Chapter 13: Lab Reports and Technical Reports: Documenting What You Did and What You Found

"The whole of science is nothing more than a refinement of everyday thinking." — Albert Einstein, Physics and Reality (1936)

Chapter Overview

Dr. Lena Foss runs a small materials-science lab. She is two years into her first faculty job, and last month she handed a draft manuscript back to a new graduate student with more red ink than black. The science was sound—the student had run a careful experiment on a new battery-electrode coating and gotten a real, interesting result. But the report was a mess. The Results section told the reader what the numbers meant before it told them what the numbers were. The Discussion repeated the Results almost verbatim and then claimed the coating "will revolutionize energy storage"—on the strength of nine samples. The Methods section said the films were "annealed at high temperature," which is the kind of phrase that makes a reproducibility-minded reviewer put down their coffee and write a single devastating word in the margin: how? The student had done good work and written it up in a way that hid the work, overstated the finding, and could not be replicated. This is the most common situation in technical writing, and it is entirely fixable.

This chapter opens Part III, where we turn from the general craft of Parts I and II to the specific genres that scientists, engineers, and technical professionals actually produce. We start with the most fundamental one: the report that documents what you did and what you found. In its academic form it is the lab report or research report, organized by a four-part convention called IMRaD—Introduction, Methods, Results, and Discussion. In its workplace form it is the technical report, built from the same bones but reorganized to put the recommendation first, because a busy decision-maker reads differently from a peer reviewer. By the end of this chapter you will be able to take a pile of results and place every sentence in the section where it belongs, write a Methods section a stranger could follow, present Results without editorializing, and write a Discussion that interprets honestly without overclaiming.

The deep idea here is one you already met in Chapter 4 (structure serves the reader, not the writer) and Chapter 5 (writing is how you figure out what you know). IMRaD is not bureaucratic box-checking. It is a structure refined over more than a century to answer, in a fixed and findable order, the four questions every reader of empirical work has: Why did you do this? What exactly did you do? What did you find? What does it mean? When you fight the structure, you are usually fighting one of those questions—most often the discipline of keeping "what you found" separate from "what it means." That separation is the threshold concept of this chapter, and crossing it is what turns a student's write-up into a scientist's report.

In this chapter, you will learn to:

• Identify what belongs in each IMRaD section—Introduction, Methods, Results, Discussion—and place any stray sentence in its correct home.
• Write a Methods section detailed and ordered enough that a competent peer could reproduce your work.
• Present Results as observations, stripping out the interpretation that belongs in the Discussion.
• Write a Discussion that interprets and states limitations without overclaiming or merely restating the Results.
• Convert an academic IMRaD report into a workplace technical report that leads with conclusions and recommendations.

📕 Engineering/Science track: This chapter is core for you, and it is the foundation for the rest of Part III—research papers (Ch 14), literature reviews (Ch 15), theses (Ch 16), grants (Ch 17), and field-specific publication conventions (Ch 35) all build on the IMRaD discipline you learn here. If you read only one section twice, make it §13.5 (Results) and §13.6 (Discussion); the boundary between them is where most reports fail. The technical-report material in §13.8 also serves the Business track, but the academic spine is yours.

13.1 What a Report Is For, and the Four Questions It Answers

Before the structure, the purpose. A lab report or technical report exists to do one job: let a reader who was not in the room understand what you did, trust your findings, and act on or build on them. Everything about the genre follows from that. The reader was not there. They did not see the apparatus, watch the run, or feel the moment the result came in. Your report is the only window they have, and a window that distorts is worse than no window at all.

That reader—peer, reviewer, future lab member, regulator, your own self in two years—arrives with four questions, in this order:

1. Why did you do this? What problem or question motivated the work, and why should I care?
2. What exactly did you do? What materials, methods, and procedures—in enough detail that I could repeat it?
3. What did you find? What were the actual observations and measurements?
4. What does it mean? How do you interpret those findings, and what follows from them?

IMRaD maps one-to-one onto these four questions. Introduction answers why. Methods answers what you did. Results answers what you found. Discussion answers what it means. The power of the convention is that it is fixed: a reader anywhere in the world, in almost any empirical field, knows that the methods live under "Methods" and the interpretation lives under "Discussion." They can navigate your report without being taught how. This is theme 5 (structure serves the reader) crystallized into a global standard. A reviewer with forty papers to read does not want to hunt for your methods; they want to find them exactly where methods always are.

🔍 Why Does This Work? Why has science converged on this particular four-part order—and made it nearly universal across fields that otherwise share little? Think about it before reading on.

Because the order tracks both how the work was logically motivated and how a skeptical reader must evaluate it. You cannot judge whether a finding is real without knowing how it was produced (Methods before Results). You cannot judge an interpretation without first seeing the raw finding it rests on (Results before Discussion). And you will not invest attention in any of it until you know why it mattered (Introduction first). The structure is not arbitrary aesthetics; it is the logical dependency chain of evaluating empirical evidence, frozen into a template. That is also why violating it—interpreting in the Results, say—feels disorienting to a trained reader: it asks them to judge a conclusion before they have the evidence to judge it with.

Here is the most important reframing in the chapter, and it connects straight back to Chapter 1. Writing the report is not a clerical task you do after the science. Drafting the Methods often reveals a step you cannot quite reconstruct—which means you did not record it carefully enough, which is a finding about your own rigor. Drafting the Discussion forces you to state what the result actually supports, and that is where you discover whether you have a real conclusion or just a hope. The report is part of the science, not its packaging. Researchers routinely find a gap in their reasoning only when they try to write it down. If you cannot write the Methods section, you do not yet fully know what you did. If you cannot write the Discussion without overclaiming, you do not yet know what your result means. Writing it is how you find out.

🔄 Check Your Understanding Match each reader question to its IMRaD section: (a) "Could I reproduce this?" (b) "Why should I care about this problem?" (c) "What do these numbers mean?" (d) "What were the actual measurements?"

Answer

(a) Methods — reproducibility is the Methods section's job. (b) Introduction — it motivates the work and states why it matters. (c) Discussion — interpretation of the findings. (d) Results — the actual observations and measurements, reported without interpretation. If you placed (c) and (d) in the same section, that is exactly the conflation this chapter exists to fix.

13.2 IMRaD at a Glance: The Map of the Whole Report

Let's see the whole structure before we walk through it section by section. A full lab or research report, in IMRaD form, typically carries these parts in this order:

Figure 13.1 (described): The IMRaD hourglass. A vertical hourglass shape. The wide top is labeled "INTRODUCTION — broad: the field and the problem, narrowing to your specific question." The hourglass narrows through a thin middle labeled "METHODS and RESULTS — narrow: exactly what you did and exactly what you found, the most specific part of the report." It then widens again toward the bottom, labeled "DISCUSSION — broadening: what your specific finding means for the wider field." A caption reads: "The report starts broad, narrows to the specific work, then broadens back out to implications." Alt-text: an hourglass diagram showing the report moving from a broad introduction, down to narrow and specific methods and results, then back out to a broad discussion of implications—the classic shape of an empirical report.

Notice the shape. The report is an hourglass: it opens broad (the field, the problem), narrows to the most specific part (precisely what you did and precisely what you found), then widens back out (what it means for the field). The Introduction funnels the reader in, from the general problem down to your exact question. The Discussion funnels them back out, from your exact finding up to its broader significance. Methods and Results are the narrow waist—the most concrete, most specific, least interpretive part of the document. We return to this hourglass in Chapter 14, where it governs the research paper; here it is enough to see that IMRaD has a logic of widening and narrowing, not just a list of boxes.

Two practical notes before we go section by section. First, the order you read the sections is not the order you write them. Most experienced researchers draft the Methods first (it is the most mechanical, and you can write it the moment the work is done), then Results, then Discussion, then the Introduction, and the Abstract dead last—because you cannot summarize a paper you have not finished. This is Chapter 5's writing process applied to a genre: plan and draft the easy, concrete parts first to build momentum, save the hard interpretive parts for when you understand your own findings. Second, IMRaD is a family of conventions, not a single rigid template. Some fields merge Results and Discussion; some split Methods into "Materials" and "Procedure"; some add a separate Limitations section. The four questions are universal; the exact headings flex by field. Follow your venue's specific format—but know that underneath any variant, the four questions are doing the work.

[📍 Good stopping point — you now have the whole map. The rest of the chapter walks each section in detail, then converts the academic report into a workplace one.]

13.3 The Introduction: State the Purpose, Don't Bury It

The Introduction answers why. Its job is to move the reader from "I know nothing about this specific problem" to "I understand the question this report answers and why it matters," in as few words as the work allows. A good Introduction is a funnel: it starts wide enough to orient any reader in the field, narrows to the specific gap or question, and ends by stating exactly what this report does about it.

The single most common failure is burying the purpose. Writers, especially students, treat the Introduction as a place to prove they did the reading—so they march through everything ever published on the topic and never quite say what this study is for. The reader finishes three paragraphs of background still asking, "Yes, but what did you do, and why?" Compare:

❌ Before (purpose buried under background): Lithium-ion batteries are an important technology in the modern world. They are used in phones, laptops, electric vehicles, and grid storage. The performance of a lithium-ion battery depends on many factors, including the cathode material, the anode material, the electrolyte, and the separator. Researchers have studied many cathode materials over the years. The cathode is where lithium ions are stored during discharge. Many coatings have been proposed to improve cathode stability. In this report, we investigate a new coating.

Six sentences in, the reader still does not know the question, the gap, or why this coating is worth a report. Everything stated is true and almost everything is irrelevant to this study. Now the funnel done right:

✅ After (funnel: context → gap → this study): Cathode coatings can extend the cycle life of lithium-ion batteries, but most reported coatings degrade above 4.3 V, limiting their use in high-voltage cells. Whether a thin alumina coating remains stable at 4.5 V is not established. This report measures the capacity retention of an alumina-coated NMC cathode cycled at 4.5 V over 200 cycles, and compares it to an uncoated control.

Why it's better: Three sentences do the whole job. Sentence one gives just enough context and names the problem (coatings degrade at high voltage). Sentence two names the specific gap (we don't know if this coating survives 4.5 V). Sentence three states exactly what the report does (measures capacity retention, named coating, named conditions, named comparison). A reader now knows precisely why this report exists and what to expect. The "before" version reviewed the field; the "after" version positioned a contribution within it. (This funnel—context, gap, study—is the seed of the literature-review synthesis you'll develop in Chapter 15.)

A workable Introduction usually moves through four beats, which you can use as a checklist:

1. Context — the broad area and why it matters, in one to three sentences. Enough to orient, not a textbook.
2. The problem or gap — what specifically is unknown, unsolved, or in dispute. This is the hinge of the whole Introduction.
3. Your response — what this report does about that gap (your question, hypothesis, or objective), stated plainly.
4. A forward pointer (optional) — a one-line preview of the approach or the finding, especially in workplace reports.

A note on tense, because it trips people up. In the Introduction, established facts take the present tense ("coatings extend cycle life"—this is true generally, not just in your study), while what you specifically did takes the past tense ("this report measured…"). Mixing them wrongly ("coatings extended cycle life" implies it only happened in your experiment) subtly misstates the science. We will not turn this into a grammar drill—Chapter 6 covered tense—but in scientific writing the tense carries meaning, so it is worth a deliberate check.

✏️ Try This Take any report or paper you have written and find the sentence that states its purpose—the one that tells the reader exactly what the report does. How many words in does it appear? If it is past the first short paragraph, your purpose is buried. Move a version of it up. The reader should know what this document is for within a few sentences.

🔄 Check Your Understanding An Introduction spends four paragraphs summarizing prior research and then ends with "In this report, we examine the problem." What two of the four beats is it missing, and which is the more damaging omission?

Answer

It is heavy on context but missing a sharp statement of the problem/gap (what specifically is unknown) and a specific statement of your response ("examine the problem" is vague—examine it how, measuring what?). The more damaging omission is the gap: without it, the reader has no reason the report needs to exist. Background without a gap is just a literature dump; the gap is what turns background into motivation.

13.4 The Methods: Write So Someone Could Actually Replicate It

The Methods section answers what you did, and it has the clearest single standard in all of technical writing: a competent person in your field should be able to read it and reproduce your work. That is the entire test. Not "did I describe what I did" but "could a stranger, with this text and no access to me, do it again and get the same kind of result?" Reproducibility is the foundation of empirical credibility; a result no one can reproduce is, scientifically, not yet a result. Your Methods section is the load-bearing wall of that credibility.

Most Methods sections fail by being too vague—written by someone who knows the procedure so well they cannot see what they have left out. This is the curse of knowledge from Chapter 2, in its most dangerous form. The phrase "annealed at high temperature" is perfectly clear to the person who ran the furnace and meaningless to everyone else. Which temperature? For how long? In what atmosphere? At what ramp rate? Watch the difference:

❌ Before (too vague to replicate): The films were deposited on the substrates and then annealed at high temperature. The coated cathodes were assembled into cells and tested. Capacity was measured over many cycles.

A reviewer reads this and cannot reproduce a single step. What deposition method? What substrate? How hot, how long, in what gas? What cell format, what electrolyte, what voltage window, what current? "Many cycles" is not a number. Every sentence raises a question it does not answer. Now the version a peer could follow:

✅ After (replicable): Alumina films were deposited on NMC-811 cathode powder by atomic layer deposition (Cambridge NanoTech Savannah S100) using trimethylaluminum and water precursors at 150 °C, for 20 cycles (≈2 nm nominal thickness). Coated powder was annealed at 400 °C for 2 hours in flowing argon (100 sccm) at a ramp rate of 5 °C/min. Cathodes were assembled into CR2032 coin cells against lithium-metal anodes with 1 M LiPF₆ in 1:1 EC:DMC electrolyte and a Celgard 2400 separator. Cells were cycled between 3.0 and 4.5 V at C/3 (0.33 mA/cm²) for 200 cycles at 25 °C using a Neware battery cycler.

Why it's better: Every step now carries the parameters a peer needs to reproduce it: the deposition method and instrument, precursors, temperature, cycle count, and resulting thickness; the anneal temperature, time, atmosphere, flow rate, and ramp rate; the exact cell format, electrolyte, separator; the voltage window, rate, current density, cycle count, and temperature. A stranger with this paragraph could rebuild the experiment. The numbers and named conditions are not pedantry—each one is a variable that changes the result, so each one must be specified. (You do not need to recognize the chemistry; notice the pattern: every vague phrase in the "before" became a specific, measurable parameter in the "after.")

Here are the principles that make a Methods section replicable, distilled:

• Specify every parameter that affects the outcome. Temperatures, times, concentrations, quantities, rates, model numbers of critical equipment, software versions. If changing it would change the result, name it.
• Report in a logical order—usually the order of execution (prepare, then treat, then measure), or grouped by stage (Materials, then Procedure, then Analysis). The reader should be able to follow it as a sequence.
• Name your equipment and materials specifically where the specific item matters (the cycler model, the reagent grade, the cell line, the software and version). Where it genuinely does not matter, you can generalize—but err toward specificity.
• Describe your analysis, not just your data collection. How did you compute capacity retention? What statistical test, with what software? An analysis you cannot reconstruct is as much a reproducibility gap as a procedure you cannot follow.
• Say what you did, not what one should do. Methods is a record of your actual procedure in the past tense ("cells were cycled"), not a recipe or a list of instructions ("cycle the cells"). The instruction-writing genre is Chapter 22; Methods is documentation of a completed act.

A word on the passive voice, because Methods is the one place the usual advice flips. Chapters 3 and 6 told you to prefer the active voice. In Methods, the passive is often the right choice and remains the convention in many fields: "the samples were annealed at 400 °C" is standard and unobjectionable, because who did the annealing is rarely the point—the procedure is. The thing that matters (the sample, the temperature) sits in the subject position; the irrelevant actor (you, a technician) is dropped. That said, the tide is shifting: many journals, especially in biology and medicine, now accept or prefer "we annealed the samples," and a string of clunky passives ("it was determined that the temperature was set such that…") is bad writing in any field. The rule is not "always passive in Methods" but "put the relevant thing in the subject position." Often that thing is the sample, not you. We treat these field-by-field voice conventions in detail in Chapter 35.

🧩 Productive Struggle Here is a Methods sentence from a software-performance study. Before reading on, list every question a reader would need answered to reproduce it: "We benchmarked the new caching layer and found it was much faster."

What a reproducible version needs

This sentence reports a result ("much faster") inside Methods—first problem; that belongs in Results. As a method it answers almost nothing. A replicable version must specify: the hardware (CPU, RAM, OS); the software versions (the application, the cache, dependencies); the workload (what requests, what data volume, read/write mix, concurrency level); the metric (latency? throughput? p50, p95?); how many runs and whether warmed up; the baseline being compared against; and how the measurement was taken (tool, sampling). "Benchmarked … much faster" hides all of it. The pattern is identical to the battery example: every vague claim conceals a parameter a stranger would need.

🔄 Check Your Understanding What is the one-sentence test for whether a Methods section is good enough?

Answer

Could a competent person in your field reproduce your work from this text alone, without access to you? Reproducibility is the entire standard. If a reader would have to ask you a question to repeat a step, that step is underspecified. (A useful proxy: hand it to a labmate who didn't do the work and ask them to find every place they'd be stuck.)

13.5 The Results: Report What Happened, Without Editorializing

The Results section answers what you found. Its discipline is the hardest one in the genre to internalize, and it is this: report the observations; do not interpret them. Results states what happened—the measurements, the trends, the outcomes—in plain, past-tense, declarative sentences. It does not say what the findings mean, why they occurred, whether they are good news, or what should follow. That is the Discussion's job, and keeping the two apart is the threshold concept of this chapter.

Why so strict? Because the reader needs to see your evidence before your interpretation, so they can judge whether your interpretation is warranted. If you blend the two—if every result arrives pre-interpreted—the reader cannot tell the observation from your spin on it, and cannot evaluate your reasoning independently. Separating Results from Discussion is an act of intellectual honesty: it shows your raw findings in the open, then argues about them in a clearly labeled place where the reader knows to be skeptical. This is theme 1 (writing is thinking) wearing lab clothes: forcing yourself to state the finding without the interpretation is how you discover whether the finding actually supports the interpretation you wanted.

The failure mode is editorializing—smuggling interpretation, judgment, or causal claims into what should be a neutral report of observations. Watch it happen:

❌ Before (Results contaminated with interpretation): The coated cells performed excellently, retaining 92% of their capacity after 200 cycles—an impressive result that clearly demonstrates the alumina coating's effectiveness at protecting the cathode from electrolyte attack. The uncoated control, as expected, performed poorly (74% retention), confirming that surface degradation is the dominant failure mode and proving that coating is essential for high-voltage cathodes.

Strip out what does not belong, and notice how much of that passage is interpretation masquerading as observation. "Performed excellently," "impressive," "performed poorly"—those are judgments. "Clearly demonstrates the coating's effectiveness," "confirming that surface degradation is the dominant failure mode," "proving that coating is essential"—those are causal and general claims the raw numbers do not, by themselves, establish. "As expected" sneaks in your hypothesis. What actually happened—the part a reader needs—is two numbers. Here is the same finding as a Results section should report it:

✅ After (Results: observation only): After 200 cycles at 4.5 V, the alumina-coated cells retained 92% ± 2% of their initial capacity (n = 9). The uncoated control cells retained 74% ± 4% (n = 9). The difference in capacity retention between coated and uncoated cells was statistically significant (p < 0.01, two-sample t-test). Coated cells also showed lower impedance growth over cycling (Figure 2).

Why it's better: Every sentence now reports an observation a reader can verify against your data: the retention numbers with uncertainty and sample size, the statistical test, the impedance trend with a figure pointer. There is no "excellent," no "proves," no causal story. The reader sees exactly what you found and can judge for themselves whether it supports the conclusion you will argue in the Discussion. The interpretation ("the coating protects the cathode from electrolyte attack") has not vanished—it has been moved to where it belongs and where the reader knows to scrutinize it. Same finding, but now the observation and the interpretation are not tangled together.

🚪 Threshold Concept: Results report what happened; Discussion reports what it means.

How novices think about it: the Results section is where you tell the reader about your findings—and "telling them about" naturally includes saying whether the findings are good, what caused them, and what they show. The interpretation feels inseparable from the result; splitting them seems artificial, even evasive. So everything pours into one place: number, judgment, cause, and significance all at once.

How experienced researchers think about it: Results and Discussion are two different cognitive acts that must be kept in separate rooms. Results is observation—what the instruments recorded, stated so neutrally that a rival who disagrees with your interpretation would still accept every sentence. Discussion is argument—what you claim those observations mean, where a reader is invited to weigh your reasoning. The boundary is the discipline of the genre: it lets the reader see the evidence uncolored, then evaluate the interpretation as a separate, contestable thing.

Once you cross this threshold, you cannot un-see editorializing. You will catch "excellent," "as expected," "demonstrates," "confirms," and "proves" creeping into Results sections—yours and others'—and you will know to move each one to the Discussion or cut it. You will also write Discussions that actually discuss, because the interpretive work is no longer used up in the Results. The two sections stop competing for the same material.

How do you catch your own editorializing? Three reliable tells. First, adjectives and adverbs of judgment: excellent, poor, surprisingly, impressively, disappointingly, remarkably. A neutral Results section has almost none. Second, causal verbs: because, due to, caused by, demonstrates, proves, confirms, indicates. Cause is interpretation; the Results report correlation and observation, the Discussion argues cause. Third, the phrase "as expected" (or "surprisingly"), which imports your hypothesis into the observation. Run a search for these words in your Results draft; each hit is a candidate to cut or relocate. (This is the editing-hierarchy discipline from Chapter 12, applied with a genre-specific checklist.)

Two clarifications, because the rule is strict but not insane. You may—and should—guide the reader's attention within Results: "Figure 3 shows that capacity retention diverged after cycle 50" is fine, because it points to what the data show, not what it means. The line is between describing the observation (allowed) and explaining or judging it (Discussion). And you do not report everything—Results is curated, presenting the findings that matter, organized logically (often around your figures and tables), not a raw data dump. The supplementary materials hold the raw data; Results holds the meaningful observations, told in an order the reader can follow. Writing well about figures and tables—captions that interpret, prose that highlights the trend—is its own skill, and Chapter 9 covered it; here, just remember that in the Results section the prose points to the data, and saves the meaning for later.

🔄 Check Your Understanding Which of these belong in Results and which in Discussion? (a) "Mean reaction time was 412 ms (SD = 38)." (b) "The slower reaction times suggest the task imposed a higher cognitive load." (c) "Accuracy did not differ between conditions (p = 0.61)." (d) "This null result may be due to a ceiling effect in our easy task."

Answer

(a) Results — a bare observation with a statistic. (b) Discussion — "suggest … higher cognitive load" is interpretation of what the observation means. (c) Results — an observation, even though it is a null result; null findings are still findings and belong in Results. (d) Discussion — "may be due to a ceiling effect" is an interpretive explanation of why the null occurred. The pattern: bare observations (a, c) go in Results; "suggests / may be due to / means" (b, d) go in Discussion.

13.6 The Discussion: Interpret Honestly, Without Overclaiming

The Discussion answers what it means. This is where interpretation finally lives—where you say what your results imply, why they might have come out as they did, how they fit (or clash) with prior work, what their limitations are, and what should happen next. The Discussion is the inverse funnel of the Introduction: it starts narrow (your specific finding) and widens out to significance (what this means for the field). It is the most intellectually demanding section, and it fails in two opposite directions.

The first failure is restating the Results. A weak Discussion simply says again, in slightly different words, what the Results already reported—adding no interpretation at all. The reader, having just read "coated cells retained 92%, uncoated retained 74%," now reads "in this study, the coated cells retained 92% of capacity while the uncoated retained 74%," and learns nothing new. A Discussion that restates is a Discussion that has not done its job. Its job is not to repeat the finding but to make sense of it.

The second, more dangerous failure is overclaiming—stretching the interpretation far beyond what the evidence supports. This is the student's "will revolutionize energy storage" on nine coin cells. Overclaiming takes a real, modest finding and inflates it into a sweeping, unsupported conclusion, and it is the single fastest way to lose a reviewer's trust. The moment a reader catches you claiming more than your data shows, they discount everything you say. Watch both failures, and the fix:

❌ Before (restates, then overclaims): In this study, the alumina-coated cells retained 92% of their capacity after 200 cycles, while the uncoated cells retained only 74%. This shows that alumina coating dramatically improves battery performance. These results prove that alumina coatings are the solution to battery degradation and will enable the next generation of electric vehicles and grid storage. Our coating will revolutionize energy storage.

The first sentence restates the Results verbatim. Then it leaps: from "92% vs. 74% over 200 cycles in coin cells at 4.5 V" to "the solution to battery degradation," "the next generation of electric vehicles," and "revolutionize energy storage." Nine coin cells over 200 cycles cannot bear that weight. The evidence supports a narrow, real claim; the prose makes a vast, unsupported one. Now an honest Discussion:

✅ After (interprets, contextualizes, bounds the claim, states limits): The 18-percentage-point improvement in capacity retention suggests that the alumina coating reduces a degradation pathway active at 4.5 V—most plausibly cathode-electrolyte side reactions, consistent with the lower impedance growth we observed (Figure 2) and with prior reports of alumina coatings suppressing surface reactivity [refs]. The effect size is comparable to that reported for coatings at lower voltages, extending the approach to a higher-voltage regime where most coatings fail.

Several limitations bound these conclusions. The test used coin cells with lithium-metal anodes over 200 cycles; performance in full cells, with graphite anodes, over the 1,000-plus cycles required for commercial use, is untested. The sample size (n = 9 per group) is modest. We measured capacity retention and impedance but did not directly characterize the coating-electrolyte interface, so the proposed mechanism remains inferential. These results justify a follow-up study in full cells over extended cycling, not a claim of commercial readiness.

Why it's better: The "after" does three things the "before" did not. It interprets (proposes a mechanism—reduced side reactions—and ties it to a second observation and prior work, rather than just repeating the numbers). It bounds the claim to exactly what the evidence supports (a real improvement in a specific high-voltage regime, comparable to known effect sizes—not "the solution to battery degradation"). And it states limitations openly (coin cells not full cells, 200 not 1,000 cycles, n = 9, mechanism inferred not measured) and ends with a proportionate next step. A reviewer reading this trusts the author, because the author has shown they know exactly how far their own data reaches. Honesty about limits is not weakness; it is the strongest credibility signal a Discussion can send.

A good Discussion typically does some subset of these moves—you need not do all, but a Discussion that does none of them is just a restatement:

• Interpret the key result: what does it mean, and (carefully) why might it have happened?
• Relate it to prior work: does it confirm, extend, complicate, or contradict what others have found? (This is where the literature-review skill of Chapter 15 pays off.)
• State limitations honestly: sample size, conditions not tested, confounds, measurement gaps, threats to validity. What can your study not conclude?
• Bound the claim: say precisely what the evidence supports and, just as importantly, what it does not.
• Point forward: what is the next experiment, the open question, the proportionate implication?

The discipline that ties these together is calibrating your language to your evidence. Scientific writing has a vocabulary of certainty, and you must match the word to the strength of the support. Prove and demonstrate are strong: reserve them for findings that genuinely earn them (which, in a single study, is rare). Suggest, indicate, is consistent with, may, appears to are appropriately tentative: they say "here is evidence pointing this way" without claiming the case is closed. A result rarely proves a mechanism; it usually is consistent with one. Writing "our data suggest" instead of "our data prove" is not weak hedging—it is accurate reporting of how strong your evidence actually is. (Chapter 7 introduced the linguistics of hedging and certainty; the Discussion is where it matters most, and Chapter 35 returns to the field-specific conventions of statistical claims.) Overclaiming is, at bottom, a calibration failure: using "prove" when your evidence only earns "suggest."

🔍 Why Does This Work? Why does openly stating your study's limitations make a reader trust you more, when it seems like it should make your finding look weaker? Consider before reading on.

Because a reviewer is, by training, hunting for the weaknesses in your work—it is their job to find what your study can't conclude. When you state the limitations yourself, you demonstrate that you already see your work clearly and are not trying to oversell it; you have done the reviewer's skeptical work for them, honestly. That signals competence and integrity, and it makes your actual claims more believable, because you have shown you only claim what the evidence supports. Conversely, a Discussion with no limitations reads as either naïve (you can't see your study's weaknesses) or evasive (you can, and you're hiding them)—and the reviewer, finding the limitations you omitted, now distrusts everything. Stating limits is how you earn the right to be believed on the things you do claim.

🔄 Check Your Understanding A Discussion section ends with: "Our results prove that intermittent fasting is the optimal diet for weight loss." The study was a 6-week trial of 30 participants. Name two distinct problems with that sentence.

Answer

(1) Overclaiming via "prove": a single 6-week, 30-person trial cannot prove anything; the language should be calibrated to "suggests" or "is consistent with." (2) Claim far exceeds evidence ("optimal"): "optimal … for weight loss" claims superiority over all other diets, which the study (presumably comparing fasting to one control, over 6 weeks) did not test—and says nothing about long-term maintenance, the outcome that matters for "optimal." The evidence supports, at most, "intermittent fasting produced more short-term weight loss than [the specific control] over 6 weeks in this sample." A third acceptable answer: no limitations are acknowledged.

13.7 Abstract, Title, and Conclusion: The Parts Around the Core

IMRaD is the spine, but a complete report has three more parts that wrap around it, and each has a distinct job.

The title is the most-read and least-revised part of any report. Far more people read your title than your abstract, and far more read your abstract than your paper—so the title is doing enormous work for its length. Make it informative: it should state what the report is actually about, ideally including the key finding or the specific system studied. "Lab 3" and "A Study of Coatings" tell a reader nothing. "Alumina Coating Improves 4.5 V Capacity Retention of NMC Cathodes" tells a searcher, a reviewer, and your future self exactly what is inside. This is the Chapter 4 lesson about informative headers, applied to the most important header in the document.

The abstract is a 150–250-word miniature of the entire report—and, like the title, it is read far more than the body. Most readers will read only your abstract; for them, the abstract is the paper. So it must stand alone and deliver the whole arc, not describe it. The structured abstract mirrors IMRaD in miniature: one or two sentences of background and purpose (Introduction), one or two on what you did (Methods), the key findings with numbers (Results), and the main conclusion (Discussion). The cardinal sin is the abstract that describes the paper instead of delivering it:

❌ Before (describes, doesn't deliver): This report investigates the effect of an alumina coating on lithium-ion battery cathodes. The methods used are described, and the results are presented and discussed. Conclusions are drawn about the effectiveness of the coating.

That abstract contains zero information. It tells the reader that methods exist, results exist, and conclusions exist—which they could have guessed. It is a table of contents wearing an abstract's clothes (exactly the failure Chapter 4 named for executive summaries). Now an abstract that delivers:

✅ After (delivers the actual content): Cathode coatings can extend battery life but most degrade above 4.3 V. We tested whether a 2-nm alumina coating, applied by atomic layer deposition, stabilizes NMC cathodes at 4.5 V. Coated and uncoated cells were cycled 200 times at 4.5 V in coin cells. Coated cells retained 92% of capacity versus 74% for uncoated controls (p < 0.01), with lower impedance growth. The coating substantially improves high-voltage cycle life in this configuration, though performance in full cells over longer cycling remains to be tested.

Why it's better: A reader who reads only this paragraph knows the gap, the approach, the actual numbers (92% vs. 74%), the significance, and the limitation. They could decide whether to read on, cite the work, or act on it—from the abstract alone. The "before" abstract forced them into the body to learn anything; the "after" respects that the abstract is, for most readers, the whole report. (Write it last, after the paper is done—you cannot summarize what you have not finished.)

The conclusion (where a report has a separate one) is a brief, high-level restatement of the main finding and its significance—the single most important takeaway, in one breath, plus the one most important implication or next step. It does not introduce new results (a result appearing for the first time in the conclusion is a structural error). In shorter reports, the conclusion often folds into the final paragraph of the Discussion; in longer ones, it stands alone. Either way, keep it short and add nothing new.

🔄 Check Your Understanding Why is "the abstract describes the paper" a failure, when describing the paper is literally what it seems an abstract should do?

Answer

Because for most readers the abstract replaces the paper rather than introducing it—they will read nothing else. An abstract that merely announces "methods are described and results are presented" delivers no actual content to that majority; it forces anyone who wants real information into the body. A good abstract delivers the findings (the actual numbers, the actual conclusion) so it stands alone. "Describes" tells the reader a paper exists; "delivers" tells them what the paper found. (Same principle as the stand-alone executive summary, Ch 4.)

📐 Project Checkpoint

Your portfolio's centerpiece is the technical report (piece 1 of 7). In Chapter 3 you drafted it at the sentence level; in Chapter 4 you restructured it top-down with an executive summary and informative headers; in Chapter 5 you fit that work into the plan-draft-revise process. Now you give it the disciplined internal structure of a real report.

This chapter's addition: restructure your report into clean IMRaD-style sections (or their workplace equivalents) and enforce the section boundaries. Concretely, do three things.

1. Sort your content into the four questions. Go through your current draft and label each chunk: is it answering why (Introduction), what I did (Methods/approach), what I found (Results/findings), or what it means (Discussion/recommendations)? Most rough drafts have these tangled—background mixed into findings, interpretation mixed into results. Separate them into the four homes. If your report is workplace-facing rather than a lab study, the sections may be Background / Approach / Findings / Recommendations, but the four questions are the same.

2. Run the editorializing audit on your "Results/Findings." Search that section for judgment words (excellent, poor, surprisingly, clearly) and causal/claim verbs (proves, demonstrates, confirms, because). For each hit, decide: is this a bare observation (keep it in Findings) or an interpretation (move it to Discussion/Recommendations)? This is the single most valuable revision you can make to a report, and it is the one the rest of this chapter exists to teach.

3. Calibrate your claims. In your Discussion/Recommendations, check every strong claim (proves, will, is the best, eliminates) against your actual evidence. Where the evidence only supports a tentative claim, swap to calibrated language (suggests, indicates, is consistent with, may). Add one honest sentence of limitations—what your report or analysis does not establish.

Deliverable: a report now organized by the four questions, with a Results/Findings section scrubbed of interpretation and a Discussion/Recommendations section whose claims match its evidence. Keep your pre-IMRaD version for the Chapter 40 reflection—the difference between "everything tangled together" and "each thing in its place" is exactly the growth this book is tracking.

Next: Chapter 14 turns this report-writing discipline toward the research paper, where the same IMRaD bones support an argument aimed at the scholarly community, and where the abstract you just learned to write becomes the difference between a paper that gets read and one that gets skipped.

13.8 Technical Reports for the Workplace: Same Bones, Recommendations First

Everything so far describes the academic report, written for peers who will evaluate your evidence. The technical report—the kind you write at work, for a manager, a client, a cross-functional team—is built from the same four questions, but it reorganizes them for a completely different reader. Understanding the difference, and when to make the switch, is what separates someone who can only write for a professor from someone who can write for a boss.

The academic reader (a reviewer, an examiner) is paid to scrutinize your evidence, reads the whole thing, and wants the conclusion last, earned by the methods and results that precede it. The workplace reader (a manager, an executive, a client) has fifteen minutes between meetings, wants to make a decision, and needs the conclusion and recommendation first. This is exactly the audience analysis of Chapter 2 and the BLUF principle of Chapter 4: the workplace report inverts IMRaD's order to lead with what the academic report saves for the end.

So the technical report's structure flips the hourglass:

The defining feature is the executive summary: a stand-alone opening (often the only part a decision-maker reads) that states the recommendation, the key findings behind it, and the action required—up front, before any methodology. The methods that anchor an academic paper's credibility get demoted in a workplace report, often to an appendix, because the manager rarely needs to reproduce your analysis; they need to act on it. The detail still exists for the one reader who wants to verify—it just no longer sits between the decision-maker and the decision. Watch the same content reorganized for the two audiences:

❌ Before (academic order sent to a busy executive): This report examines three candidate vendors for the data-warehouse migration. We first established evaluation criteria through stakeholder interviews, then issued an RFP, conducted reference calls, ran two-week proofs of concept, and scored each vendor against eleven weighted criteria spanning cost, performance, security, and support. The methodology and scoring rubric are detailed below. After completing this evaluation, weighing the trade-offs among the candidates… [600 words later] …we therefore recommend Vendor B.

For a peer reviewer, that build-up might be appropriate. For a VP deciding whether to approve a budget, the recommendation is buried 600 words deep behind a methodology she did not ask for and will not read. Now the workplace version:

✅ After (recommendation-first technical report): Executive Summary

We recommend Vendor B for the data-warehouse migration, at $480K over three years. Vendor B scored highest on our weighted evaluation (8.4/10), with the best balance of cost and query performance and no critical security gaps. Vendor A was 15% cheaper but failed two security requirements; Vendor C performed comparably but cost 40% more. We request approval to proceed by March 15 to meet the Q3 cutover.

Findings (summary) - Cost: B is mid-range ($480K); A is lowest ($410K); C is highest ($670K). - Performance: B and C comparable; A 30% slower on our benchmark queries. - Security: A failed two requirements (encryption at rest, audit logging); B and C passed all. - Support: B offers a dedicated engineer; A and C offer standard support.

[Full methodology, scoring rubric, and vendor-by-vendor detail in Appendix A.]

Why it's better: The executive can decide from the first paragraph: the recommendation, the cost, the reason, and the action-with-deadline are all in the opening four sentences. The findings she needs to sanity-check the recommendation are a scannable four-line summary. The methodology—the eleven criteria, the proofs of concept, the scoring—still exists, in the appendix, for the analyst on her team who will dig in. The "before" version made the most important reader do 600 words of archaeology to reach the one sentence she needed. Same evidence, same rigor, opposite usability—because the structure now serves her way of reading, not the writer's order of discovery.

The principles for a workplace technical report:

• Lead with the recommendation and the bottom line (BLUF, Chapter 4). The first paragraph is the executive summary, and it must stand alone.
• Make the findings scannable—a short summary, often a list or a small table, so a reader can verify the recommendation at a glance. (Workplace reports lean harder on visuals and tables; Chapter 9.)
• Demote the methodology. Keep it (credibility still matters; someone will check), but move it below the findings or into an appendix. The manager needs to act, not reproduce.
• State the action required and by when. A workplace report usually exists to drive a decision; end the summary with what you need and the deadline.
• Match length to the reader. A one-page report that gets read beats a twenty-page report that gets skimmed. Many workplace "reports" are best as a one- or two-page memo with detail appended.

Three caveats. First, this is a spectrum, not a binary: a detailed engineering test report sits closer to the academic end (a fellow engineer may need to reproduce it), while a recommendation memo to leadership sits at the far workplace end. Read your specific reader. Second, recommendations-first does not mean evidence-free: the recommendation leads, but the findings and (appended) methods must back it up, or the report is just an opinion. Third, many organizations have a house template—follow it, then apply these principles inside it, exactly as Chapter 4 advised for IMRaD. We develop the workplace genres further in Part IV (emails, proposals, status and incident reports), where this recommendation-first instinct becomes second nature.

🔄 Check Your Understanding You wrote a careful 12-page analysis. Your director asks for "the report." Do you send the 12 pages with the recommendation in the conclusion, or restructure? What is the minimum change that serves this reader?

Answer

Restructure—at minimum, add a one-paragraph executive summary at the very top that states the recommendation, the key reasons, and the action needed, and move the detailed methodology toward the back or into an appendix. You don't have to discard the 12 pages of analysis (someone may need them), but the director should not have to read to page 12 to find the recommendation. The minimum viable change is a stand-alone executive summary up front; the fuller change is flipping the whole structure to findings/recommendations-first with methods demoted. This is BLUF (Ch 4) applied to a report, driven by audience (Ch 2).

13.9 An Annotated Excellent Lab Report

Theory assembles best around an example. Below is a short, complete lab report in IMRaD form—deliberately compact, but structurally sound—with margin-style annotations naming what each part does right. (The science is a simple, real teaching experiment: measuring how the period of a pendulum depends on its length. The point is the structure, not the physics.)

TITLE
Effect of String Length on the Period of a Simple Pendulum

Informative title: it names the variables (string length, period) and the system (simple pendulum), not "Lab 4." A reader knows the content from the title alone (§13.7).

ABSTRACT
The period of a simple pendulum is predicted to depend on the square
root of its length. We measured the period of a pendulum at six string
lengths from 0.20 to 1.20 m, timing 20 oscillations at each length and
repeating three times. Period increased with length, and a plot of
period squared against length was linear (R² = 0.998), with a slope
implying g = 9.79 ± 0.06 m/s², consistent with the accepted value.
The simple-pendulum model holds well for small amplitudes at these
lengths.

A stand-alone miniature of the whole report: purpose, method (with the key numbers—six lengths, 20 oscillations, three trials), the actual result (linear, R² = 0.998, g = 9.79), and the conclusion. A reader who reads only this knows what was found (§13.7).

INTRODUCTION
The period T of a simple pendulum is predicted by theory to depend on
its length L and the acceleration due to gravity g, with longer
pendulums swinging more slowly. This relationship is widely used to
introduce harmonic motion, but student measurements often deviate at
large swing amplitudes, where the simple model breaks down. This
experiment tests how closely a real pendulum follows the predicted
length dependence across a range of lengths, and uses the result to
estimate g.

The funnel (§13.3): context (period depends on length), the wrinkle/gap (the model deviates at large amplitude), and exactly what this experiment does (test the length dependence, estimate g). Purpose stated plainly, not buried.

METHODS
A pendulum was constructed from a steel bob (52 g) suspended by
inextensible string from a fixed clamp. String length was measured from
the pivot to the center of the bob with a meter rule (±1 mm). At each of
six lengths (0.20, 0.40, 0.60, 0.80, 1.00, 1.20 m), the bob was released
from a small amplitude (<10°) and the time for 20 complete oscillations
was recorded with a stopwatch (±0.01 s). Each measurement was repeated
three times and averaged. The period was obtained by dividing by 20.
Period squared was plotted against length, and g was calculated from the
slope.

Replicable (§13.4): every parameter that affects the result is specified—bob mass, how length was measured and its precision, the six lengths, the small-amplitude condition (which matters, per the Introduction), 20 oscillations, three repeats, the analysis method. A peer could reproduce this. Past tense, passive where the actor is irrelevant ("a pendulum was constructed").

RESULTS
The measured period increased with string length, from 0.90 s at 0.20 m
to 2.20 s at 1.20 m (Table 1). A plot of period squared (T²) against
length (L) was linear (Figure 1), with a best-fit line of slope
4.03 s²/m (R² = 0.998) and an intercept not significantly different from
zero. The acceleration due to gravity calculated from the slope was
9.79 ± 0.06 m/s².

Observation only (§13.5): what happened—the periods, the linear plot, the slope, the R², the value of g—with no judgment ("excellent fit") and no interpretation ("this proves the theory"). Points the reader to Table 1 and Figure 1 without explaining what they mean. Bare findings, including the uncertainty.

DISCUSSION
The linear relationship between T² and L (R² = 0.998) is consistent with
the predicted square-root dependence of period on length, supporting the
simple-pendulum model over this range of lengths at small amplitude. The
value of g obtained, 9.79 ± 0.06 m/s², agrees with the accepted local
value (9.81 m/s²) within uncertainty, suggesting the method is sound. The
near-zero intercept indicates no large systematic offset in the length or
timing measurement.

The main limitations are the use of a stopwatch (human reaction time
adds roughly 0.1–0.2 s of timing error per reading, partly mitigated by
timing 20 oscillations) and the restriction to small amplitudes; the
model is expected to deviate at larger swing angles, which this
experiment did not test. A photogate timer and a deliberate test of the
large-amplitude regime would extend and sharpen these results.

Interpretation, calibrated (§13.6): says what the result means (consistent with the model, g agrees within uncertainty) using tentative, evidence-matched verbs ("is consistent with," "suggesting," "indicates"—not "proves"). Relates to the accepted value. States limitations honestly (stopwatch error, small-amplitude restriction) and proposes a proportionate next step. It does not restate the Results; it makes sense of them.

CONCLUSION
A simple pendulum's period followed the predicted length dependence
across 0.20–1.20 m at small amplitude, yielding g = 9.79 ± 0.06 m/s²,
consistent with the accepted value. The simple model is adequate in this
regime; its behavior at large amplitudes remains to be tested.

One-breath takeaway and the one open question. No new results. Closes the hourglass by stating the significance compactly (§13.7).

Read top to bottom, the report is the four questions answered in order: why (Introduction), what was done (Methods), what was found (Results), what it means (Discussion). Each section stays in its lane—the Results report numbers, the Discussion interprets them—and the claims never outrun the single afternoon's data. It is not long, and it does not need to be. Length is not the standard; the four questions, each answered in its place, with claims calibrated to the evidence is the standard. A report that does that in two pages beats one that does it badly in twenty.

🔄 Check Your Understanding In the annotated report, the Discussion says g "agrees with the accepted value within uncertainty, suggesting the method is sound." Why "suggesting" rather than "proving"?

Answer

Because one experiment, with stopwatch timing and a handful of lengths, provides evidence consistent with a sound method, but does not eliminate every alternative explanation (a lucky cancellation of errors, an untested amplitude regime, a single apparatus). "Suggesting" calibrates the claim to that strength of evidence; "proving" would overclaim, asserting certainty the single measurement cannot deliver. Matching the verb to the evidence is the core Discussion discipline (§13.6).

13.10 Common Mistakes and Practical Considerations

Mistake 1: Editorializing in the Results. The signature error of the genre—smuggling judgment ("excellent," "as expected") and interpretation ("this proves," "because") into what should be neutral observation. Fix: Results report what happened; the meaning goes in the Discussion. Search your Results draft for judgment adjectives and causal verbs and relocate each one (§13.5).

Mistake 2: A Discussion that just restates the Results. Repeating "we found X" in slightly different words, adding no interpretation. Fix: the Discussion's job is to make sense of the finding—interpret it, relate it to prior work, bound it, state its limits. If a sentence in your Discussion only repeats a number from the Results, it is not pulling its weight (§13.6).

Mistake 3: Overclaiming. Inflating a modest finding into a sweeping conclusion ("revolutionize," "proves," "the solution to"). The fastest way to lose a reader's trust. Fix: calibrate language to evidence (suggests, not proves), bound the claim to exactly what the data support, and state limitations openly (§13.6).

Mistake 4: A Methods section no one could replicate. Vague phrases ("at high temperature," "for a while," "the standard procedure") that hide the parameters a stranger would need. The curse of knowledge in its most damaging form. Fix: specify every parameter that affects the outcome; hand the section to someone who didn't do the work and find every place they'd be stuck (§13.4).

Mistake 5: An Introduction that buries the purpose. Paragraphs of background with no clear statement of the gap or what this study does. Fix: use the funnel—context, gap, your response—and make sure the purpose appears within the first short paragraph (§13.3).

Mistake 6: An abstract that describes instead of delivers. "Results are presented and discussed" tells the reader nothing. Fix: put the actual findings, with numbers, in the abstract; it must stand alone, because for most readers it is the paper (§13.7).

Mistake 7: Sending an academic-ordered report to a workplace reader. Burying the recommendation at the end of a methodology-first report sent to a busy decision-maker. Fix: for workplace audiences, lead with an executive summary and recommendation; demote the methods to an appendix (§13.8).

It depends — fields and formats vary. IMRaD is a family, not a single template. Some fields merge Results and Discussion into one section (common where interpretation is tightly bound to each result); some require a separate Limitations or Conclusions section; some split Methods into Materials and Procedure; engineering test reports often add a Recommendations section even in otherwise-academic formats. The four questions are near-universal; the exact headings are local. Always follow your specific venue's author guidelines or your organization's template—then apply this chapter's principles within that structure. When in doubt, find two or three well-regarded reports in your exact target venue and study how they handle each section; the conventions are learned best by reading good examples in the genre you're writing for.

A note for multilingual writers. The Methods section is, in one respect, the friendliest part of technical writing for non-native English writers: it is formulaic, repetitive, and forgiving. Past tense, often passive, short declarative sentences naming parameters—"the samples were heated to 400 °C for two hours"—is a pattern you can master and reuse. You do not need elegant prose in Methods; you need complete and unambiguous prose, and that is a matter of including every parameter, not of native fluency. The same is true of the structured abstract: it follows a fixed four-move pattern (purpose, method, result, conclusion) you can template. Structure and completeness, which are learnable, matter far more here than idiom, which is hard—so invest your effort where it pays.

Frequently Asked Questions

What does IMRaD stand for, and is it required?

IMRaD stands for Introduction, Methods, Results, and Discussion—the four core sections of an empirical report, answering why / what I did / what I found / what it means. It is the dominant structure for scientific papers and lab reports across most fields, but it is a strong convention, not a universal law: some fields merge Results and Discussion, add a Conclusions or Limitations section, or rename sections. Follow your specific venue's format, but expect the four underlying questions almost everywhere.

What is the difference between the Results and the Discussion sections?

Results report what you found; Discussion reports what it means. Results is observation—the measurements, trends, and outcomes, stated neutrally without interpretation or judgment ("coated cells retained 92%"). Discussion is interpretation—what those findings imply, why they might have occurred, how they relate to prior work, their limitations ("the 92% retention suggests the coating reduces side reactions, though…"). Keeping them separate lets the reader evaluate your evidence before your interpretation. The fastest self-check: if a Results sentence contains "proves," "because," "excellent," or "as expected," it has drifted into Discussion territory.

How detailed does a Methods section need to be?

Detailed enough that a competent person in your field could reproduce your work from the text alone, without asking you anything. Specify every parameter that affects the outcome—temperatures, times, quantities, concentrations, equipment models, software versions, and your analysis method. The reliable test: hand the section to a colleague who didn't do the work and have them mark every place they'd be unable to proceed. Each mark is an underspecified step. Vague phrases like "at high temperature" or "the standard procedure" are the usual culprits.

How do I write the Discussion without overclaiming?

Calibrate your language to your evidence and state your limitations. Use tentative, accurate verbs—suggests, indicates, is consistent with, may, appears to—rather than proves or demonstrates, which a single study rarely earns. Bound your claim to exactly what the data support (a finding about coin cells over 200 cycles is not a claim about commercial batteries). And explicitly name what your study could not establish—sample size, untested conditions, confounds. Counterintuitively, stating limitations makes readers trust your other claims more, because it shows you see your own work clearly.

What is the difference between a lab report and a technical report?

They share the same four questions but order them for different readers. A lab/research report uses IMRaD and saves the conclusion for the end, because its reader (a peer or examiner) evaluates the evidence first. A workplace technical report inverts this: it leads with an executive summary and recommendation and demotes the methodology to an appendix, because its reader (a busy decision-maker) needs to act, not reproduce. Same evidence, opposite structure—driven by audience (see Chapter 2 on audience) and BLUF (see Chapter 4 on structure).

Chapter Summary

Key Takeaways

• IMRaD answers four questions in a fixed order: Introduction (why), Methods (what I did), Results (what I found), Discussion (what it means). The convention is global because the order tracks how a skeptical reader must evaluate evidence; it serves the reader, not the writer.
• The Introduction is a funnel: context → gap → what this study does. State the purpose within the first short paragraph; don't bury it under a literature dump.
• Methods must be replicable. The standard is that a competent peer could reproduce your work from the text alone. Specify every parameter that affects the result; "at high temperature" fails the test.
• Results report; they do not interpret. State observations neutrally, without judgment words or causal claims. This is the threshold concept: Results report what happened, Discussion reports what it means.
• The Discussion interprets without overclaiming. Make sense of the finding (don't just restate it), calibrate language to evidence (suggests, not proves), and state limitations honestly—doing so makes readers trust you more.
• Workplace technical reports invert the order: lead with the recommendation and an executive summary; demote the methods. Same bones, different reader.

Action Items

1. Sort every chunk of your report into one of the four questions; separate tangled content into its proper section.
2. Audit your Results for judgment words (excellent, surprisingly) and causal verbs (proves, because); relocate each to the Discussion.
3. Hand your Methods to someone who didn't do the work; specify every place they'd be stuck.
4. Calibrate every strong claim in your Discussion to the evidence; add one honest sentence of limitations.
5. For a workplace audience, add a stand-alone executive summary up front and move methods to an appendix.

Common Mistakes

Editorializing in Results · a Discussion that restates instead of interprets · overclaiming beyond the evidence · a Methods section no one could replicate · an Introduction that buries the purpose · an abstract that describes instead of delivers · sending an academic-ordered report to a workplace reader.

Decision Framework: where does this sentence go, and how should I say it?

Spaced Review

A few questions reaching back, to strengthen retention.

1. (From Chapter 12 — editing and revision) The editing hierarchy says to fix content and structure before polishing sentences. Where in that hierarchy does "move this editorializing sentence from Results to Discussion" belong—is it proofreading, sentence editing, or structural revision? Why does doing it early save you work?
2. (From Chapter 5 — the writing process) This chapter recommends drafting a report's sections out of order—Methods first, Abstract last. How does that advice apply the five-stage writing process, and which stage does "writing the Methods reveals a step you can't reconstruct" actually belong to?
3. (Bridging — Ch 4 + this chapter) The academic report saves the conclusion for the Discussion at the end, while Chapter 4 told you to put the bottom line first (BLUF). Are these in conflict? Reconcile them—when does each apply, and how does the workplace technical report (§13.8) resolve the tension?

What's Next

You can now write a report that documents what you did and what you found—each finding in its place, claims calibrated to evidence, methods a stranger could follow. Chapter 14: Research Papers takes the next step: the same IMRaD bones, but now in service of an argument aimed at the scholarly community. A research paper is not a neutral report of an experiment; it is a claim to a contribution, positioned within a conversation of prior work, and it lives or dies on the abstract you just learned to write and the framing of its introduction. You'll meet Dr. Lena Foss again—this time turning a sound report into a paper that gets read, cited, and (after the gauntlet of peer review) published. The report is where you document the work; the paper is where you argue for its place in the field.

Practice: Exercises · Quiz Go deeper: Case Study · Case Study 2 Review: Key Takeaways · Further Reading