{
  "recordedAt": "2026-09-18T14:47:43.425Z",
  "selection": "Two selected outputs from a 20-output editorial evaluation, not an estimate of typical results. Original synthetic essay; no customer data and no detector score.",
  "promptSourceHash": "b98e2401c7b58c8de7ea5e7baaf4f26fd30a60b7fca722b3b948e48479f0ac01",
  "examples": [
    {
      "mode": "Basic",
      "writingType": "Essay",
      "model": "gpt-5.4-mini-2026-03-17",
      "inputWords": 442,
      "outputWords": 414,
      "sourceHash": "9e414f9935dc95248f501eb8285b7b2128271338de56689706a4ec8c6a2c66e6",
      "source": "A school considering a change to its writing lessons should distinguish time spent away from a draft from time spent avoiding it. In a fictional six-week pilot, two classes would write their first drafts in class. One class would revise on the same afternoon, while the other would return to the work after two days. Both classes would use the same rubric and receive feedback from the same teacher. The proposal does not include test results, and it should not be treated as evidence that either schedule is already better.\n\nThe case for a delay rests on a familiar difficulty in revision. A writer knows what a paragraph was supposed to mean and may read that intended meaning into sentences that do not actually express it. Returning later can make an unexplained link or missing qualification easier to notice. A student might discover that a paragraph announces a reason but provides no example, or that a conclusion makes a stronger claim than the evidence supports. In those situations, distance from the draft may help the writer see a problem that quick proofreading would miss.\n\nThere are limits to this argument. Waiting does not provide knowledge that the writer never had, nor does it replace reading the source material carefully. If a student's account of an event is wrong, a later revision may simply make the incorrect account more fluent. Revision therefore needs questions about the argument as well as questions about wording. What does this claim depend on? Which part of the source supports it? Is an exception important enough to change the conclusion? These questions require active checking, whether revision happens immediately or later.\n\nThe proposed comparison would also be difficult to interpret. Absences could affect how much time students receive in class, and unequal help outside school could influence the second draft. A stronger piece of writing would not automatically show that the delay caused the improvement. The teacher would need to record these limitations and compare the changes made, not just the final appearance of the papers. The pilot could be useful for deciding how to organize lessons without supporting a broad claim about every student or every kind of assignment.\n\nA delayed revision session is worth considering as part of a writing routine, but it should not crowd out planning or source checking. Students still need to decide what they want to argue, gather suitable evidence, and identify what that evidence cannot establish. The practical question is not whether waiting is a cure for weak writing. It is whether the timetable gives students a genuine opportunity to return to an argument and examine what is on the page.",
      "rewrite": "A school weighing a change to its writing lessons may find that the real issue is not revision itself, but the kind of distance revision gives. The source imagines a fictional six-week pilot in which two classes write first drafts in class. One class revises the same afternoon; the other returns after two days. Both use the same rubric and receive feedback from the same teacher. No test results are included, and the proposal should not be treated as evidence that either schedule is already better.\n\nThat delay rests on a familiar problem. Writers often know what a paragraph was supposed to mean and read that intended meaning into sentences that do not actually express it. With time away from the draft, an unexplained link or missing qualification can be easier to spot. A student may see that a paragraph announces a reason but gives no example, or that a conclusion makes a stronger claim than the evidence supports. In those cases, distance may reveal a problem that quick proofreading would miss.\n\nThe argument has limits. Waiting does not create knowledge the writer never had, and it does not replace careful reading of the source material. If a student’s account of an event is wrong, later revision may only make the incorrect account more fluent. Revision therefore needs questions about the argument, not only questions about wording: What does this claim depend on? Which part of the source supports it? Is an exception important enough to change the conclusion? Those checks still have to be active, whether revision happens at once or later.\n\nThat makes the comparison hard to read. Absences could change how much class time students receive, and unequal help outside school could shape the second draft. A stronger paper would not automatically show that the delay caused the improvement. The teacher would need to record those limits and compare the changes made, not just the final appearance of the papers. Even then, the pilot could only help with how to organize lessons; it would not support a broad claim about every student or every kind of assignment.\n\nA delayed revision session can belong in a writing routine, but it should not crowd out planning or source checking. Students still need to decide what they want to argue, gather suitable evidence, and identify what that evidence cannot establish. The practical question is whether the timetable gives them a genuine chance to return to an argument and examine what is on the page."
    },
    {
      "mode": "Stealth",
      "writingType": "Essay",
      "model": "gpt-5.6-terra",
      "inputWords": 442,
      "outputWords": 430,
      "sourceHash": "9e414f9935dc95248f501eb8285b7b2128271338de56689706a4ec8c6a2c66e6",
      "source": "A school considering a change to its writing lessons should distinguish time spent away from a draft from time spent avoiding it. In a fictional six-week pilot, two classes would write their first drafts in class. One class would revise on the same afternoon, while the other would return to the work after two days. Both classes would use the same rubric and receive feedback from the same teacher. The proposal does not include test results, and it should not be treated as evidence that either schedule is already better.\n\nThe case for a delay rests on a familiar difficulty in revision. A writer knows what a paragraph was supposed to mean and may read that intended meaning into sentences that do not actually express it. Returning later can make an unexplained link or missing qualification easier to notice. A student might discover that a paragraph announces a reason but provides no example, or that a conclusion makes a stronger claim than the evidence supports. In those situations, distance from the draft may help the writer see a problem that quick proofreading would miss.\n\nThere are limits to this argument. Waiting does not provide knowledge that the writer never had, nor does it replace reading the source material carefully. If a student's account of an event is wrong, a later revision may simply make the incorrect account more fluent. Revision therefore needs questions about the argument as well as questions about wording. What does this claim depend on? Which part of the source supports it? Is an exception important enough to change the conclusion? These questions require active checking, whether revision happens immediately or later.\n\nThe proposed comparison would also be difficult to interpret. Absences could affect how much time students receive in class, and unequal help outside school could influence the second draft. A stronger piece of writing would not automatically show that the delay caused the improvement. The teacher would need to record these limitations and compare the changes made, not just the final appearance of the papers. The pilot could be useful for deciding how to organize lessons without supporting a broad claim about every student or every kind of assignment.\n\nA delayed revision session is worth considering as part of a writing routine, but it should not crowd out planning or source checking. Students still need to decide what they want to argue, gather suitable evidence, and identify what that evidence cannot establish. The practical question is not whether waiting is a cure for weak writing. It is whether the timetable gives students a genuine opportunity to return to an argument and examine what is on the page.",
      "rewrite": "The most useful measure in a proposed change to writing lessons may not be the polish of the final paper. It may be the changes students actually make after returning to a draft. In a fictional six-week pilot, two classes would write first drafts in class under the same conditions: the same rubric and feedback from the same teacher. One class would revise that afternoon; the other would return after two days. The proposal includes no test results, so it cannot show that either schedule is already better.\n\nA stronger second draft would also require careful interpretation. Absences could alter the amount of class time available to students, while unequal help outside school could affect the later version. Even if one group produced better-looking papers, that would not automatically mean that the delay caused the improvement. The teacher would need to record those limitations and compare revisions themselves rather than judging only the final appearance of the work. Such a pilot could help organize lessons, but it could not support a broad conclusion about every student or every type of assignment.\n\nThe appeal of waiting lies in a specific difficulty. Writers often know what a paragraph was meant to say and can read that intended meaning into sentences that do not convey it. Time away from a draft can make an unexplained connection or an omitted qualification easier to see. A student may notice that a paragraph offers a reason without providing an example, or that a conclusion claims more than the available evidence supports. Distance may reveal problems that quick proofreading misses.\n\nThat benefit depends on distinguishing time away from a draft from time spent avoiding it. Waiting does not create knowledge a writer never possessed, and it does not substitute for careful reading of source material. A student whose account of an event is wrong may return later and merely make the incorrect account sound smoother. Revision must therefore address the argument as well as the wording: What does the claim depend on? Which part of the source supports it? Does an exception matter enough to alter the conclusion? Those questions call for active checking whether revision occurs immediately or after a delay.\n\nA delayed session may have a place in a writing routine, provided it does not displace planning or source checking. Students still need to decide what they want to argue, gather suitable evidence, and identify what that evidence cannot establish. The practical issue is whether the timetable gives them a genuine chance to return to an argument and examine what is actually on the page."
    }
  ]
}