What a career psychometric is for, and what it is not
It cannot tell your child what to be. It can tell you which conversations you have been avoiding, which is more useful and much less comfortable.
A report comes back a fortnight after a forty minute online test, twenty-two pages of it, and on page four there is a list of job titles: architect, biotechnologist, product designer. The family reads that list, argues over it for three evenings, forwards it to an uncle in Dubai, and never opens the other eighteen pages.
That list was the least defensible thing in the document, and whoever built the template printed it in the largest type. Something worth the money was in there, two pages earlier and far less impressive.
Three instruments, three different questions
A career psychometric is rarely one test. It is two or three bolted together under a common cover, and they answer questions with very little to do with each other.
- An interest inventory. Almost all descend from John Holland's six types: Realistic, Investigative, Artistic, Social, Enterprising, Conventional. The items ask what kind of work you would rather be doing, and the output is a rank order. Nothing in it touches competence. A student can come back high on Artistic and be unable to hold a straight line, and nothing has gone wrong, because the items never asked about drawing.
- A personality measure. The five broad traits: openness, conscientiousness, extraversion, agreeableness and emotional stability. These describe disposition: how a person tends to behave across many situations rather than what they can manage in one. Conscientiousness is the trait with the steadiest relationship to school results, which will surprise nobody who has watched a Class 11 section for a term.
- A reasoning test. The only part of the bundle that measures capacity instead of taste. Verbal, quantitative, non verbal and spatial, reported in bands. What those bands carry, and what they do not, we set out in what a cognitive profile can tell you.
Bound into one PDF under one logo, all three start to look like the same kind of evidence, and a parent turning the pages has no way to tell which of them carries weight.
Where a short form stops being honest
Length decides what an instrument may claim. The NEO-PI-R, the questionnaire the facet structure came out of, runs to 240 items and takes most of an hour. It reports thirty facets, six under each of the five traits, and defends them on eight questions per facet.
Now take a ten item short form. Two questions per trait. Two are enough to place a seventeen year old in the upper or lower third of conscientiousness against others her age, and that placement is worth having. They cannot separate orderliness from self discipline from dutifulness, because each of those needs eight questions and there are two.
When a fifteen minute questionnaire scores your child on twenty facets, each to a number out of a hundred, the precision was manufactured somewhere between her answers and the printer. Sit her again on a different Tuesday and half those numbers move. The five traits underneath will hold.
The same limit runs on the interest side. Six types in rank order off thirty questions is a fair reading. A ranked shortlist of a hundred and twenty occupations off the same thirty questions is decoration.
Preference is not ability, and effort is not in it at all
If one line survives the twenty-two pages, make it this one. Interest tells you what a student would rather be doing and personality tells you how she goes about it, and neither reaches the question of whether she will be any good at the work. Whoever implies otherwise has changed the subject.
Effort is the awkward one, the variable families care about most and the one no self report reaches. A student who tells a form she is organised and finishes what she starts is describing the person she would like to be recognised as. Both of you know whether that survives a Sunday evening.
The strongest evidence you hold about effort was never on a psychometric. It is the marksheet, because marks carry effort and consistency mixed in with everything else. Geiser and Santelices followed roughly eighty thousand students through university, and school grades beat every admissions test in the study. We work through what that licenses in what Class 10 marks can and cannot decide.
The disagreements are what you paid for
You hold three sources of evidence about one teenager: measured interest, measured personality, and what the household knows from marks, teachers and fourteen years of watching. When all three agree, you have bought confirmation. Pleasant, and not worth much.
The contradictions are the whole return. Take a boy in the second year of a JEE programme who comes back high on Social and low on Investigative. What he likes is people and the arranging of them, and nobody in that house has ever asked him what he pictures himself doing at twenty-eight. That is a question somebody owes him, not a verdict on engineering, and it is cheaper asked before the second year's fees than after the degree.
A girl who has said medicine since Class 8, and whose report puts Investigative flat and Enterprising at the top, is worth twenty minutes behind a shut door.
What you have bought is a short list of places where the evidence about your own child contradicts itself. Each is a conversation somebody has been avoiding, usually for understandable reasons and for too long.
Claims that should make you close the report
A single job title. No combination of interest and personality scores produces the sentence "your son should be a corporate lawyer". Where a person ends up turns on what the family can pay for, who they meet at nineteen, and fifteen years of work not yet done. Three or four fields to go and look at is defensible. One job is theatre.
A code handed over as identity. Four letters, a colour, an animal, a half of the brain. The four letter instruments carry a known problem: a large share of people who sit one twice are given a different code within weeks, which is hard to square with a document telling a sixteen year old what she is. Traits run along a continuum and most people sit near the middle, so cutting each into two boxes keeps the label and loses the reading.
Learning styles. Visual learner, auditory learner, VARK. Thirty years of trials, and matching teaching to a stated style does not improve what a student learns. Where it appears on a report you have learned something reliable, though not about your child.
Anything about income. A projected salary band derived from a personality questionnaire is a feeling with a rupee sign on it. Nothing in the instrument reaches that far.
How this sits beside a cognitive profile and a stream selector
Three tools, three jobs, and the trouble starts when somebody asks one of them to do another's work.
A cognitive profile measures reasoning. It comes back in bands, says nothing about what a student wants, and is the only one of the three carrying information about capability. A stream selector is a decision procedure with a deadline in it: subject marks, teacher comment and the List of Candidates date weighed together to produce one answer for one decision inside a single term. Both are useful, and neither pretends to be a psychometric.
A psychometric has no deadline and no verdict at the end. It describes preference and disposition as they stood on the day, at a grain the questionnaire length can support, and hands back the pieces that do not fit. That last part is the work, and it happens at your dining table rather than on our servers.
Ask two things: how many items is the questionnaire, and does the report name occupations. A thirty item form reporting twenty facets and eleven job titles has answered both badly. The O*NET Interest Profiler, built on the same six Holland types, is free on the web. Sit it with your child one evening and see whether the order lands where you expected.
Whether you need one at all
Plenty of families do not. If your daughter gives the same answer on two different days when nobody is standing over her, if her subject marks point the same way, and if her class teacher agrees with both, you would be paying to be told what you already hold. Put the money into a tutor instead.
It earns its hour in the other case: the student who cannot say what he wants, or says something nobody in the house believes, or has gone quiet about a subject combination chosen for him two years ago. PRISM is our version, interests and personality read alongside measured reasoning, each at the grain its instrument can support and no finer. It is allowed to come back with all three agreeing and nothing to argue about.