What a scorer is doing
Your response is read for three things, and each one is worth 0, 1, or 2 points. Six is the most anyone can earn. This page gives you the three traits in plain language, then a task to write under the clock, then two real responses to that task — one that earns a 6 and one that earns about a 2 — with notes on exactly what separates them.
Read the traits now. They tell you what the readers are weighing, and they should change how you write.
One thing to settle before the clock starts
Neither article is the “correct” one. You are not being asked which side is true, or which side you agree with. You are being asked which article makes the better-supported argument. So do not spend your time deciding what you believe. Pick the side you can argue more strongly with evidence from the articles, commit to it quickly, and spend the time building the case.
The three traits
Trait 1 — Analysis of the arguments and use of evidence
This is the trait that carries the essay. It asks one question: did you analyze the two arguments, or just react to the topic?
- Earns 2. Takes a clear position on which article is better supported, and explains why by weighing the evidence itself — how strong it is, what kind it is, what it leaves out. Draws specific evidence from both articles.
- Earns 1. States a position and uses some evidence, but the analysis is thin or one-sided — mostly summary, or evidence from only one article.
- Earns 0. Gives a personal opinion on the topic, retells the articles, or never decides which argument is better supported.
Trait 2 — Development of ideas and organization
This trait asks whether your essay is built, not just listed.
- Earns 2. A clear central claim, ideas developed with reasons and detail, a sensible order, and transitions that show how the parts connect.
- Earns 1. Some structure and some development, but uneven — ideas dropped quickly, paragraphs that do not build, weak or missing transitions.
- Earns 0. Little development; no clear plan; the reader cannot follow the argument.
Trait 3 — Clarity and command of standard English
This trait is about control of the sentence, not perfection.
- Earns 2. Generally clear and fluent, varied sentences, a mostly formal register; errors exist but do not get in the way of meaning.
- Earns 1. Readable but uneven; limited sentence variety; frequent errors that occasionally slow the reader; drifts into informal language.
- Earns 0. Errors are frequent and severe enough that meaning is often lost.
Now write one
Below is a full task. Write it under the clock before you read any further. The two responses further down are worth much less to a student who read them first — you will recognize their moves instead of finding your own, and you will not learn where your own writing actually stands.
The task
The two articles below take opposite positions on one question:
Should employers require workers to return to the office?
Read both. Then analyze both positions to determine which one is better supported by evidence. Use specific evidence from both articles in your answer.
This is not a question of which position you personally hold. Your response will be judged on how well you analyze the arguments and use evidence from the articles — not on which side you choose. Allow about 45 minutes.
Article 1
The Office Is Worth Returning To
1When the pandemic forced offices to close, many predicted that the traditional workplace was finished. Years later, the evidence tells a more complicated story — one that should make companies cautious about going fully remote.
2The most careful research comes from Microsoft. In a study of more than 60,000 employees published in Nature Human Behaviour, researcher Longqi Yang and colleagues found that when the company shifted to firm-wide remote work, employees’ collaboration networks became more isolated. People communicated more within their immediate teams but built far fewer connections across different groups. Communication also shifted away from real-time conversation toward email and instant messaging. The researchers warned that these changes could make it harder for organizations to share information and produce new ideas, because innovation often depends on loose connections between people who do not normally work together.
3The cost falls heaviest on younger and newer employees. Much of early-career learning happens informally — overhearing how a senior colleague handles a difficult client, getting a quick correction before a mistake grows, being pulled into a conversation you did not know was happening. These moments are difficult to schedule and nearly impossible to replicate over video calls. New hires who never sit beside experienced coworkers may take far longer to develop, and they often report feeling disconnected from the people they are supposed to be learning from.
4Culture and trust also depend on presence. Shared physical space gives people repeated, low-stakes contact — the brief exchanges that build the familiarity teams rely on when work becomes difficult. Trust built only through scheduled meetings tends to stay narrow and transactional.
5None of this means every worker must be in an office five days a week. But it does mean that the office serves real functions that remote work has not yet matched: mentorship, the spread of ideas across boundaries, and the slow accumulation of trust. Companies that ignore these functions in the name of flexibility may not notice the damage until their pipeline of skilled, well-connected employees has quietly thinned.
Article 2
Flexible Work Is Here to Stay
1The push to force workers back into offices rests on an assumption that the evidence does not support: that being in a building makes people work better. The most rigorous study available suggests otherwise.
2Stanford economist Nicholas Bloom and his colleagues ran a controlled experiment with more than 1,600 employees, randomly assigning some to a hybrid schedule — three days in the office, two at home — and others to full-time office work. The results, published in the journal Nature in 2024, were striking. Hybrid workers were just as productive as their fully in-office counterparts, and they were promoted at the same rate. But the hybrid group quit their jobs about a third less often. Reducing turnover by that much saves companies enormous sums, since replacing a trained employee is costly and slow.
3The benefits to workers are equally concrete. According to U.S. Census Bureau data, the average American spends roughly 27 minutes commuting each way — close to an hour a day, much of it unpaid and stressful. Remote and hybrid arrangements return that time to workers, who report using it for sleep, family, and exercise. These are not small quality-of-life gains; over a year, they add up to weeks of reclaimed time.
4Flexibility also widens who can participate in the workforce at all. Parents managing childcare, people with disabilities for whom a daily commute is a serious barrier, workers caring for aging relatives, and people living far from major job centers are all better able to hold steady employment when work does not require being in one fixed place every day. A rigid return-to-office rule effectively shuts many of these people out, narrowing the talent companies can draw from.
5Critics argue that remote work weakens collaboration. That concern deserves attention — but the answer is hybrid scheduling, which Bloom’s experiment shows preserves in-person contact without sacrificing the gains in retention and access. The realistic choice today is not office versus home. It is whether companies will adopt flexible models supported by strong evidence, or cling to old habits that the data no longer justifies.
Your 45 minutes
Stop here if you have not written yet. What follows is two complete responses to the task above, with the reasoning behind each score. Reading them first will not teach you to write one; it will only teach you to recognize one.
After the clock — two responses to the same task
Both responses below were written to the task you just did. Read each one straight through before reading the notes. Then read your own against them and be honest about where it falls. The point is not to score perfectly on a first attempt; it is to see, sentence by sentence, what the readers reward.
Scores 6 — two points on each trait
A response that earns a 6
Both writers are arguing about the same thing — whether requiring people to be in an office produces better work — but only one of them uses evidence that can settle it. Article 2, “Flexible Work Is Here to Stay,” is the better-supported argument, because the study at its center was built to answer that question and the study in Article 1 was not.
The strength of Article 2 is the design of the experiment it rests on. Bloom and his colleagues did not survey people or watch a company after a change; they randomly assigned more than 1,600 employees to hybrid work or to full-time office work. That matters, because random assignment is what lets a difference in the results be blamed on the schedule rather than on the kind of person who would choose it. The writer then reports three separate outcomes — productivity, promotion rate, and quitting — and two of them fail to help the writer’s own side. Hybrid workers were not more productive, only equal, and they were promoted at the same rate, not a better one. The gain is in turnover alone, about a third fewer people leaving. A writer who reports the results that do not help him is describing evidence rather than selecting it, and that is a reason to trust the third result.
Article 1 is not empty. The Microsoft study is large, more than 60,000 employees, it was published in a scientific journal, and the writer explains a mechanism instead of only asserting a harm: innovation depends on “loose connections between people who do not normally work together,” and remote work thinned those connections. That is a real finding, and Article 2 never denies it.
But the finding does not reach the conclusion drawn from it. The study measured how people communicated — more inside their teams, less across groups, more by message and less by voice. It did not measure whether the company produced less work, or worse work. The writer crosses that gap with a single word: the changes “could make it harder” to share information and produce ideas. Could is not did. The study also has nothing to compare itself against, because the shift was firm-wide and happened at a moment when everything else about working life changed at once. And the two paragraphs that follow carry no evidence at all. New hires “often report feeling disconnected” names no survey, no number, and no source, and the paragraph about trust is entirely assertion.
Article 2 has holes of its own. The commuting figure measures time given back to workers, not work performed, so it sits inside an argument about productivity without belonging to it. The paragraph about parents, disabled workers, and caregivers is reasoning about who gets shut out rather than a measurement of anyone being shut out — persuasive and undocumented at the same time.
What finally separates them is that Article 2 answers its opponent and Article 1 does not. Article 2 names the collaboration objection, grants that it deserves attention, and points to hybrid scheduling as the arrangement that keeps in-person contact while holding on to the retention gain. Article 1 never mentions turnover, which is the one outcome Article 2 actually measured. An argument that leaves its opponent’s strongest number untouched is the weaker of the two, however serious its concerns.
Scores about 2 — the errors below are deliberate
A response that earns about a 2
I think people should be able to work from home. Passage 2 is right because working from home is better for workers and it saves them time. The average American spends 27 minutes commuting each way which is almost an hour a day and thats time you could spend with your family or sleeping. I know this from my own life because when I had a job far away I was tired all the time and I had no time for anything else.
Passage 1 says the office is worth returning to and that people need to be around each other to learn. Maybe that is true for some jobs but not all jobs. Some people work better alone. The writer says new employees feel disconnected but I think you can learn things over video too, we have the technology now.
Also a lot of people cant work in an office, like parents and people with disabilities. So they should be allowed to work at home. Companys should think about their workers and not just about profits. In conclusion flexible work is here to stay like the title says and I agree with it.
What separates them, trait by trait
Trait 1 — analysis and evidence
- The 6 earns 2. It says in its first paragraph which article is better supported and why. It names evidence from both — the random assignment, the three outcomes, the 60,000 employees, the “could make it harder” sentence — and says what each piece proves and fails to prove. It also concedes two weaknesses on its own side.
- The 2 earns 0 or 1. It answers a different question. “Passage 2 is right” is a verdict on the topic, not on the arguing. Article 1 is dismissed with “maybe that is true for some jobs” and personal belief. The one number it uses, 27 minutes, is repeated from the article rather than weighed. Nothing from Article 1 is quoted or examined.
Trait 2 — development and organization
- The 6 earns 2. Six paragraphs, each doing one job, in an order that could not be shuffled: the claim, the winning evidence, the other side stated fairly, where the other side fails, its own side’s weaknesses, then the point that decides it.
- The 2 earns 1. Three paragraphs that could be put in any order without loss. Ideas are raised and dropped in a sentence. The closing line restates the article’s title and adds nothing.
Trait 3 — clarity and command of English
The 6 earns 2. Sentences vary in length and shape. The register stays formal without becoming stiff. There is nothing showy in it.
The 2 earns 1, not 0. The errors are real — thats, cant, Companys, and the comma splice in “you can learn things over video too, we have the technology now” — and the register drifts into speech: I think, I know this from my own life. But the meaning is never lost. A reader can follow every sentence. That is the difference between a 1 and a 0 on this trait, and it is why fixing the apostrophes would not move this essay much.
Read your own
- Find your claim. Point to the sentence that says which article is better supported. If you cannot find it, or it is really a sentence about which side you agree with, that is the first thing to fix.
- Count your sources. How many specific pieces of evidence did you take from the article you did not choose? If the answer is none, Trait 1 caps at 1.
- Find one weakness you named. Did you say what is wrong with any piece of evidence, on either side? Saying only what is right is summary, not analysis.
- Check the shape. Could your paragraphs be reordered without loss? If yes, you have a list rather than an argument.
- Read it aloud. Sentences you stumble over reading are sentences a scorer stumbles over too.