Holland’s RIASEC model: six types, one hexagon and two readings

In 1959, the American psychologist John Holland proposed that work interests could be described by a few types, and that the same types also described work environments. The theory was revised over decades, and its most cited edition, from 1997, describes six types: Realistic, Investigative, Artistic, Social, Enterprising and Conventional. The initials give the model its name, RIASEC.

The central idea is congruence: people who work in a setting that resembles their interests tend to choose that path, to be more engaged, to perform better and to stay. It is a hypothesis about tendencies in groups of people, studied for decades, and not a rule about anyone’s fate.

Almost everyone has a little of each type. What changes from person to person is how strong each one is and the order among them.

The most common summary of a profile is the code made of the three highest letters, in order. An “IAS” profile enjoys investigative activities most, then artistic ones, then social ones. The code is handy for conversation and for searching occupations, but it is a summary: two people with the same code can have quite different profiles in the numbers.

That is why the Map shows all six scores, not only the letters. When two letters are tied, your report says that the order between them is not a real difference.

Holland drew the six types at the corners of a hexagon, always in the same order: R, I, A, S, E and C. Neighboring types are more alike, and opposite types are less alike: Realistic and Social, Investigative and Enterprising, Artistic and Conventional.

Two classic readings of the model come from the hexagon. The first is consistency, the distance between the first and the second letter of the code:

• high, when the two letters are neighbors (I and A, for instance);
• medium, when one type sits between them (I and S);
• low, when they are opposites (A and C).

Low consistency only means the combination is less common. It is not a flaw or an incoherence: some jobs call for exactly the tastes that rarely go together. In the Map, consistency is reading information and does not enter the occupation scoring.

The second reading is differentiation, the distance between the highest and the lowest type. A well-differentiated profile has clear peaks and valleys. A poorly differentiated, or flat, profile appears when a person enjoys almost everything or almost nothing.

In the Map, the scale runs from 0 to 4, and the profile is treated as flat when the gap between the highest and the lowest type is below 0.75. This cutoff is provisional and will be calibrated with data. A flat profile is not a bad result: the report says so openly and suggests exploring more than one field before narrowing down, instead of forcing an order the numbers don’t support.

The six types show up in research from many countries. The perfect hexagon doesn’t: a meta-analysis with data from 18 countries did not support the same circular order outside the US (Rounds & Tracey, 1996). The Brazilian adaptation of the O*NET® Interest Profiler, with 603 university students, found the six components, but the structural analysis did not confirm the hexagon (Teixeira & López Castillo, 2020).

That is why the Map rests on the six scores and the code. The hexagon is the drawing that organizes the reading, and the occupation scoring uses no distance on the hexagon.

The effects are modest and real. In a review of 92 studies, the fit between a person’s profile and the job’s was related to performance at 0.32, against 0.16 for the interest score alone (Nye et al., 2017). A meta-analysis of 74 studies found links with training performance and with leaving a job (Van Iddekinge et al., 2011).

The link with job satisfaction is smaller: 0.19 in a review of 105 studies (Hoff et al., 2020). And interests measured at the end of high school predicted life outcomes 10 years later, beyond what IQ and personality already explained (Stoll et al., 2017). It is a good reason to measure interests alongside reasoning, not instead of it.