Holland Codes Explained: The RIASEC Model of Career Interests

Holland codes turn a vague sense of what you enjoy into a few letters you can search with. Your Holland code is the two or three strongest of six work-interest types, ranked: Realistic, Investigative, Artistic, Social, Enterprising and Conventional, or RIASEC for short. The free O*NET Interest Profiler from the US Department of Labor scores you on those six types and uses them to suggest occupations.¹ The psychologist John Holland first published the theory behind them in 1959.²

Before you trust the results, know one limit. The latest systematic review we found, published in 2020 by Kevin Hoff and colleagues, found that a close match between your interests and your job goes with higher job satisfactionjob satisfaction: How positively a person judges and feels about their job, as a whole or in parts such as pay, colleagues or the work itself. It is usually measured by questionnaire and is related to engagement, but it is not the same thing.Full entry in the glossary, but a little more weakly than expected.³ So treat your code as a way to widen the list of careers you consider. If you are not yet sure you want to leave your field, begin with deciding if a career change makes sense for you.

What a Holland code measures: your interests

A Holland code measures interests, meaning what you like to do, not what you are good at. The ONET Interest Profiler Manual, published in 2021 for the US Department of Labor, describes interests as trait-like preferences for activities and settings that motivate effort, and it builds its six scales on Holland’s 1997 version of the theory. Your code is your highest scores in order: the first letter is your strongest interest, and the second and third add detail. ONET codes jobs the same way, and the manual’s example is personal financial advisors, coded ECS: mainly Enterprising, with Conventional and Social work mixed in.¹ An advisor still keeps careful records and still listens to worried clients, even though selling and persuading lead.

Holland’s central idea was that people and workplaces can be described in the same six terms.¹ Someone with strong Social interests is drawn to teaching, helping and caring; a school or a clinic is a Social environment that asks for exactly that. The third edition of his book, Making Vocational Choices, was published in 1997 by Psychological Assessment Resources (PAR), a company that also sells tests built on the model; its subtitle names both halves, vocational personalities and work environments.⁴

The manual sums up each type in a line and gives its own example careers.

Read your own code as a ranking of what draws you, with the lower letters still present. It is not a label that rules careers in or out, and a job coded differently from you may still hold plenty of the work you like.

How the free O*NET Interest Profiler finds your matches

The ONET Interest Profiler scores your answers on the six RIASEC scales, compares the shape of your profile with an interest profile for each occupation in ONET’s database, and lists the closest matches at the level of training you choose. It is free, and its results link to more than 900 US occupations.⁵ The web version asks you to picture yourself doing 30 work activities and rate each from strongly dislike to strongly like, setting aside pay and schooling for now.⁶

1. 1Rate activitieslike or dislike, ignoring pay and training
2. 2Score six scalesyour top three letters form your code
3. 3Choose a Job Zonehow much preparation you have or want
4. 4Compare shapesyour peaks against each job profile
5. 5Read the matchesBest, Great and Good Fit

Two design choices shape the list you get. First, expert raters following a standard coding procedure gave every occupation its own six-number interest profile, and the Profiler checks how closely the pattern of your highs and lows matches each one, not how high your scores are. Second, O*NET sorts occupations into five Job Zones by how much education, training and experience they need, because every interest area contains jobs at several levels; the same profile can produce one list at Job Zone Three, which needs a medium level of preparation, and another at Job Zone Five, which needs extensive preparation. The manual suggests comparing careers in your current zone with those in a zone you could reach later, and it labels the matches Best Fit, Great Fit or Good Fit.¹

Shape matching has a practical upside. Someone who mildly likes everything and someone who loves everything get similar suggestions if their peaks fall in the same places, so a cautious answerer is not penalized for being cautious.

A paid alternative, the Self-Directed Search, was written by Holland himself, with Melissa Messer on the fifth edition, and is sold by PAR, which also publishes his 1997 book.⁷ Aptitude tests answer a different question: what you can do, not what you like. Personality-type tools such as the Myers-Briggs come from a separate model and do not produce a Holland code.

The manual reports that scores on the longer 60-item version stayed highly stable when people took it twice, and that the 30-item version built for phones is also adequately consistent, though its reliability figures are a little lower than the longer form’s.¹ Two of the manual’s editors, James Rounds and Kevin Hoff, also co-wrote several of the studies discussed below, so read it as the developers’ own account rather than an independent review.

For your own run, take the Profiler once at the zone that fits your plans, then look at the next zone up before you rule a field out.

The RIASEC hexagon: why neighboring types go together

Holland placed the six types around a hexagon in a fixed order, R-I-A-S-E-C, so that types next to each other are more alike than types further away or directly across. The O*NET manual’s example is that Realistic work with tools and machines sits close to Investigative work in science and technology.¹

1. Neighbors: types next to each other, such as Realistic and Investigative, are more alike than types farther around
2. Opposites: types directly across, such as Realistic and Social, sit furthest apart and are less closely related than neighbors
3. Your code: your three strongest types in order; in this example S, then I, then A

The shape carries a practical message. Read around it and Social sits across from Realistic, Enterprising across from Investigative, and Conventional across from Artistic. A code made of neighbors, such as RI or SE, describes interests that sit close together in the model. A code that joins opposite corners, such as RS, pulls in two directions. Someone with that pattern may need a role that mixes both, such as teaching a practical trade, or may meet one of the interests outside work; that is our reading, not a research finding.

Does the hexagon match real data? In 1993, Terence Tracey and James Rounds tested it with a structural meta-analysismeta-analysis: A study that combines the results of earlier studies on the same question into one overall estimate. Pooling makes the estimate more precise, but it cannot repair the studies it pools: a meta-analysis of surveys is still survey evidence.Full entry in the glossary of 104 published correlation tables from 1965 to 1989. Holland’s ordering fitted adequately and beat a rival three-group model that Gati had argued for in 1991.⁸

Outside mainstream US samples the picture was less tidy. A 1996 follow-up by the same researchers, covering data from 18 countries and from US ethnic groups, found that the circular order did not hold equally across cultures, and none of the models tested described the US ethnic-group samples adequately.⁹ Both tests are decades old, and we found no newer review of the same scope that we could read in full.

Treat the neighbors-are-closer rule as a tendency, best supported in US data. If you live elsewhere, read your second and third letters loosely.

Does matching your Holland code make you happier at work?

The best evidence shows only a modest link, and it cannot show cause. Kevin Hoff and colleagues’ 2020 meta-analysis found interest fit, the match between your Holland code and your job’s, linked to higher job satisfaction, but slightly more weakly than the authors expected. It also says earlier meta-analyses, which reported no statistically significant link, were held back by problems including low statistical power.³

Myth

Find the job that matches your Holland code and you will be happy in it.

Fact

Across decades of studies, interest fit is linked to job satisfaction only modestly; it is more strongly linked to performance.

A 2005 meta-analysis by Orit Tsabari, Aharon Tziner and Elchanan Meir, pooling 26 studies published from 1988 to 2003, shows why the question is slippery: the link depended on how fit was scored and which interest questionnaire was used, and it differed between studies from the US and from Israel.¹⁰ The answer depends partly on the ruler, which is one reason a single test result should not carry a big decision.

The study

Moderate evidence

Does interest fit predict satisfaction? Hoff's review of 105 studies, 2020

A closer match between people’s interests and their jobs was linked to higher overall job satisfaction, with a correlation of 0.19, which the authors call slightly lower than expected. The link was notably stronger for satisfaction measures that capture how people rate their career choice in general.³

A correlation of that size is small. Across a large group it is real, but for any one person it leaves most of the story about liking a job to other things. The studies are also correlationalcorrelation: A measure of how closely two things move together, running from minus one, where one rises as the other falls, through zero, meaning no link, to plus one. It says how strong the relationship is, not what causes it, and it is not a percentage.Full entry in the glossary, which means they cannot tell you that moving to a matching job would lift your satisfaction.

Two people with the same Conventional-first code might do the same bookkeeping job. One has a fair manager and a team she likes; the other has neither. Their shared code cannot tell you which of the two dreads Monday.

Where interest fit counts more: effort and performance

Interest fit is more strongly linked to how well people perform than to how satisfied they feel in the job, according to Hoff and colleagues’ 2020 review.³ A 2012 meta-analysis of 60 studies spanning more than 60 years, led by Christopher Nye with Rong Su, James Rounds and Fritz Drasgow, found interests related to performance and persistence at work and in education, and fit measures were more closely tied to performance than interest scores alone.¹¹

The same team’s larger 2017 update, of 92 studies, found the correlation for fit twice that for interest scores alone, and it set out to address a disagreement with an earlier review.¹² That review, Chad Van Iddekinge and colleagues’ 2011 meta-analysis, found weak links for a single RIASEC scale and stronger ones for interest measures built around a particular job or vocation.¹³ Both teams agree that interests are linked to performance; they differ on how much the fit calculation adds.

Why would interest show up in performance? The O*NET manual describes interests as preferences that motivate goal-oriented behavior, which fits the pattern: interest steers where people put their effort.¹ A new analyst who finds data puzzles absorbing may stay with a hard problem longer than a colleague who is there for the pay, and that extra time adds up. The evidence is correlational, though, so the direction is not proven.

When you compare options, weigh fit as a guide to where you would keep improving, and test that guess with real work before you rely on it.

Using your Holland code without overreading it

A Holland code works best as a starting list and a shared vocabulary. Each career the Profiler suggests links to O*NET’s page for that occupation, with its tasks, knowledge, education and pay, which is where the real checking starts.¹ For roles that look promising, comparing the skills your current job and a target job use shows what would carry over and what you would need to build.

Choosing between the careers on your list is a bigger process with its own steps.

The bottom line

Your Holland code ranks the kinds of work that draw you, using six types Holland arranged on a hexagon, and the free O*NET Interest Profiler turns it into a list of US occupations at the level of training you choose. Fit with a job is linked to performance more than to job satisfaction, so use the code to widen your options, then check each one against the actual tasks.