How to write a lot: a practical guide to productive academic writing - Silvia Paul J. 2019

Writing Journal Articles

Scholarly journals are like the mean jocks and aloof rich girls in every 1980s high-school movie: They reject all but the beautiful and persistent. A manuscript must be appealing and shiny to catch the eye of the cool crowd that picks which papers get published. And it must be determined and resilient to squeeze its way past the many dozens of other suitors competing for the editor’s affections. So how do we write manuscripts that are both pretty and gritty?

After a couple decades in the muddy trenches of scholarly journals—as an author, reviewer, and editor—I’ve learned that the process of peer review is largely reasonable, predictable, rational, and fair—not entirely, and not overwhelmingly, but largely. And if we understand how any reasonably predictable system works, we can nudge it to do what we want.

This chapter talks about beauty and persistence, the two things writers can control to have better odds at the journals. For the first, beauty, we’ll describe how to write better papers. As shocking as it sounds, better papers have better odds. For the second, persistence, we’ll describe how to interact with journals, which strike beginners as cold and fickle. We’ll focus on IMRAD articles—the Introduction, Method, Results, and Discussion structure common in the social and health sciences—but the general advice applies to everyone on the outside looking in.

PICK YOUR AUDIENCE FIRST

Obviously, we should write our articles with the assumption that people will read them. But who? Nearly all scholarly work can be crafted to appeal to different audiences, so you need to pick one before writing. Our articles aren’t so different from rock songs. A good song can be arranged to make the hipsters bob their heads in a small basement club, rearranged to appeal to selfie-snapping teens in a huge arena, and rearranged yet again when the lead singer embarks on an emotionally overwrought acoustic coffeehouse tour after getting out of rehab.

Articles have themes and hooks and arcs that can be arranged to appeal to different crowds. Instead of “writing an article,” we should always “write an article for . . .” There’s no one-size-fits-all structure, no plain-vanilla format, for a scholarly field’s articles. We might teach a plain-vanilla format to undergraduates in research methods classes, but what works in rehearsal doesn’t always make the crowd dance. For example, perhaps you’ve just wrapped up a study of how adults with depression talk to their children about goals and values. This work could go to journals in clinical psychology, developmental psychology, family studies, motivation and emotion, or social psychology. Each area would expect to see something different—in length, tone, emphasis, details, and references—so you should pick your audience first.

Choosing your target journal is thus the first step in writing. If you know who you’re writing for, you can craft the paper to appeal to them. Most of your vexing writing decisions can be solved by using that journal’s published articles as models. How long should the Introduction be? Should you make a table of your sample’s demographic statistics? Do you have to have sections devoted to limitations or future directions? Should you post online data and supplemental materials? Should the tone be detached and stuffy or personal and earthy? Do what the articles in that journal typically do.

WRITING AN IMRAD EMPIRICAL ARTICLE

Writing a journal article is like writing a screenplay for a romantic comedy—you need to learn a formula. But instead of meet-cutes and quirky best friends, we have the IMRAD: Introduction, Method, Results, and Discussion.

Outlining and Planning

On my list of maladaptive practices that make writing harder, Not Outlining is pretty high—just above Typing With Scratchy Wool Mittens, just below Training My Dog to Take Dictation. Outlining is writing, not a prelude to “real writing.” Like hypochondriacs, writers who don’t outline are convinced that they’re afflicted with a mystifying illness—the fake malady of writer’s block, in this case (see Chapter 3). After trying to write blindly, they feel frustrated and complain about how hard it is to generate words. “Clear thinking becomes clear writing,” said Zinsser (2006, p. 8). We’re not doing improv, so let’s collect our thoughts before stepping onto the stage.

Your article should always fit within the typical length of published articles in your target journal, but it’s better to be on the brief end. Short is good. When you read journal articles, how often do you wish that the authors would keep their momentum going for another eight pages? Some authors are like self-indulgent jam bands who keep riffing on the same themes—yet another secondary finding, another future direction, another arcane implication. Just enough is usually just right.

The Title and Abstract

Most readers who come across your article will see only the title and abstract, so make them count. A good title balances generality and specificity—say what your article is about, but don’t be so specific that your article sounds technical and tedious. If tempted to write a trendy, topical, or comical title, think about how it will sound in 10 years. Will future researchers get the joke? Readers in our digital age find our articles with electronic databases, so your abstract should be stuffed with the keywords that you would want found in a database search, even if the abstract’s style suffers.

Introduction

Writers fear the Introduction, the hardest section to write, and for good reasons. This section gets the most scrutiny from readers and reviewers, so the article’s success hinges on how well we can make our case. For most articles, the time-honored books-and-bookends formula (Silvia, 2015, p. 97) will make your Introduction sleek and compelling. This formula divides the Introduction into brief opening and closing sections that flank longer chunks of ideas, much like small bookends holding books upright.

§ Your Introduction starts with a brief overview of your work, usually one or two paragraphs long, that sets the stage. This first bookend, often called the “Pre-Introduction” or the “Intro to the Intro,” starts with the global issues that animate your work and then funnels into a snapshot of your project. This section usually ends with a sentence that hits the paper’s primary purpose, such as a sentence starting with, “In the present research, we examined. . . .” A good Pre-Introduction gives a snapshot of the paper’s animating problem.

§ After your first bookend, you have your books. When outlining, think of this section as a set of chunks—usually two, three, or four main pieces. Ultimately, most Introductions have only two or three parts, each marked by a heading. Once you figure out those parts, writing the Introduction is simple. Your first section, for example, might summarize the state of knowledge about your problem; your second section might introduce a complicating issue (why does it happen? how does it work?); and your third section might describe how the complicating issue can be addressed. Likewise, your first section might describe one theory, your second might describe a competing one, and your third might explain how to evaluate which one is better. The details, of course, will vary, but cleaving your ideas into two to four chunks is a good outlining tactic.

§ Finally, you conclude your Introduction with the other bookend, which starts with a heading called something like The Present Research. So far, you have given an overview of your problem (your Pre-Introduction) and developed your reasoning for the research (your two to four books). By now, the reader understands your study’s context and significance. This final section gives a snapshot of your research and explains how it answers your guiding question—it might take one to four paragraphs, depending on the level of detail. Conclude this section with the heading that begins your Method section (Method or Study 1).

The books-and-bookends formula is crisp: It introduces the reader to your problem, lays out the theories and research relevant to the problem, and illustrates how your research will solve the problem. It leads both the reader and the writer on a straight path and discourages straying from the main points. You’ll find exceptions to this formula—for short reports, a single section with no headings suffices—but the books-and-bookends template works well for most articles.

Method

Method sections are easy, the low-hanging fruit that weak-willed writers pick first when starting a new manuscript. Writing them is even easier if you choose your journal first, because you can use their published articles as models. Method sections aren’t glamorous, but they reveal how carefully you conducted your research (Reis, 2000). Like introductions, they follow a formula made up of subsections. The first, Participants or Participants and Design, describes the size and characteristics of the sample and, for experiments, the experimental design. If your study hinged on equipment, you’ll need a subsection called Apparatus. A Measures subsection is helpful when your research used assessment tools, such as neurocognitive tests, interest inventories, and self-report measures of attitudes.

After these subsections, you have the Procedure subsection, the heart of your Method. In this section, describe what you did and said. Reviewers pay close attention to the procedure subsection, and you don’t want to look like you’re hiding something. Provide a lot of detail about your independent variables and dependent variables. Your rhetorical goal is to connect your procedures with the procedures used in published articles. If your experiment used a manipulation that has been used before, cite representative past experiments, even if the manipulation is well-known. If you invented the manipulation, cite research that used similar manipulations or research that implies that your manipulation is reasonable. Connecting your procedures to past research allays concerns about the validity of what you did.

Reviewers want to know how you measured your dependent variables. If your dependent variables are well established, cite articles that developed or used the scales. For professional tests, cite the test manuals as well as recent articles that used the tests. For self-report scales, list the scale values—for example, 7-point scales can be 1 to 7, 0 to 6, or −3 to +3—along with any labels that anchored the scale (e.g., 1 = not at all, 7 = extremely). If your dependent measures were physiological or behavioral, briefly describe past research that supports the validity of your measure.

Results

The Results section is where you display and describe your findings. Your statistics should be like paintings in a museum hallway that readers walk through—each major piece gets plenty of space and some handy interpretive text. Order, emphasis, and selectivity are important.

A good Results section will use headings to segregate the dense clusters of findings. Your first heading marks a subsection for the ugly bits (Salovey, 2000). This part, usually called something like “Data Reduction and Analysis Plan,” contains essential but tedious analytic facts. They need to be in there somewhere, but concentrating them in one place makes the Results sleeker. This section, for example, is where you would describe how you formed composite scores, evaluated reliability, tested statistical assumptions, and crunched the numbers. Such arcana are best shoveled into this first section, which is the subtle maintenance closet at the start of the hallway.

Next, you’ll need a heading for the main findings. The most central findings come first—your primary hypotheses, your primary measures—and decline from there (Salovey, 2000). It’s key to know when to stop. Researchers can be easily enchanted by peripheral findings, but we can’t describe everything. Just as all museums have heaps of paintings in the basement vault that they could have displayed, you probably have heaps of peripheral findings that you find fascinating but would clutter the walls. Readers get confused, and your Results section loses steam as readers slog through the morass of marginalia. Peripheral findings can be stuffed into footnotes or, better yet, online supplemental information that can be accessed by the few readers who share your enchantment.

When describing your findings, use a remind-describe-explain format. At the start of a segment, remind readers of your hypothesis, describe the outcome of the analysis, and then briefly explain what it means. Your Discussion will have the extended recaps and interpretations; here, you have the brief, tasteful sign next to the painting. For example, here’s a blunt, context-free segment:

According to a t-test, the mean depression scores were significantly different between the treatment and wait-list control groups, t(78) = 3.32, p < .001.

Readers would rather read this, however:

Did the intervention affect levels of depression? A t-test on BDI scores (see Table 2) revealed that the intervention did significantly affect depression scores, t(78) = 3.32, p < .001. As our model predicted, people who received short-term social skills training were less depressed than people in the wait-list control condition.

A museum-quality Results section achieves a sleek and spare look by moving as much information as possible to tables and figures. If we’re describing an experiment with four cells, for example, no one wants to read dense paragraphs stuffed with means, standard errors, and confidence intervals for each outcome for each condition. But if we make an elegant figure with the means and error bars, a comprehensive table of descriptive statistics, or both of them, our readers are doubly happy. They can see both an easy-to-read paragraph as well as tables and figures that convey much more statistical detail than the paragraph could possibly hold. Your Results section will obviously always contain numbers, but aim for discourse over digits whenever possible.

Discussion

The Discussion steps back and puts your findings in context. The typical one is a drama in three parts.

§ Start with a recap, a one- to three-paragraph overview. After sifting through the piles of nuts and bolts in your Method and Results, your readers want to be reminded of what you’re trying to build. Your recap thus goes back to the big issues in the Introduction. Begin with the conceptual issues, funnel toward the hypotheses, and then describe how your results inform the big questions that animate your article. A good recap resembles a long abstract of the Introduction and Results.

§ After the recap, build connections with past work by discussing a couple ways in which your work matters. Perhaps your findings have implications for how your readers should think about past theories, methods, practices, or articles. We all see our work as rich with implications, but pick the most interesting and important ones. If you discuss more than three, you’ll start to sound like a self-indulgent jam band.

§ The third section wraps up any remaining issues. Any quirks or problems in the Results to confront? Any directions for future research? Any practical applications? Any limitations? Most of the topics in this section are optional at some journals but required at others. In some fields, you must always have a section on limitations and a section on future directions; in other fields, some papers have them but others don’t. You should consult recent articles in your target journal. If most of the articles don’t include these sections, you can omit them for the sake of a sleek Discussion. If you like, you can end with a short concluding paragraph, but it’s optional.

About that limitations section: Your undergraduate research-methods instructor told you to end your Discussion with a section on limitations; your thesis committee probably wanted this section, too. Describing limitations is a useful educational exercise, but it’s often pointless in an article intended for a professional journal. Most of what pass for limitations are merely directions for future research. Yes, it would have been nice to have a larger sample with a broader range of ages, regions, and cultures; yes, it would have been nice to have even fancier methods and more time points; yes, it’s conceivable that a different study that uses different measures with a different sample would find something different. Other limitations are so generic to an area of research that it’s irksome to reread the same paragraph of ritual self-flagellation in every article. As always, do what the articles in your target journal do—but if they don’t indulge in the ritualistic self-abasement of the generic limitations paragraph, omit it.

References

Although not as glamorous as an Introduction or as brawny as the Results, your References section deserves to be done well. Your references say a lot about how you view your work. Apart from documenting the sources that influenced your ideas, they position you within a field of scholarship. If you’re sending your paper to a family studies journal, for example, but rarely cite articles from those journals, you’ll look like an interloper who wants an audience without taking the trouble to connect to it.

SUBMITTING YOUR MANUSCRIPT

When is your fledgling manuscript ready to leave the nest? The frazzled and flaky submit their papers too early, thinking, “I’ll just send it off now and clean it up later when I resubmit it.” The perfectionists, on the other hand, can’t bear to stop tweaking and sanding and polishing as they fret about the imagined scowls of reviewers who noticed an errant their/there typo. In this case, the perfectionists are probably right. Editors are more likely to invite a revision when the first draft is tight because the author seems like someone who would resubmit a revision without much drama.

Before submitting your pristine manuscript, don’t forget to read the journal’s instructions to authors. Submission guidelines vary between journals and change over the years, so they are always worth double checking. These days, virtually all journals manage submissions via an online system—everything else seems suspiciously retro. If a journal wants you to mail a hard copy of your paper to the editorial office, be sure to ask where the Pony Express courier should deliver it.

Regardless of how you submit your manuscript, you’ll need to write a cover letter to the editor. Most people dash off a boilerplate letter with the standard disclosures and statements; a few write a treatise that summarizes the manuscript and extols its many merits and charms. If your paper’s merits and charms aren’t self-evident, a long and labored cover letter won’t help you—keep it short and simple. Some journals invite you to list a few possible reviewers and to note anyone who shouldn’t review it. Don’t be shy about suggesting a few people who could give your paper an informed reading, but don’t be crass about suggesting your bros and besties from grad school or anyone with a clear conflict of interest. You’ll lose credibility with the editors, who have long memories for such shenanigans.

UNDERSTANDING REVIEWS AND RESUBMITTING YOUR MANUSCRIPT

What should we do once we submit our paper? Turn to the next one, naturally. We don’t reward productive writing by not writing (see Chapter 3), so chip away at your backlog while you wait to hear back from the journal. Experience shows that the editor’s action letter will arrive at the most inconvenient time, usually when a grant proposal is due in 2 weeks or you’re 85% done with another manuscript.

When the editor’s action letter arrives, read it. If you feel the need for some “emotional distance” before reading the reviews, strive instead for “intellectual closeness.” Your writing backlog is too vast and obdurate to care about your fragile fears and feelings. Setting aside the reviews for a few days is precious self-indulgence. Read the reviews and make a plan. The editor’s decision can take three forms: (a) the manuscript has been accepted; (b) the door is cracked open for a resubmission; or (c) the door is closed, locked, and sealed with crime-scene tape. Acceptance decisions are usually obvious. The editor says the manuscript has been accepted and tells you to expect some forms; sometimes, the editor accepts a manuscript pending minor changes. Although it’s rare that the first submission of a manuscript is accepted outright, it happens—one more reason to submit excellent first drafts.

When the door is open, the editor is willing to consider a revised version of your manuscript. This category ranges from encouraging letters that imply likely acceptance to discouraging letters that imply a long slog of revision. Wide-open doors involve easy changes, such as rewriting parts of the text or adding information. Barely open doors involve tedious changes, such as collecting more data and rewriting most of the text. If an editor warns you that the revised manuscript will be treated as a new submission, he or she is hinting that major revisions are needed. And when the door is closed, the editor never wants to see your manuscript again in any form or language. Don’t antagonize the editor by resubmitting a revised draft as a new manuscript or sending a whiny letter of protest, which is the researcher version of entitled grade-grubbing. It’s more dignified to take some lumps, rework the paper, and send it somewhere else.

The word reject in a decision letter doesn’t necessarily mean that you can’t resubmit the manuscript. Many editors use reject to refer to any manuscript that they aren’t accepting outright: They’re “rejecting” your first draft but expect to accept the revised one. In such cases, the journal’s online portal is usually easier to interpret because it classifies a decision into practical categories, like “Accept With Minor Revisions” or “Major Revision.” If the door is open for resubmission, you should almost always revise and resubmit. You have cheated the gods of rejection rates, so collect some burnt offerings, do a happy dance, and then revise your paper as quickly as possible. It is rarely rational to send your paper elsewhere instead of resubmitting it, but it might be if the journal wants changes that you’re unwilling or unable to make. But beware the temptation to submit elsewhere when faced with a daunting revision. It’s a subtle way of avoiding the tedious labor of revising.

After you commit to revise and resubmit your manuscript, you need to make a plan. Examine the editor’s letter and the reviews and extract the action points—the targets for change. Many reviews are discursive and meandering; a long review might have only a few action points. Underline each comment that implies a change—adding something, rewording something, amputating something. After you identify the action points, revise the manuscript quickly. Your paper is close to publication, so don’t slow down now (see Chapter 3).

For each action point, you have three options. First, you can make the change. Editors and reviewers have a lot of good ideas, so embrace change. Nothing puts you in the dramatic-and-erratic category like refusing to make simple changes, such as combining tables, deleting a figure, or reducing the word count by 10%. If you have an itchy urge to lash out at strangers who disagree with you on obscure and trivial matters, perhaps there’s an Internet nearby where you can get it out of your system. Second, you can resist a suggestion. Along with their many good ideas, editors and reviewers have some that seem ill-considered, uninformed, or unproductive. If you don’t make a suggested change, your revision letter will need to spend some time explaining your reasons. And third, you can punt the decision back to the editor. You’ll occasionally see a suggestion that is probably unimportant but you’re not sure how to handle it, such as when reviewers suggest changing your title, chopping the paper in half, or adding a bunch of figures for peripheral findings. In such cases, it’s fine to punt the issue to the editor: “To save space and keep the paper focused, we decided not to add the four new figures suggested by Reviewer 2, but we could if you think it would improve the paper.”

When you resubmit your manuscript, you’ll need to send a cover letter that describes how you handled the criticisms and comments. Here is where the publishing game is won or lost. Editors are busy, harried people who have a backlog of decision letters to write and plenty of good manuscripts to pick from. Like the rest of us, they appreciate a quick win. If you send them a reasonable, comprehensive, and low-drama cover letter, they can get a quick win by accepting your manuscript and moving on to the next one in their pile. But many letters are dramatic and erratic, seething with indignation and wounded pride, and others read like postcards from a dreary vacation (“thanks for your comments—here’s our revised manuscript”).

Here’s what a good revision letter looks like:

§ Create headings, organized by reviewer, for each set of action points. Make a set of headings—start with Editor’s Comments followed by Reviewer 1’s Comments, Reviewer 2’s Comments, and so on. Within each heading, address each action point in the order in which it was raised in the reviews. Numbered lists are easier for editors than a discursive, essay-style letter, so keep it crisp and clipped.

§ Tackle each action point with a three-part system: (a) briefly summarize the comment or criticism; (b) describe what you did in response to this comment, if anything, and cite page or line numbers in the revised manuscript when possible; and (c) discuss how your changes resolve the comment. When several reviewers raise the same action point—contrary to beginners’ complaints, reviewers usually agree—you should mention the issue again each time it is raised. Simply dispatch it by noting the comment and referring to the number of your earlier discussion.

§ Don’t be fawning and obsequious. Editors don’t expect you to refer to the reviewers’ comments as masterful, wonderful, or insightful. You’ll sound gratingly ingratiating.

And what do we do after submitting our revision letter and revised manuscript back to the journal? We turn to the next paper in the backlog, naturally.

FREQUENTLY GRUMBLED GRUMBLINGS ABOUT JOURNALS

“They’re Just Going to Reject My Paper”

Many writers get paralyzed by their fear of criticism and rejection. They’ll put off finishing their manuscript because they’re nervous about what happens after it leaves the nest and enters the cold, hard world of peer review. They imagine one reviewer scowling, another reviewer grimacing, and an editor slowly shaking her head as she fumbles for her enormous red REJECT stamp. And it is a hard world out there. Manuscripts, like sea turtle hatchlings, face many hazards during their harrowing journey from the nest to the open sea, and many never make it.

But rejection is a fact of academic life that we must accept. Most journals reject most of the papers they receive, so expecting rejection is rational. If a journal rejects 80% of its submissions, then its base rate of acceptance is 20%. Without any other information, 20% is the only rational estimate of our paper’s chance of getting accepted. Because good journals rarely have acceptance rates above 50%, we should marvel at how many people get published, not despair at the inevitability of rejection.

“That’s bleak,” some might say. “How can you be motivated to write if you expect rejection?” I suppose we all write manuscripts for the same reason baby sea turtles scrabble toward the sea. Writing is a species-typical behavior for academics, much like reading old books and voting in local elections. We feel an urge to share our ideas, but these ideas must go through the harrowing journey of peer review. This process is imperfect—a few sickly sea turtles make it to sea, and a few hardy ones get picked off by seabirds—but in the long run peer review sharpens our ideas and strengthens our field.

“I Can’t Fit Everything in”

Journal articles are small vessels for our ideas, so beginning writers get stuck when they can’t cram all their thoughts into their article. “I can’t fit it all in,” they say. “The manuscript gets too long when I say everything I want to say.” This is a good problem to have, I suppose, because it means you have a lot to say. If a research topic is worth studying, a short journal article shouldn’t exhaust our ideas. Some of the mental surplus can be used as the seeds for other articles or for talks, book chapters, and essays. You might have a thesis that needs a book-sized vessel (see Chapter 7).

But most of our excess ideas are drab and dreary: arcane implications that few readers will care to hear; long explications of secondary, tertiary, and quaternary findings; and lengthy reviews of the literature that your thesis committee wanted to see but your readers already know. People are more likely to read and understand articles that are focused and compact—that make a few important points well—so most of the secondary ideas should be chopped. Hacking down the brush and brambles of peripheral ideas can feel wrenching, but your readers will see farther when they’re gone.

“They’re Going to Make Me Change Everything”

Some writers struggle to finish a paper because they believe that all their hard work will be undone by meddlesome reviewers. “Why bother making it great,” they think, “when they’re just going to make me change everything anyway?” My peevish and stubborn side resonates with this “You’re not the boss of me!” outlook, but it isn’t true. Editing a journal isn’t like restoring a vintage car, where you strip it down to the metal frame and rebuild it with new parts. Instead, editors just scrap your jalopy of a paper and select a new, shiny one that won’t embarrass the neighbors. Editors get more good papers than they have space for, so they’ll reject your paper if it needs overhauling.

We must be willing to make big changes to our papers. Science holds published research to high standards and uses peer review to provide quality control. Our scholarly journals are our public and permanent record. Your article will be printed on acid-free paper and archived on library shelves for eternity, however long that is these days. Progress is faster when people connect their work to others’ ideas, apply methods that meet modern standards, and confront awkward questions about their research. Journals are not the place for us to pitch wild ideas, dump malformed projects, and uncork opinionated rants. Fortunately, the Internet invented blogs and social media for us to do all that.

So, yes, we will be asked to change our papers, and sometimes those changes will be extensive. And in virtually every case, those changes make our papers better.

CONCLUSION

When struggling to write their first article, some writers lament, “Why would they care about my research?” If they refers to the world at large, they probably won’t track down and read your article. But if they refers to researchers in your subfield, then you should expect some interest in your article. Your paper might be rejected once or twice before it finds a home, but a good paper will always find a good home. To write good articles, pick your journal first, outline according to the standard templates, submit great first drafts, and craft excellent resubmission letters.

After you have published a few articles, you’ll find that the world of peer-reviewed journals isn’t scary, merely slow. And after mastering the articles game, you might be ready for a bigger challenge—writing a book.