Your Confidence Is Measuring How Easy Something Felt, Not How Right You Are
A well known experiment found that generating more evidence of your own assertiveness can paradoxically make you feel less assertive, because the effort of remembering felt harder. The finding says something important about what confidence is actually built from.
Ask a group of people to list six times they acted assertively, then ask a second group to list twelve times. You would expect the second group, having generated twice as much actual evidence, to rate themselves as more assertive afterward. In a well known experiment using this exact setup, the opposite happened. The group asked for twelve examples rated themselves as less assertive than the group asked for only six.
The explanation researchers arrived at has nothing to do with the content of what people actually remembered and everything to do with how the remembering felt. Six examples of your own assertiveness come to mind fairly easily for most people. Twelve is hard. You run out of clear memories partway through and have to strain for the rest. That strain, the felt difficulty of the task itself, ends up mattering more to a person's final self judgment than the actual number of examples sitting in front of them. If generating evidence for a claim about yourself felt hard, people conclude, on some level, that the claim must not be very true, regardless of how much evidence they in fact produced.
This general pattern, known in cognitive psychology as processing fluency, has been documented well beyond this one experiment. Psychologists Adam Alter and Daniel Oppenheimer have shown across dozens of studies that people use the felt ease or difficulty of a mental task as a stand in for judging the quality of whatever that task produced, whether the task is recalling a memory, reading a statement, or estimating how well they know a subject. Statements printed in a clearer font get rated as more likely to be true than the identical statement in a harder to read font. Facts that come to mind quickly get treated as more reliable than facts that take a moment of digging, independent of whether the quick fact is actually the more accurate one.
This general pattern, known in cognitive psychology as processing fluency, has been documented well beyond this one experiment.
The effect shows up in places that have nothing to do with self judgment at all, which is part of what makes it convincing rather than a just so story invented after the fact. Alter and Oppenheimer found that in the weeks immediately following a company's stock market debut, shares with names or ticker symbols that were easier to pronounce reliably outperformed shares with more awkward, harder to say names, a gap that had nothing to do with the underlying businesses and everything to do with how smoothly investors could process the label. If fluency shapes something as impersonal as which stock ticker looks more trustworthy, it is not a stretch that it also shapes something as personal as how capable you feel in a given moment.
Applied to personal confidence specifically, this has a fairly direct and slightly unsettling implication. The feeling of confidence is not a reliable readout of competence. It is closer to a readout of familiarity and fluency, how easily an answer, a memory, or a sense of your own capability comes to mind in the moment. Early on at any new skill, everything about it feels effortful by definition, which means confidence will be low almost regardless of how well the underlying learning is actually going. Confidence tends to rise later mainly because the skill has become more fluent and familiar, which is a real signal worth trusting eventually, but is a completely different thing from the raw amount of genuine evidence for your own competence at any given moment.
The practical use of knowing this is mostly defensive. A task that feels unusually hard right now, a conversation, a presentation, a new job, is not reliable evidence that you are doing it badly, since difficulty and quality are measured by two separate systems that people habitually confuse for one. And on the other side, a swift, easy sounding answer, from yourself or from someone else, is not automatically the correct one just because it arrived without friction.
Ease is information about how familiar something feels. It was never actually information about whether it is true, and a man who remembers that distinction has a real defense against the specific, quiet kind of self doubt that shows up not because he is doing badly, but simply because something is still new.