Two colleagues both answer “Often” to a statement about planning their work ahead. One keeps a weekly plan on paper. The other keeps a plan in their head and a stack of late tasks. Same answer, different realities. That gap is the central problem with any self-rating, including the Competency Test, and this article is about how wide it can get, why it appears, and how to use a self-rating well anyway.
Why the same answer can mean different things
A self-rating is a judgment, not a measurement. When you decide whether you “often” do something, you are comparing yourself with some yardstick, and you rarely choose the yardstick on purpose. This is the reference-frame effect: the people around you set the scale.
Picture a person who plans reasonably well, working on a team of exceptionally organized colleagues. Surrounded by color-coded calendars, they may rate themselves low on planning. Move them to a team where deadlines are discovered on the day and the same person may feel like a model of order. Their behavior has not changed; the comparison has. Words like “sometimes” and “often” also differ from person to person, so even the answer options are not fixed units.
This is partly why the Competency Test makes no comparison with other people. There is no norm group, so the result describes only how you describe yourself, with all the limits that implies. What a competency test is covers what that means for reading your levels.
What research on self-assessment says, in plain terms
Researchers have studied self-assessment for a long time, and the broad picture is consistent even where the details are debated. Most of this work concerns ability and performance rather than everyday habits, so treat it as a caution about self-ratings in general, not as a verdict on any one test.
- Kruger and Dunning (1999) found that people who performed worst on tests of skills such as reasoning and grammar tended to rate their own performance as much better than it was, while the strongest performers tended to rate themselves somewhat lower than their results justified. Our piece on the Dunning-Kruger effect covers that pattern in depth.
- A review by Dunning, Heath and Suls (2004), which looked at health, education and the workplace, concluded that people’s views of their own skills are often only loosely related to how they actually perform.
- A synthesis of many studies by Zell and Krizan (2014) found that people do have some insight into their abilities, but only a modest amount, and that it differs from one domain to another.
- An earlier meta-analysis by Mabe and West (1982) found that how well self-evaluations matched outside measures varied a great deal, and that they matched better when people expected their answers to be compared with something real.
None of this says self-ratings are worthless. It says they are noisy, and they tend to be less reliable when nothing will ever check the answer. It is also reasonable to expect more insight in work where feedback is quick and clear, since a missed deadline is hard to miss, than in work where feedback is vague.
Three pressures that tilt a workplace self-rating
- Social desirability. Most of us want to look competent, even to ourselves. Habits that sound admirable (“I plan ahead”) are easier to over-endorse than ones that sound dull.
- Consistency. Once you have answered “Often” a few times in a row, the next answer drifts the same way. Researchers who study common method bias (Podsakoff et al., 2003) point out that when one person supplies all the information, in the same format and at the same moment, shared influences such as the wish to look good and the wish to seem consistent can shape the answers. Reverse-worded statements are one defense: in the Competency Test, ticking “Almost always” all the way down does not produce a perfect score. That blocks one pattern. It does not remove the wish to look good.
- Recency. The latest events are the easiest to recall. A rough last two weeks can color how you rate the past year, and a great project last month can do the opposite.
Why a self-rating is still worth having
The honest answer to “is it accurate?” is “partly, and you cannot tell which part from the inside”. That sounds bleak, but it is where the value starts. A self-rating makes a private, fuzzy picture explicit, and an explicit picture can be examined, challenged and compared.
Most of its value is as a conversation starter. If you rate yourself Strong on communication and a colleague mentions they often have to ask you to repeat the point, you have found something neither the level word nor the remark would have shown alone. The gap between your picture and someone else’s is information. Naming concrete behaviors, such as “I check that people understood me”, also gives you something to watch for that “I’m a good communicator” never does. The same applies before an interview: how competency-based interviews work shows how a rough self-picture turns into usable examples.
Three ways to cross-check your own ratings
- Look for evidence. For your highest and your lowest competency, write down two recent, specific episodes with a rough date. If a Strong rating comes with no episode you can recall, treat it as a question rather than a result. If a low one comes with examples you had forgotten, that is worth knowing too.
- Ask a colleague. Choose someone who has watched you work and ask a question that is hard to answer with a pleasantry: “When did you last see me close a loop quickly, and when did you last see me let one hang?” Listen without defending. One view is only one view, so ask two people if you can.
- Re-rate after four weeks. Keep a short note each Friday about one occasion when each area came up, then take the Competency Test again. Four weeks is long enough for real occasions to happen and short enough to remember. If the result moves, check your notes before deciding whether your habits changed or only your attention did. Both are useful; they are different things.
Using the result without leaning on it
A workable rule: the more a result surprises you, in either direction, the more it deserves a check before you act on it. A surprising high is a prompt to find the evidence. A surprising low is a prompt to ask whether you compared yourself with an unusually strong team, or were thinking of one bad week.
Whichever way it goes, keep the claim small. The Competency Test on JobCannon shows how you describe your own habits; it does not measure real skill and does not predict job success. For the skills side of the picture, the Skills Audit is its closest sibling. For general rules on rating skill honestly, see how to self-assess your skill level honestly, and once you have picked an area to work on, build one work competency in 30 days.