Stories that sound like facts

Some conclusions have been carried so long they no longer feel like conclusions. They present as neutral self knowledge. "I am not good with people." "I work better under pressure." "I need more time before I start."

Two are worth looking at closely.

"I work better under pressure" is usually a description of the only condition under which the person has tested themselves. Work started at the last minute is compared against work never started, and it wins. It is rarely compared against the same work started a week earlier.

That comparison has actually been run. Ninety nine executive-education professionals were assigned either evenly spaced deadlines or deadlines they set themselves. The imposed-deadline group scored 88.8 against 85.7 on papers and 86 against 77 on the final project. In a second study, evenly spaced deadlines beat self-imposed deadlines, which beat a single end deadline, on errors detected and time on task (Ariely & Wertenbroch, 2002). People do impose costly early deadlines on themselves, roughly 42 days early on the first assignment, and they still do it worse than a fixed external schedule.

Note what that study does not say. Self-imposed deadlines beat no interim deadlines. The claim is not that self-management fails; it is that it underperforms structure.

"I need more time before I start" has a similar shape. The additional time is real, the requirement is not, and there is a well-replicated reason to distrust the estimate. Students predicting their own thesis completion gave a best guess of 33.9 days and a worst case of 48.6. The actual mean was 55.5 days, worse than their own pessimistic estimate, and only about 30 per cent finished within their prediction (Buehler, Griffin & Ross, 1994). In a later study 13 per cent finished by their 50 per cent confidence deadline.

The general rule: a story becomes dangerous when it stops describing an experience and starts controlling a future decision.

Some self descriptions are accurate and permanent, and treating them as stories to dismantle is dishonest. The distinction is not whether the statement is comfortable. It is whether it names a fixed constraint or a repeated outcome.

The stories that cause most trouble are the ones that were once true under conditions that no longer apply. Someone genuinely poor at presenting at twenty three, in an organisation that gave no feedback and no practice, formed the conclusion accurately. Fourteen years later the conclusion is intact and the evidence is fossil.

Watch for the ones carrying a small social benefit. "I am not good with people" is a conclusion and an exemption: it removes the obligation to initiate, and people will route around you. Conclusions with a benefit attached are the most stable, and they are rarely surrendered to argument.

There is a measurement problem underneath all of this that the course should not hide. Ten self-report procrastination instruments were compared against actual behaviour in 235 people. The best predicted days to completion at 0.19; several predicted it not at all; none predicted deadline-action pacing (Vangsness et al., 2022). Three experience-sampling studies found trait self-reports did not explain individual delay at all, while momentary judgements about the specific task did.

Read that in both directions. It supports the pessimism about narrative: your articulate account of your own pattern is weakly related to what you do. It also undercuts using that account as a diagnosis. This is why the exercises ask for calendars and dated acts rather than for self-ratings.