phase of his research was prior to the advent of computers, so the Pearson
correlations had to be computed by hand using mechanical adding machines. The amount
of human error in these calculations is unknown but some clerical errors have been
documented (Goldberg, 1993). The result of these computations was a 14-square-foot
table of correlations that was so large that it had to be spread out on the floor of a
gymnasium for examination. Cattell’s research team literally crawled around the floor,
inspecting the correlations to find the largest values. When they found correlations
greater than .83, Cattell considered the clusters identical (measuring the same thing).
Correlations between .45 and .83 were considered to indicate clusters that were similar
enough to belong together. Any three clusters that belonged together he called a triad;
four clusters, a tetrad, and five clusters, a pentad. By grouping the 171 clusters into triads,
tetrads, and pentads, Cattell reduced the number of clusters from 171 to 67 (Cattell, 1943)
or 69 (Cattell, 1946). Cattell later reported that he retained only 58 clusters that were
replicated in further analyses. Over a period of time, by simply looking at and thinking
about the content of the clusters, he dropped many of them, added others, and eventually
reduced the number of clusters to 35, 36, or 42 (John, Angleitner, & Ostendorf, 1988).
Finally, using the University of Illinois' first computer, Illiac I, Cattell submitted
ratings on the 35 broad clusters to a statistical procedure called factor analysis. Numerous
variants of factor analysis exist, but they are all designed to reduce data by identifying
sets of measurements that are related to each other, yet relatively unrelated to other sets
of measurements. Each set of related measurements is called a factor. Cattell employed a
variant of factor analysis that included what is called oblique rotation, which allows the
factors to overlap somewhat. This is rarely done today, as researchers prefer orthogonal
rotation, which produces factors that are completely distinct and independent of one
another. Factor analyses with oblique rotation tend to produce more factors than analyses
with orthogonal rotation. Cattell’s factor analysis produced 12 lexical factors.
Not satisfied with this number, Cattell added four personality constructs that he
felt were necessary for covering all essential personality traits. He created a self-report
questionnaire, still in use today, called the 16PF, to measure the 16 personality factors.
Cattell therefore contributed to both the lexical research program by reducing Allport's
list of prime trait terms to 35 clusters or 12 factors and to the questionnaire research
program by constructing a questionnaire based on the lexical factors. The precise way in
which the 16PF led to the Five-Factor Model in questionnaire research will be explained
in the later section on this history of the questionnaire research program. The next
question for the lexical research program is how Cattell's 12 lexical factors eventually
became five factors.
Ernest Tupes and Raymond E. Christal. Ernest C. Tupes and Raymond E. Christal
were two relatively unknown research psychologists working at Lackland Air Force Base
in Texas during the 1950s and 60s. The project that eventually gave them recognition as
fathers of the Big-Five Model was an attempt to use peer ratings of personality to select
pilots. Their peer rating study was motivated by a concern that self-reports could be too
easily faked. Ironically, their peer rating system was never used for pilot selection out of
concern that peers might provide biased ratings in order to improve their own chances of
being selected.
Tupes and Christal took interest in an article by Donald Fiske (1949), who
conducted a factor analysis of simplified descriptions of 22 of Cattell’s clusters and
reported five factors instead of 12. Fiske forwarded several possible reasons for finding
five rather than 12 factors, one of which was his decision to use a rotation that resulted in
relatively independent rather than overlapping factors. Tupes and Christal (1961/1992)
followed up on this suggestion by reanalyzing the staff ratings and teammate ratings from
Fiske's sample and the ratings from two of Cattell's samples, using orthogonal rather than
oblique rotation. They similarly analyzed ratings of Cattell's 35 personality clusters with
data from four new samples: three large samples of Air Force Officer Candidate School
graduates and a large sample of students from the Air Force Command and Staff School
Class. In all eight data sets, they found clear evidence of the same five factors, which
they labeled Surgency, Agreeableness, Dependability, Emotional Stability, and Culture.
Scholars today agree that the discovery of the five major personality factors by
Tupes and Christal might have languished in obscurity because their results were reported
in an Air Force technical report, except that Warren Norman, a professor at the
University of Michigan, read their report and attempted a replication with students at his
Warren Norman. Rather than use all 35 of Cattell's personality clusters, Norman
(1963) chose four clusters that he thought best represented each of the five factors
identified by Tupes and Christal. He then had members of four groups of students (ROTC
seniors, two fraternity samples, and dormitory residence-hall men) nominate one-third of
the other members of his group who best represented one pole of a personality cluster
(e.g., talkative) and the one-third who best represented the other pole (e.g., silent). The
summed ratings were subjected to factor analysis with orthogonal rotation. Norman
(1963) reports a clear replication of Tupes and Christal (1961/1992), although he chose
slightly different names for some of the five factors.
The factor names and summary labels for the personality cluster scales used by
Norman are as follows: I. Extroversion or Surgency (Talkative—Silent; Frank, Open—
Secretive; Adventurous—Cautious; Sociable—Reclusive); II. Agreeableness
(Goodnatured—Irritable; Not Jealous—Jealous; Mild, Gentle—Headstrong; Cooperative
—Negativistic); III. Conscientiousness (Fussy, Tidy—Careless; Responsible—
Undependable; Scrupulous—Unscrupulous; Persevering—Quitting, Fickle); IV.
Emotional Stability (Poised—Nervous, Tense; Calm—Anxious; Composed—Excitable;
Not Hypochondrical—Hypochondriacal); and V. Culture (Artistically Sensitive—
Artistically Insensitive; Intellectual—Unreflective, Narrow; Polished, Refined—Crude,
Boorish); Imaginative—Simple, Direct). This ordering of the five factors, complete with
Roman numerals, became a standard in subsequent lexical research (e.g., Hofstee, de
Raad, & Goldberg, 1992).
Norman ended his article by quoting and commenting on the following
concluding remarks in Tupes and Christal's technical report, "It is unlikely that the five
factors identified are the only fundamental personality factors. There are quite likely
other fundamental concepts involved among the Allport-Odbert adjectives on which the
variables used in the present study were based [p. 12]" (Norman, 1963, p. 582). Norman
(1963) suggested that the next step in the lexical research program should be "to return to
the total pool of trait names in the natural language—there to search for additional
personality indicators not easily subsumed under one or another of these five recurrent
factors" (p. 582).
Returning to the total pool of trait names is exactly what Norman (1967) did in
subsequent research. He repeated what Allport and Odbert (1936) had done with the 1925
unabridged edition of Webster's New International Dictionary, using the more recent
third edition of Webster’s New International Dictionary. Although he reported finding
about 9046 additional trait words, he said that most of the “new” terms were previously
identified words with added prefixes or suffixes. Only 171 truly new words were added
to the Allport and Odbert list. From this complete list, Norman selected what he
considered to be 2800 prime trait terms. He reduced this set of terms to 1,431 by
eliminating terms that University of Michigan students had trouble understanding. He
then organized these 1,431 terms into 75 categories based on his understanding of their
similarities in meaning. However, he did not subject his set of trait terms to factor
analysis. That task was undertaken by Lewis R. Goldberg.
Lewis R. Goldberg and John M. (Jack) Digman. Despite the apparent potential of
the five recurring personality factors for providing a basic taxonomy for personality
research, that potential was not realized until decades later. Critiques of personality
assessment during the late 1960s and throughout the 1970s virtually stopped the
publication of personality measurement research in what was sometimes called "the
decade of doubt" (Digman, 1996). Robert Hogan, today one of the most important
personality psychologists in the history of the Big-Five Model, reported (Hogan & Foster,
2016), "In the late 1960s and early 1970s, there was a wall of resistance against
personality assessment in academic psychology. The first 13 papers R. Hogan submitted
for publication came back without being reviewed. Editors said things like 'Everyone
knows personality assessment doesn’t work, so we’re not going to review this paper.'" (p.
Lewis R. Goldberg of the University of Oregon and the Oregon Research Institute
helped to make personality measurement research acceptable again by extending the
work of Warren Norman. Goldberg administered the Norman's 1,431 trait terms, along
with some additional terms, to a sample of university students. This methodology differs
considerably from prior lexical research in that research participants rated themselves on
hundreds of individual trait words rather than being rated by others on clusters of
personality traits. When he summed the 1,431 ratings according to Norman's 75
categories and submitted the scores to factor analysis, they produced the familiar five-
factor structure. The Zeitgeist was still negative for personality measurement research,
however. Instead of publishing his results in peer-reviewed journals, Goldberg presented
them in a symposium organized by Jack Digman for the annual meeting of the Western
Psychological Association (Goldberg, 1980) and described them in two invited book
chapters (Goldberg, 1981, 1982). In his 1981 chapter he coined the term "big five" (in
lower case and in quotation marks).The 1980s saw an improvement in the climate for
publishing personality measurement research. By the end of the decade, Goldberg (1990)
was able to publish his extensive factor analyses of Norman's trait terms.
There was another reason that Goldberg did not advocate for the Big-Five Model
until 1990. In an historical review of the Big-Five Model, Goldberg (1993) said that
initially he and Norman were both skeptics about the Big-Five structure. Norman's
skepticism is what led him to repeat the Allport and Odbert (1936) lexical search and to
develop his own 75 clusters of personality trait terms. While acknowledging throughout
the 1980s that the Five-Factor Model was compelling, Goldberg (1993) wrote that he
actually preferred an alternative three-factor model (general evaluation, assertiveness,
and impulse expression) proposed by colleague Dean Peabody (Peabody & Goldberg,
1989). Another close colleague, Jack Digman, was at first equally skeptical about the
Five-Factor Model, believing that as many as 10 factors were necessary to account for
teachers' ratings of their students' personalities.
However, when Digman examined Cattell's studies carefully, he found clerical
errors in two of Cattell's matrices (Goldberg, 1993). On reanalysis of six data sets,
including Cattell's data, Digman and Takemoto-Chock (1981) found a striking
convergence toward the Five-Factor Model. "Regardless of whether teachers rate
children, officer candidates rate one another, college students rate one another, or clinical
staff members rate graduate trainees, the results are pretty much the same" (Digman &
Takemoto-Chock, 1981, pp. 164-65).
Digman's conclusions were echoed by Goldberg (1981). Upon analyzing
Norman's 75 clusters, he wrote "it hardly matters what number of factors are extracted,
since the loadings оn the first five factors are always nearly the same. . . . Clearly, there
is something to this structure. It is not simply а matter of extracting а particular number
of factors or using а particular type of rotationa1 algorithm. These аге data speaking for
themselves" (Goldberg, 1981, p. 160).
Convinced that the Five-Factor Model was the most robust, replicable model of
personality trait words, Goldberg (1992) turned to the development of marker scales for
the five factors. His 1992 article presents both a set of 100 individual trait words and a set
of 35 bipolar rating scales anchored pair of trait words for measuring the five major
personality factors. He continued to collaborate with other researchers on lexical research
(e.g., Hofstee, de Raad, & Goldberg, 1992; Saucier & Goldberg, 1998). But by the end of
the 1990s he turned his focus from individual trait words to short phrases indicative of
personality (Goldberg, 1999), creating a public domain website for his International
Personality Item Pool (IPIP; Goldberg, et al, 2006). The IPIP ( contains
a number of different public domain inventories for measuring the Big Five with short
Digman also played a crucial role in promoting the Big-Five Model. In addition to
his own publications supporting the B5M, he authored an article for the Annual Review
of Psychology summarizing research in support of the model (Digman, 1990). Digman
and Goldberg both discussed at length alternative labels for the five factors. Digman
preferred the label Will to Achieve for the Conscientiousness factor Digman & Takemoto-
Chock (1981); however, that label never became mainstream. Goldberg, on the other
hand, successfully substituted the label Intellect for Norman's fifth factor, Culture.
Goldberg (1993) noted that the Culture label was an artifact of Cattell's removing
intellect-related personality words in favor of an actual intelligence test. Both Digman
and Goldberg also introduced a more subtle change in the labeling of the first factor,
using the spelling Extraversion rather than Norman's Extroversion. Extraversion was the
spelling preferred by the originator of the term, Carl Jung (Kaufman, 2015).
During the 1980s, Goldberg and Digman's lexical research also had an enormous
impact on the direction of personality questionnaire research. During this decade, major
personality inventories were authored or revised to align with the Big-Five Model, and
other major personality inventories began to be reinterpreted in terms of the five factors.
The details of how this happened are described in the next section.
The Personality Questionnaire Research Program
A defining feature of the lexical research program is that it has been descriptive
and atheoretical. The only "theory" underlying this research program is the lexical
hypothesis—the assumption that, over millennia, human beings noticed important
individual differences in each other and coined words for them. The job of academic
psychologists in the lexical tradition has been simply to measure how people use these
words to describe one another and to use statistical methods to model the basic
dimensions within personality description through ordinary language.
In contrast to the descriptive, atheoretical nature of lexical research, much
personality questionnaire research has involved measurement of purportedly scientific,
theoretical constructs (McCrae & John, 1992). Many psychologists, including Cattell,
believed that personality words from ordinary language were not sufficient for a truly
scientific description of personality (McCrae, 1990; McCrae, Costa, & Piedmont, 1993).
They designed self-report questionnaires, not to measure ordinary traits such as shyness,
but to measure constructs proposed by theorists such as Henry Murray’s need for
Affiliation or need for Harm avoidance. Cattell refused to use ordinary English words for
the constructs on his questionnaire, opting for terms he made up such as Parmia and
A sampling of well-known personality questionnaires indicates a diversity of
theoretical constructs. Gordon Allport's Harvard colleague, Henry Murray, devised a
theory of 20 "manifest needs" that he considered to be important determinants of human
behavior. This list served as the basis for two widely-used personality inventories, Alan
Edwards' Edwards Personal Preference Schedule and the Douglas Jackson's Personality
Research Form. The Allport-Vernon-Lindzey Study of Values was designed to assess
adherence to the six major types of values proposed by Eduard Spranger and thought to
be important in vocational choice. Jung's theory of psychological types was the basis for
the Myers-Briggs Type Indicator (MBTI). Hans Eysenck developed a behavioristic-
genetic-physiological theory of traits that distinguish normal people from criminals and
the mentally ill and created the Eysenck Personality Inventory and Eysenck Personality
Questionniare to assess these traits. J. P. Guilford and Wayne Zimmerman’s Guilford-
Zimmerman Temperament Survey (GZTS) was created by factor analyzing existing
inventories of theoretical constructs. Harrison Gough constructed the California
Psychological Inventory to assess folk concepts—the concepts used by ordinary people in
everyday life to describe human behavior.
The above are but a few examples of questionnaires developed between 1900 and
1970 (see Goldberg, 1971, for a more thorough review). Three noteworthy points about
these inventories are as follows: (1) many bear the name of the persons who developed
them; (2) each developed a loyal following devoted to the theory behind the inventory;
and (3) researchers devoted to a particular inventory tended to work in isolation of
research based on other inventories. As a consequence, it was difficult to see if these
inventories were really measuring different concepts or whether they measured
something in common. Order within the cacophony of questionnaire research did not
emerge until the construction of new personality inventories by Robert Hogan and by
Paul T. Costa, Jr. and Robert R. McCrae.
Robert Hogan. Robert Hogan was a student of Harrison G. Gough, author of the
California Personality Inventory (CPI). After conducting research with the CPI for
several years at the Johns Hopkins University, Hogan decided it would be more
convenient to create his own personality inventory for research and consulting. Well
aware of the Norman (1963) study, Hogan chose to construct his inventory around the
five factors described by Norman. As a first step toward constructing such an inventory,
he worked with graduate student John A. Johnson to see if the CPI could be scored to
yield scores on Norman's factors (Johnson, 1997). Hogan then placed Johnson in charge
of constructing the initial item pool and developing the preliminary scales for a new
inventory, the Hopkins Personality Inventory. After Johnson and Hogan left Hopkins,
Hogan revised and renamed the inventory, publishing it as the Hogan Personality
Inventory (HPI; Hogan, 1985; Hogan & Hogan, 1992). Hogan Assessment Systems has
used the HPI successfully in hundreds of studies of personnel selection and
organizational behavior.
Paul T. Costa, Jr. and Robert R. McCrae. As indicated in the earlier section on
Raymond Cattell, Cattell's 16PF questionnaire contained 16 different scales. Paul T.
Costa, Jr. and Robert R. McCrae had included the 16PF in a major longitudinal study, but
they believed that the scales of the 16PF were too redundant, with overlapping meanings.
They therefore conducted a cluster analysis on the items in the 16PF (Costa & McCrae,
1976). Two item clusters revealed by the analysis were Neuroticism and Extraversion,
familiar dimensions in the history of personality questionnaires, especially due to the
research of Eysenck (Goldberg, 1971). The third cluster, Openness to Experience, was
rarely discussed in the literature. Believing that they could write better items to measure
these three domains more validly, McCrae and Costa (1983) authored their own
inventory, which they called the NEO Inventory. It contained three “domain scales” to
measure the factors N, E, and O, but each of the domain scales contained six “facet
scales” that measured narrower aspects of each domain. For example, Neuroticism
contained facet scales for Anxiety, Angry Hostility, Depression, Self-Consciousness,
Impulsiveness, and Vulnerability.
In 1978 Costa and McCrae moved from Boston to accept new positions with the
National Institute of Aging in Baltimore, where they joined the Baltimore Longitudinal
Study of Aging (BLSA). Shortly after their arrival in Baltimore, Hogan invited them to
Johns Hopkins where he and his graduate students presented a seminar on the HPI
project. Hogan tried to convince Costa and McCrae that their NEO measured only three
of the important five factors identified by Norman. Costa and McCrae left the seminar in
disagreement with Hogan. Costa argued (much as Allport might have) that Agreeableness
and Conscientiousness were merely “social evaluations” rather than substantial
personality traits. Agreeableness reflects our personal like or dislike of people and
Conscientiousness reflects whether they share our moral values, he said. According to
Costa, these dimensions say more about observers' personal values than about the
objective personalities of the people they are judging.
However, soon after the meeting at Hopkins, Costa and McCrae became aware of
the work presented by Digman and Goldberg at the WPA symposium in Hawaii. They
invited Lewis Goldberg to present his research at the NIA Gerontology Center in
Baltimore in 1983. Digman's and Goldberg's data convinced Costa and McCrae that
Agreeableness and Conscientiousness were important domains, so they created two
scales for their inventory and renamed the NEO Inventory the NEO Personality Inventory
(NEO PI; Costa & McCrae, 1985). Perhaps because they were aware of the imminent
publication of the HPI (Hogan, 1985), Costa and McCrae's NEO PI was rushed to press
without the development of facet scales for Agreeableness and Conscientiousness. These
facet scales were developed later and were included in a revised version of the NEO PI
(NEO PI-R; Costa & McCrae, 1992).
Once the NEO PI was published, Costa and McCrae’s research program became
the most influential force promoting a Five-Factor view of personality questionnaires.
Costa and McCrae had at their disposal one of the largest longitudinal databases in the
world. Participants in the BLSA had already completed a large number of personality
measures and had also provided personality ratings from informants who knew them
well. Whenever Costa and McCrae wanted to test a new measure, they could administer it
to the BLSA participants. To demonstrate the power of the Five-Factor Model, Costa and
McCrae methodically compared scores on their NEO PI and NEO PI-R to scores on all of
the major personality inventories. From their results they argued that all major
inventories measured at least some of the five factors, and none produced important
information beyond the five factors. The B5M/FFM became a kind of standard against
which any other measure could be compared, and the NEO PI-R became the most widely
used personality questionnaire in FFM research.
Although the lexical research program and questionnaire research program
became integrated when Costa and McCrae incorporated Agreeableness and
Conscientiousness into the NEO PI-R, certain unresolved issues remained. Debates about
the best way to conceptualize the factors continued, especially about the fifth factor,
usually labeled Intellect in the B5M and Openness to Experience in the FFM. Further
research was necessary to see whether the rating scales employing trait adjectives or
questionnaire employing full-sentence items were preferable. Finally, researchers
continued to investigate ways of dividing the five broad factors into narrower
components and ways of combining the five factors into higher-level super factors. These
topics are addressed in the next major section of this entry.
Current Issues in Big-Five Model Research
Conceptualizing the Big Five Factors
Discerning the meaning of factors produced by any factor analysis always
involves subjective judgment on the part of the researcher. These judgments are informed
by the magnitudes of factor loadings for the factor. A factor loading can be thought of as
the correlation between an item and a factor. Factor meanings are derived by locating the
items with the highest loadings on the factor and intuiting a common theme running
through the items. For example, the following items in Norman's (1963) rating study
showed high loadings on Factor I in sample C: Talkative-Silent (.90), Frank, Open-
Secretive (.78), Adventurous-Cautious (.79), and Sociable-Reclusive (.86). (These items
also showed similarly high loadings in the other samples.) Norman decided that the
common theme running through these items is the same theme identified by Tupes and
Christal (1961/1992) in their analyses of these items, alternatively called Extroversion or
Norman (1963) also chose labels that were the same or similar to the labels used
by Tupes and Christal (1961/1992) for the remaining factors. Both labeled Factor II
Agreeableness. Tupes and Christal labeled Factor III Dependability and Norman labeled
this factor Conscientiousness. Both used Emotional Stability to label Factor IV and
Culture to label Factor V.
But judging the similarity of factors in different studies requires more than simply
comparing the labels used in each study. What matters is whether the same items show
similarly high loadings across studies. In this case, the four items with the highest
loadings on Factor III in the Norman (1963) study also showed high loadings in the
Tupes and Christal (1961/1992) study. This implies factor similarity across the studies
despite the different labels. Assuming that the labels Dependability and
Conscientiousness indicated different interpretations of the factor would have been an
example of what Jack Block (1995) called the jangle fallacy. The jangle fallacy refers to
mistakenly assuming that different labels for two psychological measures imply two
different psychological concepts.
Determining whether factors are the same across studies becomes more complex
when the same items are not used in each study because one cannot compare item factor-
loadings. Although Costa & McCrae's (1992) NEO PI-R had its origin in a cluster
analysis of Cattell's 16PF, which itself was derived from ratings that led to the lexical Big
Five, the fifth factor in the NEO PI-R questionnaire is labeled Openness to Experience
rather than Intellect. Although the content of items in the Openness to Experience scale
seems similar to trait terms that define the Intellect factor in lexical studies, McCrae
(1990) has claimed that the technical, scientific concept Openness to Experience is not
well-represented in ordinary language. He admitted that a number of trait words from one
lexical measure of Intellect (artistic, clever, imaginative, ingenious, insightful, inventive,
original, sharp-witted, and witty) do partially describe Openness to Experience, but he
pointed out that other trait words from this Intellect scale showed higher loadings one of
the other four factors. He concluded that "the fifth factor identified in lexical studies is
best interpreted as a variant of the more psychologically fundamental dimension of
Openness to Experience" (McCrae, 1990, p. 123) and asserted that the facets of the NEO
PI-R Openness to Experience scale cover the domain of the fifth factor more completely
than any lexical scale.
McCrae's assertions did not, however, close the debate on the best interpretation
of the fifth factor. In 1994, an entire special issue of the European Journal of Personality
was devoted to articles on interpretations of the fifth factor (de Raad & Van Heck, 1994).
McCrae maintained his position in his contribution to the special issue. As indicated
earlier in the section on the history of personality questionnaires, questionnaire authors
tend to remain committed to the conceptual basis of their questionnaires. In contrast to
McCrae, several other contributors to the special issue independently suggested that the
core of the fifth questionnaire factor is better represented by the Openness to Aesthetics
and Openness to Ideas facets of the NEO PI-R Openness scale than the other four
Openness facets. They concluded that the conceptual core of Factor V, revealed by both
questionnaire research and lexical studies, is best represented by the label Creative
Intellect or Imagination.
Interpretation of the other four factors has been much less contentious, although
there have been disagreements about where particular traits belong in the Big-Five
Model. Johnson and Ostendorf (1993) reviewed these disagreements before proposing a
way to resolve them. They noted that some researchers believe that positive emotions are
an aspect of Extraversion, while others locate positive emotions in Agreeableness. Some
researchers conceptualize Conscientiousness as a proactive tendency to methodically
strive toward the achievement of goals. This interpretation locates ambition as part of
Factor III and aligns with Digman and Takemoto-Chock's (1981) view of this factor as
Will to Achieve. Other researchers regard Conscientiousness as a restraining force, a
denial of anti-social impulses leading to conformity to social expectations and norms. For
them, ambition is energetic social engagement reflected in the term Surgency, an
alternative label for Extraversion. Researchers who favor the proactive view of
Conscientiousness see conformity as an aspect of Agreeableness and nonconformity as an
aspect of Openness to Experience.
Johnson and Ostendorf (1993) further observed that Impulsivity has been a
particularly slippery concept in psychology. Eysenck initially saw impulsivity as part of
Extraversion until he relocated the concept to his Psychoticism dimension, often
interpreted as low Agreeableness and low Conscientiousness in the B5M (McCrae &
John, 1992). In Costa & McCrae's (1992) NEO PI-R, impulsivity is a facet subscale
within their factor IV domain scale, Neuroticism. Finally, Johnson and Ostendorf (1993)
noted that the location of terms related to intelligence such as intelligent, intellectual, and
perceptive have depended up how a researcher conceptualized Factor V. Those favoring
the lexical view of the fifth factor as Intellect have located intelligence-related terms
within this factor, while those favoring an Openness view of the fifth factor have located
such terms within Factor III.
Johnson and Ostendorf's (1993) resolution to these disagreements (see also
Johnson, 1994a) drew upon an extension of the five-factor lexical model called the
Abridged Big-Five Circumplex (AB5C; Hofstee, de Raad, & Goldberg, 1992). The
impetus for the AB5C model was the observation that the very best efforts to identify a
set of personality trait words that show high loadings on one and only one factor (e.g.,
Goldberg, 1992) have not been totally successful. Hofstee, et al. (1992) decided that this
is because very few—if any—personality trait words reflect only one factor. More often,
trait words show appreciable factor loadings on two or more factors. If one were to
completely specify a particular trait word in terms of the orthogonal Big-Five factors, one
would have to include its loadings on all five factors. Geometrically, each trait word
would be represented by a point in five-dimensional hyperspace, with the five loadings
corresponding to the coordinates on the five axes.
Fortunately, Hofstee el al. (1992) argue, a five-dimensional model is unnecessary
because most personality terms show only one appreciable secondary factor loading in
addition to the main loading. Personality terms can therefore be located in one of the 10
two-dimensional planes defined by a pair of Big-Five factors; hence the expression
"Abridged 5-dimensional Circumplex" (AB5C). For example, the terms alert, ambitious,
firm, and purposeful showed a primary positive loading on Conscientiousness with a
secondary positive loading on Extraversion (Hofstee, et al., 1992). In AB5C terms, they
would be classified as III+I+ traits. In contrast, the terms careful, cautious, punctual,
formal, thrifty, principled, and circumspect show a primary positive loading on
Conscientiousness but a secondary negative loading on Extraversion (denoting a
tendency toward introversion). These terms would therefore be designated III+I-. Such
differences in secondary loadings, argue Johnson and Ostendorf (1993), could explain
differences in the way researchers conceptualize and measure the five factors and locate
trait terms within the factors. A researcher who conceptualizes Factor III as organized
purposefulness is likely to use a measure of that dimension that contains many III+I+
items. A second researcher who conceptualizes Factor III as inhibitory impulse control is
more likely to use a measure of that dimension that contains many III+I- items. The fact
that terms such as ambitious show appreciable loadings on both Conscientiousness and
Extraversion could explain why some researchers locate the term on Factor III and some
on Factor I.
Geometrically, all of the III+I+ and III+I- terms listed above (as well as I+III+,
I+III-, I-III+, I-III-, III-I+ and III-I- terms) can be graphed in a plane where Extraversion
forms the vertical axis, and Conscientiousness, the horizontal axis (Hofstee et al., 1992).
Their exact distance and direction from the origin (0,0) can be defined by coordinates
representing the magnitudes of loadings on the two factors. Hofstee, et al. (1992)
suggested a further simplification, ignoring the distance from the origin and projecting all
Conscientiousness-Extraversion-related terms onto a circle (circumplex) with its center at
the origin. The terms would then be described simply by their degree location within the
360-degree circle. Within the Factor I x Factor III circumplex, some terms with their two
highest loadings on these factors will have secondary loadings that are so relatively small
that they will appear within 15 degrees of either the Factor I or Factor III axes. These
I+I+ or I-I- and III+III+ or III-III- terms Hofstee, et al. (1992) called factor-pure. Other
terms lie between the axes but closer to one than the other. For example, ambitious lies
closer to III+ than I+ and is therefore designated as III+I+, an extraverted form of
conscientiousness. A I+III+ term such as competitive lies closer to Extraversion than
Conscientiousness and is therefore seen as a conscientious form of extraversion.
By taking into account the secondary loadings of personality trait terms, the
AB5C lexical model can account for variations and commonalities in the measurement
and interpretation of the five factors. For example, Factor V scales originating in the
work of Norman (1963) and Goldberg (1992), who interpreted the factor as Culture or
Intellect, respectively, were found to have a V+III+ (controlled intellect; see Peabody &
Goldberg, 1989) character. In contrast, the Factor V scale based on adjectives chosen by
McCrae and Costa (1985), who interpret Factor V as Openness to Experience, had a
V+I+ (expressive intellect; see Peabody & Goldberg, 1989) character. Hogan and
Johnson (1981) authored two different sets of Factor V adjective scales. The first they
initially called Intellectance, later renamed Mentality (Johnson, 1997), which was
designed to reflect an Intellect version of the factor, and a second scale they initially
called Ego Control, and later, Novelty (Johnson, 1997) which is similar to Openness to
Experience. The Intellectance/Mentality scale was found to have a V+III+ character,
while the Ego Control/Novelty scale had a V+I+ character.
According to Johnson and Ostendorf's (1993) analyses, researchers who measured
the Big Five with trait adjectives chose adjectives whose secondary loadings supported
their preferred variant of each factor (e.g., Factor V as intellect vs. openness to
experience; Factor III as organized purposefulness vs. inhibitory impulse control).
Johnson (1994a, 1994b) found that personality questionnaire scales from different
research programs also showed different secondary factor loadings when mapped onto
the AB5C model.
Although the AB5C model can explain different interpretations of the five factors,
it cannot say which interpretations are more absolutely correct or fundamental. There are
no grounds for saying that organized purposefulness (III+I+) or inhibitory impulse
control (III+I-) is a better way to conceptualize or measure Conscientiousness.
Furthermore, there is no absolutely correct rotation of a pair of factors in defining any of
the ten circumplexes in the AB5C model. Prior to the AB5C model, McCrae and Costa
(1989) observed that the Wiggins Interpersonal Circumplex, defined by a vertical axis of
status (dominant vs. submissive) and a horizontal axis of love (warm vs. cold-hearted)
was a 30-to-45-degree rotation of a circumplex defined by their Extraversion and
Agreeableness factors. To McCrae and Costa, Wiggins' dominance is a disagreeable form
of their own Extraversion, while Wiggins' warmth is an agreeable form of Extraversion.
But from Wiggin's perspective, Extraversion could be called a warm form of dominance,
and Agreeableness, a submissive form of warmth.
Johnson and Ostendorf (1993) and Johnson (1994a), although affirming that the
five factors cannot be absolutely, objectively defined, offered a method of resolving
differences across lexical and questionnaire research programs. They suggested
aggregating data from theoretically diverse measures of the five factors and then applying
Hofstee, et al.'s AB5C algorithm. Items mapped as "factor-pure" (e.g. I+I+, I-I-, II+II+,
etc.) from the aggregated data would represent a core of agreement across diverse
research programs. In their analyses they found the following consensual cores for each
factor: Factor I, Social Communicativeness (extraverted, frank/open, fun-loving,
sociable, talkative, straightforward); Factor II, Softness (acquiescent, mild/gentle, soft-
hearted); Factor III, Constraint (careful, fussy/tidy, hardworking, neat, punctual,
scrupulous, thrifty, well-organized); Factor IV, Freedom From Negative Affect (calm);
and Factor V, Creativity (artistic, creative, imaginative).
Preferred Formats for Assessing the Big-Five Factors
Despite the fact that the B5M/FFM originated in analyses of individual trait
adjectives, trait-adjective scales such as those proposed by Goldberg (1992) have been
used in research and practice far less often than questionnaire measures of the five
factors. Goldberg (1999) provides three reasons why individual trait words are not ideal
for personality assessment. First, the finite number of trait words in any language limits
researchers' ability to assess complex nuances of personality. Second, trait adjectives and
type nouns are relatively abstract. This abstractness can be associated with ambiguity
about the meanings of these words without behavioral or contextual specification. Third,
one-to-one translations of single words often cannot be found across languages, limiting
international, collaborative research. Also, researchers such as Block (1995) and McCrae
(1990; McCrae, Costa, & Piedmont, 1993) have questioned whether words from ordinary
language can be sufficient for a scientific analysis of personality, because a mature
science (for example, physics) uses concepts that transcend the appearances found in
ordinary experience and expressed in everyday language. Astronomy tells us that the
earth orbits the sun rather than vice-versa, biology tells us that whales and bats are
mammals like us, not fish or birds, and physics tells us that time is relative rather than
Consequently, Hofstee, et al. (1992) and Goldberg (1999) recommended assessing
personality with short phrases rather than individual trait words from ordinary language.
This is the format of items in the International Personality Item Pool (IPIP; Goldberg, et
al, 2006), which contains several inventories for assessing the five factors, and the items
in the widely-used Big-Five Inventory (Soto & John, 2016). Very short, adjective-based
measures such as the Ten Item Personality Inventory (TIPI; Gosling, Rentfrow, &
Swann, 2003) continue to be used in projects with serious time constraints, but
questionnaires employing short phrases or sentences are used more often, particularly
when researchers want to assess subdomains of the Big-Five factors (e.g., Johnson,
2014). The NEO PI-R remains a widely-used personality questionnaire for measuring the
five factors and is sometimes called the "gold standard" for Five-Factor measurement
(Muck, Hell, & Gosling, 2007).
Big-Five Subfactors and Superfactors
Cattell. The Big-Five Model has almost always been seen as hierarchical, which
is to say that each of the five broad traits in the model has been defined by a cluster or
interrelated, narrower traits or subfactors. Each of Cattell's (1943) basic, lexical rating
scales was not a single trait adjective, but, rather, the labeling of a set of adjectives
judged by Cattell to have a common underlying meaning. For example, he listed his first
rating variable (Cattell, 1943, p. 490) as follows:
Alert . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Absent-minded
Observant, vigilant, omnipercipient Dreamy, indefinite, depersonalized
Therefore, even each basic rating scale in Cattell's original study was hierarchical,
representing a broader meaning perceived by Cattell in a set of individual trait words.
Cattell then clustered his 171 ratings even further by identifying ratings that correlated
empirically. His resulting 60 clusters therefore represented yet another hierarchical level
—clusters of adjective clusters.
In subsequent rating studies (e.g., Cattell, 1947), Cattell augmented the 35 rating-
scale clusters he retained for factor analyses by adding descriptions of each pole of the
rating scales. Raters were no longer given rating scales anchored only by sets of trait
words. Instead, the rating scales contained Cattell's summary interpretations of each pole.
For example, instead of a scale anchored by Cooperative . . . Obstructive (with a few trait
adjectives listed under each term; Cattell, 1943) Cattell (1947, p. 201) used the following
rating scale:
1. Readiness to cooperate
Generally tends to say yes
when invited to cooperate.
Outgoing. Ready to meet
people at least halfway.
Finds ways of cooperating
despite difficulties.
2. Obstructiveness
Inclined to raise objections to
a project, cynical or realistic.
"Cannot be done." Uninterested
or unfavorable attitude to
joining in. Inclined to be "difficult."
These written descriptions encompassed a range of behaviors, perhaps introducing yet
another level of hierarchical abstraction. Cattell's description-enhanced rating scales were
the rating scales used by Tupes and Christal (1961/1992) and by Norman (1963).
Therefore, when the Big Five were identified by these researchers as five very general,
abstract dispositions, Cattell had already created, through his own personal judgment,
several layers of hierarchical organization beneath the Big Five.
Norman and Goldberg. Subsequent lexical studies by Norman (1967) and
Goldberg (1990) returned to the level of individual trait adjectives, but, again, trait words
that were judged to be similar in meaning were grouped together to form synonym
clusters. Goldberg (1990) demonstrated that factor-analyzing Norman's 75 clusters
yielded the Big Five, but he questioned whether it was necessary to include all 1,431 of
the adjectives chosen by Norman, and whether one might group the adjectives differently,
using more objective techniques. Goldberg (1990) therefore selected 479 terms from
Norman's list and regrouped them into 133 synonym clusters with the help of
lexicographers using dictionaries and synonym finders. After collecting new ratings on
the 479 terms, Goldberg (1990) further reduced the terms to 339 trait adjectives
organized into 100 clusters. Goldberg (1990, p. 1223) described the factors that emerged
from a factor analysis of the 100 clusters as "nearly perfect examples of the Big-Five."
The 100 clusters (25 for Extraversion, 32 for Agreeableness, 23 for Conscientiousness, 9
for Emotional Stability, 11 for Intellect) remain as one standard intermediate level
between individual trait adjectives and the broad lexical Big Five.
Hogan. Before items were written for the earliest version of the Hogan
Personality Inventory, Robert Hogan (1986) had his graduate students think of as many
different ways a person could present himself or herself to create an impression on the
Big Five. For example, Hogan's students thought of six ways people present themselves
as agreeable or likeable: being easy to live with, being even-tempered, caring about
others, trusting others, liking people, and taking others' opinions into consideration.
These six Factor II categories became targets for writing items. After data for these
provisional items were collected, item analyses confirmed which items were answered
similarly within each category of each Big Five dimension. These item clusters, generally
varying in length between three and eight items, were called homogeneous item
composites or HICs. Further analyses after additional data collection led to reassignment
of some items to other HICs and even some HICs to other Big Five factors (Hogan &
Hogan, 1992). Some of the HIC reassignments contradicted typical ways of thinking
about lower levels for the Big Five. For example, the Not Depressed HIC was reassigned
from Adjustment (Hogan's label for Emotional Stability) to Ambition (One of two HPI
Extraversion scales) because Hogan observed that ambitious people are rarely depressed.
Costa and McCrae. In the questionnaire tradition, Costa and McCrae's (1995)
facet subscales remain the most widely-used scheme of organizing the Big Five into
narrower traits. Costa and McCrae (1995) arrived at their facets by compiling a list of
traits that have been widely discussed in the literature on trait measurement and chosing
six traits that they thought best represented each of the Big Five domains. For example,
they decided that Neuroticism (the opposite end of Emotional Stability) was best
represented by Anxiety, Angry Hostility, Depression, Self-Consciousness, Impulsiveness,
and Vulnerability. They confirmed their rational assignment of items to these facet
subcategories and facets to the five domains by statistical analyses.
Intermediates between facets and domains. DeYoung, Quilty, and Peterson (2007)
demonstrated that it is possible to assess personality at an intermediate level between
Costa and McCrae's facets and the broad Big-Five domains. By factor-analyzing facet
scores from the NEO PI-R and an IPIP version of AB5C facet scales within each of the
five domains, DeYoung, et al. (2007) found two factors within each domain that were
broader than facets but narrower than the Big Five. Ten-item IPIP scales were developed
to assess each of the ten intermediate aspects of the Big Five and were placed in the
public domain at
Superfactors. Others have proposed that the Big Five traits themselves are
interrelated, forming even broader superfactors. Five orthogonally-rotated factors from
any factor analysis are statistical abstractions that are forced to be independent of one
another. But the actual rating scales or questionnaire scales from different Big-Five
domains do in fact correlate with one another and therefore be considered to point to
higher-level factors.
Typically, when oblique rather than orthogonal rotation is used, researchers have
found that Agreeableness, Conscientiousness, and Emotional Stability form one higher-
order factor, while Extraversion and Intellect (or Openness to Experience) form a second
higher-order factor (Digman, 1997). Digman labeled the first superfactor α and the
second superfactor β. Digman discussed the α and β superfactors from a number of
different theoretical perspectives and concluded that α represents the dispositions that
encourage normative moral development, accommodating one's self to the expectations
of society. Indeed, the three Big Five traits under α are the traits that tend to increase over
the life course. In contrast, Digman interpreted superfactor β as a tendency toward self-
enhancement, personal growth and the achievement of status and power. The typical
developmental path for the two Big Five traits under β is curvilinear: an increase during
the first part of the life course and a decrease toward the end. DeYoung, Peterson, and
Higgins (2002) provided an additional interpretation of α and β, which they call Stability
and Plasticity, in terms of the serotonergic and dopaminergic neurotransmitter systems.
They suggested that that Stability encourages conformity, while Plasticity discourages
Finally, Musek (2007) has suggested that if the α/Stability and β/Plasticity
second-order factors are allowed to correlate through oblique rotation, one can also posit
a third-order general superfactor, similar to the g factor in intelligence research. Musek
(2007) discusses possible interpretations of the General Personality Factor or GPF.
Noting that the GPF represents a blend of all aspects of personality that are positively
valued, he dismissed the idea that the GPF is just an artifact of biased responding.
Instead, he suggested that the GPF represents a profile of positive emotionality,
motivation, well-being, and self-esteem.
Despite the uncertain, nebulous path provided by Cattell's initial reduction of
Allport's 4,504 prime trait terms into 12 or 16 personality factors, the Big Five Model
eventually emerged from Cattell's work as a clear, widely-accepted framework for
conceptualizing the universe of personality traits at a broad level. As with any scientific
model, the BFM has not been without critics (e.g., Block, 1995, McAdams, 1992).
Furthermore, a number of issues about the BFM have not been fully resolved. These
include the best way to conceptualize and measure each of the Big-Five domains, the
replicability of the Big-Five across cultures and languages, whether significant factors
beyond the Big Five exist, the optimal way of dividing the Big Five into subfactors, and
whether the Big Five should be combined into superfactors.
Nonetheless, progress on these issues has been achieved, and research on the Big
Five has given a clear picture of their genetic origin, development over the lifespan, and
ability to predict significant life outcomes. Initially a purely descriptive endeavor
involving ordinary language, Big-Five theorizing has grown to include concepts ranging
from neurotransmitters to dyadic interactions to human evolution (Wiggins, 1996). The
documented success of the Big-Five Model for organizing and generating personality
research has led to a wide acceptance of the model today.
Allport, G. W., & Odbert, H. S. (1936). Trait-names: A psycho-lexical study.
Psychological Monographs, 47 (Whole No. 211). DOI: 10.1037/h0093360
Block, J. (1995). A contrarian view of the five-factor approach to personality
description. Psychological Bulletin, 117, 187-215. DOI:10.1037/00332909.117.2.187
Cattell, R. B. (1943). The description of personality: Basic traits resolved into
clusters. Journal of Abnormal and Social Psychology, 38, 476-506. DOI:
Cattell, R. B. (1946). Description and measurement of personality. World Book:
New York, 1946. DOI: 10.1037/0022-3514.63.1.146
Cattell, R. B. (1947). Confirmation and clarification of primary personality
factors. Psychometrika, 12, 197-220. DOI:10.1007/BF02289253
Costa, P.T., Jr., & McCrae, R.R. (1976). Age differences in personality structure:
A cluster analytic approach. Journal of Gerontology, 31, 564–570. DOI:
Costa, P. T., Jr., & McCrae, R. R. (1985). The NEO Personality Inventory
manual. Odessa, FL: Psychological Assessment Resources.
Costa, P. T., Jr., & McCrae, R. R. (1992). Revised NEO Personality Inventory
(NEO PI-R) and NEO Five-Factor Inventory (NEO-FFI): Professional manual.
Psychological Assessment Resources, Odessa, FL.
Costa, P. T., Jr., & McCrae, R. R. (1995). Domains and facets: Hierarchical
personality assessment using the revised NEO Personality Inventory. Journal of
Personality Assessment, 64, 21-50. DOI: 10.1207/s15327752jpa6401_2
de Raad & Van Heck G. L. (Eds.) (1994). The fifth of the big five [Special Issue]
European Journal of Personality, 8(4).
DeYoung, C. G., Peterson, J. B., & Higgins, D. M. (2002). Higher-order factors
of the Big Five predict conformity: Are there neuroses of health? Personality and
Individual Differences, 33, 533-552. DOI: 10.1016/S0191-8869(01)00171-4
DeYoung, C. G., Quilty, L. C., & Peterson, J. B. (2007). Between facets and
domains: 10 aspects of the Big Five. Journal of Personality and Social Psychology, 93,
880-896. DOI: 10.1037/0022-3514.93.5.880
Digman, J. M. (1990). Personality structure: Emergence of the Five-Factor
Model. Annual Review of Psychology, 41, 417-440. DOI:
Digman, J. M. (1997). Higher-order factors of the Big Five. Journal of
Personality and Social Psychology, 73, 1246-1256.
Digman, J. M ., & Takemoto-Chock, N. K. (1981). Factors in the natural language
of personality: Re-analysis, comparison and interpretation of six major studies.
Multivariate Behavioral Research 16, 149-170.
Fiske, D. W. (1949). Consistency of the factorial structures of personality ratings
from different sources. Journal of Abnormal and Social Psychology 44, 329-344. DOI:
Goldberg, L. R. (1971). A historical survey of personality scales and inventories.
In P. McReynolds (Ed.), Advances in psychological assessment (Vol. 2, pp. 292-336).
Palo Alto, CA: Science and Behavior Books.
Goldberg, L. R. (1980, Мау). Some ruminations about the structure of individual
differences: Developing а common lexicon for the major characteristics of human
personality. Symposium presentation at the meeting of the Western Psychological
Association, Ноnolulu.
Goldberg, L. R. (1981). Language and individual differences: The search for
universals in personality lexicons. In L. Wheeler (Ed.), Review of Personality and Social
Psychology: Vol. 2 (рр. 141-165). Beverly Hills, CА: Sage.
Goldberg, L. R. (1982). From Ace to Zombie: Some explorations in the language
of personality. In C. D. Spielberger & J. N. Butcher (Eds.), Advances in personality
assessment (Vol. 1, pp. 203-234). Hillsdale, NJ: Erlbaum.
Goldberg, L. R. (1990). An alternative "description of personality": The Big-Five
factor structure. Journal of Personality and Social Psychology, 59, 1216-1229. DOI:
Goldberg, L. R. (1992). The development of marker scales for the Big-Five factor
structure. Psychological Assessment, 4, 26-42. DOI: 10.1037/1040-3590.4.1.26
Goldberg, L. R. (1993). The structure of phenotypic personality traits. American
Psychologist, 48, 26-34. DOI: 10.1037/0003-066X.48.1.26
Goldberg, L. R. (1999). A broad-bandwidth, public domain, personality inventory
measuring the lower-level facets of several five-factor models. In I. Mervielde, I. Deary,
F. De Fruyt, & F. Ostendorf (Eds.), Personality Psychology in Europe, Vol. 7 (pp. 7-28).
Tilburg, The Netherlands: Tilburg University Press.
Goldberg, L. R., Johnson, J. A., Eber, H. W., Hogan, R., Ashton, M. C.,
Cloninger, C. R., & Gough, H. C. (2006). The International Personality Item Pool and the
future of public-domain personality measures. Journal of Research in Personality, 40,
84-96. DOI: 10.1016/j.jrp.2005.08.007
Gosling, S.D., Rentfrow, P.J., & Swann, W.B. Jr. (2003). A very brief measure of
the Big-Five personality domains. Journal of Research in Personality, 37, 504–528. DOI:
Hofstee, W. K. B., de Raad, B., & Goldberg, L. R. (1992). Integration of the Big
Five and circumplex approaches to trait structure. Journal of Personality and Social
Psychology, 63, 146-163. DOI: 10.1037/0022-3514.63.1.146
Hogan, R. (1985). Hogan Personality Inventory manual. National Computer
Systems, Minneapolis, MN.
Hogan, R., & Hogan, J. (1992). Hogan Personality Inventory manual. Hogan
Assessment Systems, Tulsa, OK.
Hogan, R., & Foster, J. (2016). Rethinking personality. International Journal of
Personality Psychology, 2, 37-43.
Hogan, R., & Johnson, J. A. (1981, August). The structure of personality. Paper
presented at the 89th Annual Convention of the American Psychological Association, Los
Angeles, CA.
John, O. P., Angleitner, A., & Ostendorf, F. (1988). The lexical approach to
personality: A historical review of trait taxonomic research. European Journal of
Personality, 2, 171-203. DOI: 10.1002/per.2410020302
Johnson, J. A. (1997). Seven social performance scales for the California
Psychological Inventory. Human Performance, 10, 1-30.
Johnson, J. A. (1994a). Clarification of factor five with the help of the AB5C
model. In B. De Raad & G. L. Van Heck (Eds.), The fifth of the big five [Special Issue].
European Journal of Personality, 8, 311-334. DOI:10.1002/per.2410080408
Johnson, J. A. (1994b). Multimethod replication of the AB5C model of
personality traits. In B. De Raad, W. K. B. Hofstee, & G. L. M. Van Heck, Personality
psychology in Europe, Volume 5 (pp. 42-49). Tilburg, The Netherlands: Tilburg
University Press.
Johnson, J. A. (2014). Measuring thirty facets of the five factor model with a 120-
item public domain inventory: Development of the IPIP-NEO-120. Journal of Research
in Personality, 51, 78-89. DOI: 10.1016/j.jrp.2014.05.003
Johnson, J. A., & Ostendorf, F. (1993). Clarification of the five factor model with
the Abridged Big Five-Dimensional Circumplex. Journal of Personality and Social
Psychology, 65, 563-576. DOI:10.1037//0022-3514.65.3.563
Kaufman, S. B. (2015, August 31). The difference between ExtrAversion and
ExtrOversion [Blog post]. Retrieved from
McAdams, D. P. (1992). The five-factor model in personality: A critical appraisal.
Journal of Personality, 60, 329-361. DOI: 10.1111/j.1467-6494.1992.tb00976.x
McCrae, R. R. (1990). Traits and trait names: How well is Openness represented
in natural languages? European Journal of Personality, 4, 119-229. DOI:
McCrae, R. R., & Costa, P. T., Jr. (1992). Discriminant validity of the NEO-PIR
facet scales. Educational and Psychological Measurement, 52, 229-237. DOI:
McCrae, R. R., & Costa, P. T., Jr. (1983). Joint factors in self-reports and ratings:
Neuroticism, extraversion and openness to experience. Personality and Individual
Differences, 4, 245-255. DOI:10.1016/0191-8869(83)90146-0
McCrae. R. R., & Costa. P. T., Jr. (1989). The structure of interpersonal traits:
Wiggins's circumplex and the five-factor model. Journal of Personality and Social
Psychology, 56, 586-595.
McCrae, R. R., Costa, P. T., Jr., & Piedmont, R. (1993). Folk concepts, natural
language, and psychological constructs: The California Psychological Inventory and the
five-factor model. Journal of Personality, 61, 1-26. DOI: 10.1111/1467-
McCrae, R. R., & John, O. P. (1992). An introduction to the five-factor model and
its applications. Journal of Personality, 60, 175-215. DOI: 10.1111/j.1467-
Muck, P. M., Hell, B., & Gosling, S. D. (2007). Construct validation of a short
five-factor model instrument: A self-peer study on the German adaptation of the Ten-
Item Personality Inventory (TIPI-G). European Journal of Psychological Assessment, 23,
166-175. DOI 10.1027/1015-5759.23.3.166
Musek, J. (2007). A general factor of personality: Evidence for the big one in the
five- factor model. Journal of Research in Personality, 41, 1213–1233.
Norman, W. T. (1963). Toward an adequate taxonomy of personality attributes:
replicated factor structure in peer nomination personality ratings. Journal of Abnormal &
Social Psychology, 66, 574-583 DOI: 10.1037/h0040291
Norman, W. T. (1967). 2800 personality trait descriptors: Normative operating
characteristics for a university population. Ann Arbor: Department of Psychology,
University of Michigan. ERIC Document Number xt/ED014738, retrieved from
Peabody, D., & Goldberg, L. R. (1989). Some determinants of factor structures
from personality-trait descriptors. Journal of Personality and Social Psychology, 57, 552-
567. DOI:10.1037/00223514.57.3.552
Saucier, G., & Goldberg, L. R. (1998). What is beyond the Big Five? Journal of
Personality, 66, 495-524. DOI: 10.1111/1467-6494.00022
Soto, C. J., & John, O. P. (2016). The next Big Five Inventory (BFI-2):
Developing and assessing a hierarchical model with 15 facets to enhance bandwidth,
fidelity, and predictive power. Journal of Personality and Social Psychology, Advance
online publication. DOI: 10.1037/pspp0000096
Tupes, E. C. and Christal, R. E. (1961/1992). Recurrent personality factors based
on trait ratings (USAF ASD Tech. Rep. No. 61-97). Aeronautical Systems Division,
Personnel Laboratory: Lackland Air Force Base, TX. Reprinted as Tupes, E. C., &
Christal, R. E. (1992). Recurrent personality factors based on trait ratings. Journal of
Personality, 60, 225-251. DOI: 10.1111/j.1467-6494.1992.tb00973.x
Wiggins, J. S. (ed.) (1996). The Five-Factor Model of personality: Theoretical
perspectives. New York: Guilford Press.
... Proper analysis and ordering of these data in terms of adequately selected psychological personality theory (e.g. OCEAN 4 type or other, more extensive and precise) [4] will allow screening of recipients and reaching a targeted sales offer to those whose psychological profile guarantees acceptance of a given project. As a result, a quick purchase will be not a surprise. ...
Full-text available
The ideas presented by Bernard Rudofsky in the exhibition and book entitled Architecture Without Architects are being proposed by the author as the starting point of discussion on the condition of the architectural profession. Rudofsky focused the world’s attention on architecture formed by the forces of Nature and as a result of natural development processes. The article raises the question of the future of the architectural profession seen from the perspective of the present time, whilst the technological development, in particular automation and computerization of design and construction processes, has gone so far that it is possible in the near future that man might be replaced by machines, which in turn may lead to a deep change in the architectural profession, or to its eventual complete disappearance
... Note 1. Although the labels the Big-Five and the FFM are often used interchangeably to refer to the consensus on their importance as major constructs of personality, they derive from different research traditions (see, e.g., Johnson, 2017). ...
Previous attempts to identify personality traits that enhance inclination to social anxiety (SA) have been limited by a tendency to focus on selected traits in isolation, rather than examining their interactions. Additional research is needed to better understand whether and how these dimensions are linked to SA. In a prospective study, it was examined how interactions between the Big Five personality factors predict SA symptoms. A total of 135 individuals, aged 18–50 years, were recruited. Personality traits were measured at baseline, and SA symptoms were assessed one month later. Results showed that low emotional stability was an independent predictor of higher levels of SA. Additionally, two significant interactions emerged: the interactions between extraversion and openness, and between openness and agreeableness predicted SA symptoms. At high openness, higher extraversion was associated with significantly lower levels of SA, suggesting that the interaction provides incrementally greater protection against SA. Thus, extraverts are likely to be protected against social anxiety symptoms, but more so the more open they are. Moreover, at high levels of agreeableness, low openness has been shown to be uniquely predictive for higher levels of SA symptoms, indicating that the combined effect of openness with agreeableness may be more important to SA than either trait in isolation. These findings highlight the importance of testing interaction effects of personality traits on psychopathology.
Most copyrighted personality inventories facilitate norm-referencing through illustrative tables, yet their application to the many fields relevant to personality measurement is constrained by the need for stakeholders to possess the requisite financial resources to access them. Using an IPIP-NEO-300 dataset from Johnson’s IPIP-NEO data repository, we created open-source norm tables for different age groups (14–17 years; 18–25 years; and 30+ years) within a combined standardization sample from the United Kingdom (UK) and Ireland ( N = 18,591). The newly created tables are freely available online ( ), and there is no need to ask for permission to modify them. We provide general instructions that can be used to create open-source personality trait norms for other countries, settings, and age groups, as well as gender-specific norms. There is great potential for these norms to be used in various settings and their open-source freedoms may encourage future collaborations and investigations.
Full-text available
Purpose This study strives to improve one’s understanding of tailored messaging as an organizational communication strategy that amplifies processes of organizational change at an individual level of personality traits. Design/methodology/approach A scientific experiment was conducted to test the effects of tailored messages on self-reported employee engagement during an organizational change process. Findings The results show that tailored messaging improves employee engagement for change when messages fit the specific needs of different personality types. Conversely, message tailoring can lower employee engagement when messages do not match personality types. Further, message tailoring has different impacts at different stages of a change project. Research limitations/implications An employee's ability to change as a function of his professional skill set as well as the project type (e.g. digital transformation project, post-merger integration project, leadership change project) should not be neglected in an overall model that aims to explain the success factors of change management. Practical implications Obviously, proper targeting, timing, as well as the implementation of a valid, legal and feasible method for identifying an employee's personality as well as other individual characteristics are equally important and challenging to improve change management outcomes. Originality/value This study adds value to the discussion on the efficacy of message tailoring as a communication strategy for organizational change.
The aim of this study was twofold. Firstly, it explored the relationships between morningness, Big Five personality traits, depression, and problematic Internet use (PIU). Secondly, it examined the possible mediating role of depression in these relationships. The following measures were used: the Internet Addiction Test, the Composite Scale of Morningness, the International Personality Item Pool–Big Five Markers–20, and the Center for Epidemiologic Studies Depression Scale. The participants in the study were 398 university students 18 to 30 (M = 20.37, SD = 2.29) years of age. The study found that PIU Total, PIU Mental Disorder, and PIU Time Management Disorder are positively related to depression and negatively related to morningness, Extraversion, Agreeableness, Conscientiousness, Emotional Stability, and Openness. PIU was more often found in individuals with an evening chronotype. The study confirmed the role of depression as a mediator in the relationship of personality traits and morningness to PIU Mental Disorder and PIU Time Management Disorder. For PIU Mental Disorder, the study revealed a full mediation effect of depression in the case of morningness, Emotional Stability, and Extraversion, and partial mediation in the case of Agreeableness and Intellect. For PIU Time Management Disorder, there was a full mediation effect of depression in the case of Emotional Stability, Extraversion, and Agreeableness, and partial mediation in the case of morningness and Intellect. The results highlight the importance of including depression when investigating the relationship between morningness and personality in studies of PIU.
Full-text available
Робота присвячена адаптації українською мовою методики TIPI (10-пунктового п’ятифакторного опитувальника рис особистості, розробленого С.Гослінгом, П.Ренфру і В.Свонном). Описано методологічні засади розробки методики автором та процедуру здійсненої адаптації. Дослідження проводилось в декілька етапів: апробація дослівного перекладу (233 особи), перевірка розширеного переліку дескрипторів і узгодження остаточної версії опитувальника (405 осіб), перевірка ретестової надійності методики (136 осіб). Опитувальник TIPI-UKR показав хороші результати при оцінці внутрішньої узгодженості шкал (крім шкали «емоційна стабільність»). Конструктна валідність методики перевірена за допомогою аналізу кореляційних зв’язків шкал TIPI-UKR з відповідними шкалами п’ятифакторних методик «Локатор великої п’ятірки» (адаптація Л.Бурлачука, Д.Корольова) та П’ятифакторного опитувальника особистості (адаптація А.Хромова). Перевірка тест-ретестової надійності показала стабільність і повторюваність результатів методики на високому рівні. Ключові слова: п’ятифакторна теорія особистості, психологічна діагностика, TIPI, адаптація тестів, валідність, надійність.
Full-text available
Three studies were conducted to develop and validate the Big Five Inventory-2 (BFI-2), a major revision of the Big Five Inventory (BFI). Study 1 specified a hierarchical model of personality structure with 15 facet traits nested within the Big Five domains, and developed a preliminary item pool to measure this structure. Study 2 used conceptual and empirical criteria to construct the BFI-2 domain and facet scales from the preliminary item pool. Study 3 used data from 2 validation samples to evaluate the BFI-2’s measurement properties and substantive relations with self-reported and peer-reported criteria. The results of these studies indicate that the BFI-2 is a reliable and valid personality measure, and an important advance over the original BFI. Specifically, the BFI-2 introduces a robust hierarchical structure, controls for individual differences in acquiescent responding, and provides greater bandwidth, fidelity, and predictive power than the original BFI, while still retaining the original measure’s conceptual focus, brevity, and ease of understanding. The BFI-2 therefore offers valuable new opportunities for research examining the structure, assessment, development, and life outcomes of personality traits.
This personal historical article traces the development of the Big-Five factor structure, whose growing acceptance by personality researchers has profoundly influenced the scientific study of individual differences. The roots of this taxonomy lie in the lexical hypothesis and the insights of Sir Francis Galton, the prescience of L. L. Thurstone, the legacy of Raymond B. Cattell, and the seminal analyses of Tupes and Christal. Paradoxically, the present popularity of this model owes much to its many critics, each of whom tried to replace it, but failed. In reaction, there have been a number of attempts to assimilate other models into the five-factor structure. Lately, some practical implications of the emerging consensus can be seen in such contexts as personnel selection and classification.
The IPIP-NEO (Goldberg, 1999) is a 300-item inventory that measures constructs similar to those in the NEO Personality Inventory (NEO PI-R; Costa & McCrae, 1992). Despite evidence for its reliability and validity, the IPIP-NEO is even longer than the original 240-item NEO PI-R. This article details the development of a 120-item version of the IPIP-NEO from an Internet sample (N=21,588) and the subsequent testing of its psychometric properties in Goldberg’s (2008) Eugene-Springfield community sample (N=481), two additional large Internet samples (Ns =307,313 and 619,150) and a local sample (N=160). Results indicate that the psychometric properties of the 120-item IPIP-NEO compare favorably to the properties of the longer form.
An experiment was arranged whereby all three judges independently classified into the four selected columns 300 representative words, drawn from the total list according to a principle of representative distribution. The results of this study may be expressed in percentages of the total number of terms that each pair of judges assigned to identically the same columns. Taking only the instances where all three judges agree, we find 141 words or 47 per cent of the list, whereas 6.25 per cent represents the chance expectation. Examining this average agreement more closely we next determine the peculiarities of each individual judge when his placements are compared with those of the other two judges. This analysis calls attention to perhaps the principal source of unreliability, namely the tendency of each judge to have a mental set of "leniency" favoring the inclusion of marginal or doubtful terms in one column rather than another. Four outside judges selected 130 of the 300 terms as strange and unfamiliar to them. The agreement of the three editors for this group of terms averaged only 45 per cent, as against 47 per cent for the total list, and 48 per cent for the remaining 170 more familiar terms. Apparently established usage and familiarity enhances but slightly the reliability of the placement.
Psychometric studies of the organization of the "natural language of personality" have typically employed rating scales as measurement medium and factor analysis as statistical technique. The results of such investigations over the past 30 years have varied greatly, both with respect to number of factors and with respect to the constructs generated. Re-analysis of the correlations of six studies, including the classical work of Cattell, indicated that the domain appears to be well described by five factors, with some suggestion of a sixth. The five factors were related across studies, using the Kaiser-Hunka-Bianchini method. Generally, the factors were highly related, with most indices of relatedness exceeding .90. The five-factor model was tested by the multiple-group method, used to factor a large-scale study of teachers' ratings of children. With slight modification of the originally hypothesized structure, the five-factor model accounted for the observed relationships quite well. The five constructs suggested by the factors appear to be domains of research effort and theoretical concern which have long been of interest to psychologists.