Morality

Honest people underreport the truth just to avoid looking like liars — the same fear driving your finished-task guilt

Writing a task you already finished onto your list — just to cross it off — is not cheating. Completion is a real reward your mind registers, and that little line through the words does honest work. The guilt some people feel about it rests on a misread of what cheating even means. So why does a harmless checkmark feel like padding?

Three sealed questions

This page makes three promises before it starts. Each seal opens at the section that keeps it — if one stays shut, the page owed you better.

Before the sections open, the figures this page stands on — each one carrying its own source.

The published figures, first

your answer first

The current published figures for Psychology says people who write already-finished tasks on their to-do list just to cross them off aren’t cheating — completion is a real cognitive reward, and the crossing-off is doing legitimate work, quoted straight from the official source with their receipts — before anything else on this page.

Your own amount depends on your income, assets and circumstances — confirm it with the official source.

Is writing a done task on your to-do list cheating?

Start with the fear behind the habit. You finish something, then jot it down just to draw a line through it. A quiet voice says that feels like padding. It is not. Cheating means deceiving someone to gain something you did not earn. Adding a real, finished task and crossing it off deceives no one and takes nothing. The work happened. The line simply records it.

Researchers who study honesty have spent years measuring how and when people actually cheat, and their picture is far narrower than the guilt suggests. Cheating shows up when there is a payoff and a person to fool. A checkmark on your own list has neither. So the completion you feel is a genuine reward for finishing, and marking it down is a fair way to see your effort.

Does crossing off a finished task give a real reward?

Yes, and the reward is worth naming. Completion registers as a small, genuine payoff — you set out to do a thing, you did it, and the record confirms it. Writing the task after the fact does not manufacture that feeling out of nothing, because the work behind it was real. The line through the words points at effort that already happened.

Set that beside dishonesty, which the research ties to a prize you did not earn and a person you mislead. A finished-task checkmark carries neither an unearned reward nor a fooled audience. So the good feeling is honest on its face. Treat the list as a mirror of your day, and the crossing-off becomes a fair way to close a loop.

First — a note to someone else

Someone you love is having their hardest day with replication right now. Not you today: them. What would you actually say?

Everyone believes their own eyes. The spotter below is where that confidence goes to be audited.

Cognitive-bias spotter

Biases aren’t stupidity — they’re shortcuts that were useful on the savanna and misfire in a world of headlines and spreadsheets. The research on replication above runs on the same machinery. Check the ones that sound like you; naming a bias is most of defusing it.

How often do people actually lie? A 44,050-decision study

One of the largest looks at everyday dishonesty comes from Gerlach, Teodorescu and Hertwig (2019). Their meta-analysis pooled 565 experiments and 44,050 separate decisions across the main cheating tasks — sender-receiver games, die rolls, coin flips, matrix puzzles. About 42% of men and 38% of women lied. Most who cheated cheated only a little. Very few pushed a lie to the maximum.

Two things nudged the numbers: the size of the reward, and the honesty people saw around them. That matters here because it shows what real cheating needs — a prize worth grabbing and a rule worth bending. Writing down a task you already did offers no prize and bends no rule. Hold that scale in mind the next time the checkmark feels suspect, and let the record stand.

The figures below put your private checkmark next to what measured dishonesty actually looks like.

Cross-examination

This finding survived re-testing — so put it on the stand. File the objection every skeptic reaches for, and read the record’s answer. Sometimes your objection wins; that is exactly why the rest holds.

A page should show its load-bearing numbers plainly. These are this one’s, receipts attached.

The numbers this page stands on

Not decoration — load-bearing. Each figure below is a cited finding’s own sentence, receipt attached; the rest of the page is built around them.

Why people returned lost wallets more often when they held more cash

People are more honest than cold self-interest predicts, and one striking field test shows it. Wallets with no money were reported about 40% of the time. Wallets holding about $13.45 were reported about 51% of the time. In a three-country test with about $94 inside, reporting reached 72%.

More cash meant more returns, the reverse of what pure greed would do. Country by country, reporting ran from 14% to 76%. The pattern is steady: given the chance to pocket a stranger’s money, most people hand it back. Weigh that honesty against a private checkmark, and the checkmark reads as the mild, harmless thing it is.

The return rates by wallet contents show the pattern at a glance.

The picture of the numbers

Rates hide inside sentences. Here the page’s own cited figures are drawn instead: a hundred small squares, the record’s share of them lit. The exact figure rides beside every picture, untouched.

A page should show what it grows from. Here is the tree — every leaf quoted, every receipt attached.

What this page grows from

Every branch below is one of this page’s own sections; every leaf is a finding it stands on, quoted exactly, receipt attached.

The Ten Commandments honesty study that did not replicate

A famous 2008 finding claimed that asking people to recall the Ten Commandments made them cheat less, a sizable effect at d = 0.48. It became a go-to example of how a quick moral nudge steadies our behavior. When many labs tried to reproduce it, the effect disappeared.

Verschuere and colleagues (2018) ran a multi-lab Registered Replication Report — 19 core replications with 4,674 people, and 25 replications in all totaling 5,786. The moral-reminder group actually solved 0.11 more matrices than the neutral group, landing at d = -0.04. That points the opposite way from the original claim. The reminder did nothing. So the idea that a small prompt polices our honesty did not survive a fair, large retest.

The weight of the evidence

Every study is a brick. The ones that support a claim stack on one pan; the failed re-tests and honest boundaries stack on the other. Watch where the beam settles before you believe — and tap any brick for its receipt.

You have read enough about minds in general. This one maps yours — drawn live from your answers, with a citation under every claim.

The Cartographer

You’ve been navigating by a map you’ve never seen. Over thirty quick questions we’ll draw it — every line cited to the science, none of it invented. Answer honestly, not aspirationally; there are no right answers, only true ones.

The signed-pledge study PNAS retracted

Another headline idea met the same fate. A 2012 paper argued that signing an honesty pledge at the top of a form, rather than the bottom, made people report more truthfully. Kristal and colleagues (2020) put it to a preregistered test with 1,235 people, plus five further studies covering 4,559 more. They found no effect. PNAS then retracted the 2012 paper on 13 September 2021 after its data were found to be fabricated.

Two of the most repeated stories about nudging honesty collapsed. Honesty turns out to be sturdier and less theatrical than those demonstrations implied. Your habit of logging finished work never belonged in that moral drama, so give the guilt a rest and treat the checkmark as plain record-keeping.

Now for something slower. Step into the room and let it read how you move through it.

The Replication Room

This asks nothing about replication directly. You wake in a dark room and simply move through it; the room watches, and at the end draws the shape you left behind. For delight, not diagnosis.

And if any of this touched something raw, the help below is real, free, and answers at all hours.

Finding support

In crisis right now?+
In the US, the 988 Suicide & Crisis Lifeline gives free, confidential support — call or text 988. Anywhere in the world, findahelpline.com lists a line for your country.
Looking for ongoing help?+
Consider speaking with a licensed counselor. psychologytoday.com lets you search by concern and by insurance.

Why honest people underreport the truth to avoid looking like liars

Here is the answer to the question at the top. The checkmark can feel like padding for a reputation reason, not a moral one. Choshen-Hillel, Shaw and Caruso (2020) call it “lying to appear honest.” In a billable-hours scenario with 115 lawyers, 18% of those who had truly worked 90 hours said they would falsely report fewer, because the honest number looked too good to be believed. Lawyers who worked an ordinary 60 hours almost never did this.

Read that closely. People sometimes bend the truth downward to protect how they look, even when they did nothing wrong. The unease over crossing off a done task runs along the same track. You fear it reads as inflating your output, so you second-guess an honest record. Name that worry for what it is — reputation management — and the pressure eases.

Here is the same science again, told the way an evening actually goes.

Your replication, the film

Everything above describes how replication works. This describes yours. The film asks a few short things as it plays and builds each chapter from your answers — nothing is uploaded; it’s drawn on your screen and gone when you leave.

Remember the note you left? It has been waiting for you.

A note from a stranger

a note from a stranger

How to use a to-do list without feeling like you’re gaming it

A list works best when it mirrors real effort and gives you an honest read on progress. Write tasks at a size you can actually finish, log the ones you complete, and let the crossing-off mark true work. The reward of completion is real, and there is no one to deceive. A few plain habits keep the list honest and useful.

Mind solver

Most psychology writing ends where your real problem begins. Describe what’s going on in your own words; this is a signpost, not a diagnosis — it maps your words to what people in similar spots find useful, and above all to WHEN and where to bring in a real human. It never replaces one.

For the skeptics — rightly so — here is the research this page stands on, figure by figure.

What the research shows

  • Primary meta-analysis of 19 replications (n=4,674; 25 total replications, total N=5,786): moral-reminder group solved 0.11 MORE matrices than the neutral group (95% CI -0.09 to 0.31; Cohen's d = -0.04), numerically opposite the original d = 0.48.A multi-lab Registered Replication Report (Verschuere et al., 2018) of Mazar, Amir & Ariely's (2008) finding that recalling the Ten Commandments reduces cheating found no such moral-reminder effect, undercutting a widely cited dishonesty demonstration.
  • Across 17,000+ wallets in 355 cities / 40 countries, the 'owner' was emailed for 40% of no-money wallets vs 51% of wallets with ~$13.45 (P<0.0001); in the 3-country big-money condition (~$94) reporting reached 72%. Mean country reporting (conditions combined) ranged 14-76%.In Cohn et al.'s (2019) global lost-wallet field experiment, return rates rose with money in the wallet, contradicting rational self-interest.
  • Across 565 experiments and 44,050 decisions (sender-receiver, die-roll, coin-flip, matrix tasks), 42% of men and 38% of women lied; dishonesty was usually moderate (most cheat 'a little', few maximally), with reward magnitude and observed norms moderating it.Gerlach, Teodorescu & Hertwig's (2019) meta-analysis quantified how much people cheat across the main deception paradigms.
  • Shu, Mazar, Gino, Ariely & Bazerman's 2012 PNAS 'sign-at-the-top' effect failed in a 2020 PNAS replication (Kristal et al.; one preregistered direct replication n=1,235 plus five conceptual studies n=4,559, no effect); PNAS retracted the 2012 paper on 2021-09-13 after its data were found fabricated.The influential claim that signing an honesty pledge BEFORE (vs after) reporting reduces dishonesty did not survive replication and the original paper was retracted for fabricated data.
  • people underreport extremely favorable true outcomes because the truth would look like a lie.: In a billable-hours scenario with 115 lawyers, 18% of those who truthfully worked 90 hours said they would falsely underreport their hours (to avoid seeming dishonest), whereas those who worked the in-range 60 hours rarely did — a reputational, not self-serving, form of dishonesty.Choshen-Hillel, Shaw & Caruso (2020) documented 'lying to appear honest'

Common questions about crossing off finished tasks

Short answers to the doubts that come up most, grounded in what honesty research actually found.

One more thing before you go. Fold this page into a single sentence of your own — the when and the how, decided now.

The One Sentence

Everything above, folded into a single sentence you write yourself. Pre-deciding the when and the how is the strongest follow-through move the research knows.

And in the spirit of every receipt above: here is how the page itself was built, device by device.

How this page works on you

Every device this page uses to hold your attention, named and sourced. Sites built on dark patterns cannot print this panel without confessing; a site built on receipts can end with it.

This is general information about the mind, not therapy or a diagnosis. If things feel hard, please consult a professional. In a crisis, reach a free, confidential crisis hotline right away; findahelpline.com lists one for your country.

this page

Built on the record, not on vibes

ppw.kuleuven.be · tier S
Dishonesty moral reminder rrr 2018
A multi-lab Registered Replication Report (Verschuere et al., 2018) of Mazar, Amir & Ariely's (2008) finding that recalling the Ten Commandments reduces cheating found no such moral-reminder effect, undercutting a widely cited dishonesty demonstration.: Primary meta-analysis of 19 replications…
mpib-berlin.mpg.de · tier S
Dishonesty truth about lies meta gerlach2019
Gerlach, Teodorescu & Hertwig's (2019) meta-analysis quantified how much people cheat across the main deception paradigms.: Across 565 experiments and 44,050 decisions (sender-receiver, die-roll, coin-flip, matrix tasks), 42% of men and 38% of women lied; dishonesty was usually moderate (most cheat…
davetannenbaum.github.io · tier A
Moral civic honesty wallet money cohn2019
In Cohn et al.'s (2019) global lost-wallet field experiment, return rates rose with money in the wallet, contradicting rational self-interest.: Across 17,000+ wallets in 355 cities / 40 countries, the 'owner' was emailed for 40% of no-money wallets vs 51% of wallets with ~$13.45 (P<0.0001); in the…
pnas.org · tier A
Dishonesty signing at top retracted failed
The influential claim that signing an honesty pledge BEFORE (vs after) reporting reduces dishonesty did not survive replication and the original paper was retracted for fabricated data.: Shu, Mazar, Gino, Ariely & Bazerman's 2012 PNAS 'sign-at-the-top' effect failed in a 2020 PNAS replication…
apa.org · tier A
Dishonesty lying to appear honest choshenhillel2020
Choshen-Hillel, Shaw & Caruso (2020) documented 'lying to appear honest': people underreport extremely favorable true outcomes because the truth would look like a lie.: In a billable-hours scenario with 115 lawyers, 18% of those who truthfully worked 90 hours said they would falsely underreport…

This article was last reviewed on July 22, 2026. Psychology is a living science — where findings are contested or have failed to replicate, we say so in the text.