The sentence that works best in a busy house is usually the one nobody wants to say out loud. Not the sharp version. Not the long, careful, three-clause version either. Just the plain one: “Please put your shoes on now.” “I need the table cleared before dinner.” “Would you pick Sam up on Thursday?”
Most of us reach for something softer. We turn the request into a question, add a trailing “okay?”, or hint twice and hope. It feels kinder. In our experience it is mostly just harder to follow, and there is a growing set of research suggesting we are also wrong about how the plain version lands.
What the 2.6 percent study actually measured
In 2022, Nicole Abi-Esber, Jennifer Abel, Juliana Schroeder and Francesca Gino published a paper in the Journal of Personality and Social Psychology called “Just letting you know…”: Underestimating others’ desire for constructive feedback. It ran five experiments with 1,984 participants, plus a small field study with 155.
The Artful Age
A weekly letter on aging well, family across generations, and the creative life after the kids leave home.
The field study is the one people remember. Researchers administered surveys with something visible on their own faces, food or marker, and counted how many participants mentioned it. Only 2.6 percent said anything.
Across the experiments, people consistently underestimated how much the other person wanted to be told. The authors describe two reasons pulling in the same direction: we weigh our own discomfort in the moment heavily, and we overestimate how bad the other person will feel. When participants were prompted to take the receiver’s perspective, the gap narrowed more than with any other approach the authors tested. In the final experiment, withholding feedback tracked with less improvement for the person who did not get it.
Worth being precise about what this is. It is a set of studies on adults giving feedback to other adults, mostly in lab and online settings, not a study about families and not a study about children. It is a strong result across five experiments rather than a single finding, but it does not tell you what to say to your seven-year-old on a Tuesday. What it does establish is that the instinct to say nothing, or to say something vaguer than what we mean, is close to universal and is calibrated badly.
Why the softened version is harder for a child to follow
The parent-training world has been direct about directness for a long time. The Parent-Child Interaction Therapy handout on giving commands, used in clinician training, opens by saying the first step in getting compliance is to be direct with your child. Its examples are almost boringly plain: “Please keep your hands to yourself.” “Please hold onto the shopping cart.” “Please sit on the couch.”
The contrast it draws is with two things. One is the vague instruction, like “now behave in the store,” which names a standard rather than an action. The other is the negative one, “stop jumping on the bed,” which tells a child what to quit without telling them what to do instead.
That handout is clinician training material for families working on genuinely difficult behavior, not a rule for every household, and it should be read that way. But the underlying observation matches what we have seen across a lot of afternoons at a lot of kitchen tables.
The question form is the one that catches most of us. “Do you want to put your shoes on?” is a real question, and children answer real questions. So is “shall we tidy up?” if the answer no is available. When we have watched a request go nowhere and then land on the second try, the difference is often not volume or firmness. It is that the second version contained an action and a subject.
We are badly calibrated about how a plain ask lands
The other piece of research here is older and, to us, the more surprising one. In 2008, Francis Flynn and Vanessa Lake published “If you need help, just ask: Underestimating compliance with direct requests for help” in the same journal. They sent people out to ask strangers for small favors, and first asked them to predict how many people they would have to approach before someone agreed.
The predictions were off in the same direction every time. For filling out a questionnaire, participants guessed they would need to ask about 20 people. It took about 10. For borrowing a cell phone to make a call, they guessed 10 and it took 6. For asking someone to walk them to the campus gym, they guessed 7 and it took just over 2. In a study with fundraising volunteers for the Leukemia and Lymphoma Society, volunteers expected to approach roughly 210 people and approached about 122, and they underestimated the average donation as well, guessing about $48 against an actual $64.
People say yes far more often than we expect them to.
Again, the limits are real. These were adults, largely on and around a US campus, asking strangers for small and low-cost favors. Whether the same gap holds when you ask a neighbor for a school-run swap for the third week running is not something this paper can tell you. What is useful is the direction and the size of the error, which showed up across every scenario they tried.
Direct is not the same as blunt
This is the part that goes wrong most often when people take the idea home.
The 2022 feedback research is about information the other person wanted and did not get. It is not a mandate to hand out criticism nobody asked for. A person with marker on their face wants to know. A person who is proud of a thing they made is not asking for a review.
That distinction matters a lot with children and with their work. Being direct about a request is one thing: “I need you to wash your hands before you touch that.” Being direct about what a child has drawn is a different category entirely, and the plainest, most useful sentence there is usually a description rather than an assessment. What you are looking at is a child in the middle of doing something, not a finished piece submitted for comment.
The other common misread is swapping softness for sharpness. A direct request said warmly is still direct. “Please put your shoes on now, we are leaving in five minutes” is not softer for being said kindly, and it does not become clearer for being said louder. Tone and clarity are separate dials, and the research above is only about the clarity one.
Directness also is not evenly distributed across families or cultures, and none of the studies here speak to that. Some households run on hints and understand each other fine. If yours does, this is not a correction.
A smaller version to try this week
Pick one recurring friction point. Shoes, screens off, the recycling, the same conversation with the same person every Thursday. For that one thing, replace the question with a statement that contains an action and a time. Then leave everything else alone.
The lightest possible version, if that feels like too much: keep the sentence you already use and drop the trailing softener. “Can you get your shoes on, okay?” becomes “Can you get your shoes on.” One word.
What tends to happen is not dramatic. Some requests land faster. Some do not, because the child is three, or tired, or in the middle of something that matters to them, and no amount of clear phrasing outranks that. Directness removes ambiguity from a request. It does not remove a two-year-old’s opinion about leaving the park.
The part that surprised us most was not what happened with the children. It was noticing how many adult sentences we had been hedging for no reason at all, and how few of them turned out to need it.