You are here: Resources --> Guidelines for preparing journal manuscripts
Vous êtes ici : Ressources --> Guidelines for preparing journal manuscripts

Guidelines for preparing journal manuscripts

The guidelines in this document will help you create a manuscript that will be easier to edit (thus, less time-consuming and expensive to edit), easier for reviewers to review, easier for the journal to publish, and easier for readers to understand without error. I will provide descriptions for Microsoft Word 2007 and later versions; other word processors and versions of Word will have similar settings. For much more extensive instructions and a learning plan, please see my book Writing for Science Journals.

This guide contains the following sections:

• Before you begin writing
• Manuscript preparation checklist
• Author information
• Abstract
• Introduction
• Methods and Materials
• Numbers, variables, and equations
• Results
• Discussion
• Conclusions
• References and literature citations
• Creating figures
• Creating tables
• Final advice

Note: Every journal has author guidelines, and you should follow them if they contradict the advice in this document. If you are preparing a manuscript for a conference or symposium, the organizers may send you guidelines on how to prepare your manuscript. To avoid delays, follow the journal or symposium guidelines.

Before you begin writing

• Create a strong outline that defines the contents and sequence of your paper, and rigorously review this outline with help from your colleagues or graduate students before you begin writing. I have provided a tutorial on writing good outlines.
• I will be happy to edit your outline before you begin writing. In my experience, this increases the productivity of writing sufficiently that it negates the cost of this additional editing.
• Ensure that all authors read and approve the manuscript before you submit it for editing or for review by the journal. NEVER skip this step, as it is essential for finding and correcting errors and other problems. Anything your co-authors do not understand will be difficult or impossible for readers (including me) to understand.
• To minimize errors, ALWAYS ask one or more of your co-authors to verify the data that has been entered into your statistical software and to repeat all calculations independently to ensure that they are correct. In the worst case, errors in the literature can endanger human lives. At a minimum, they will mislead future researchers and waste research dollars.

Manuscript preparation checklist

I have provided a Microsoft Word template that contains many of the changes described in this set of guidelines. If you will be using your own template to create the document, please make the following changes if you have not already done so:

• Text alignment: If you mostly write in Chinese, Japanese, or other languages that use a grid (#) to align characters or control the character spacing, turn off this setting when you are writing in English. It can cause serious problems for Western reviewers; in some cases, they may be unable to read your manuscript. To make this change, select the Page Layout tab of the ribbon, open the menu for “Align” and select “Grid Settings”. In other versions of Microsoft Word, this setting may be found under the Format menu: select "Document", then in the tab labeled "Document grid", select "No grid".
• Text size: Most journals require that all text be set to 10, 11, or 12 points in size. Of these, 12 points is most common because it is easiest to read, so I recommend that you use that size. If you find that the text is too large on your screen, you can solve that problem by adjusting the magnification (“zoom”) setting in the View tab of the ribbon or under the View menu. If your computer's mouse has a scroll wheel, you can also hold down the Control key and rotate that wheel. This changes the magnification interactively until the size is comfortable for you to read. Do not waste your time manually adjusting the text size for different parts of the document; journals prepare manuscripts for publication using automated software that will eliminate all your hard work and apply the journal’s formats. Only use smaller text when it is necessary to (for example) fit a table on a single sheet of paper.
• Text fonts (typefaces): Use Times New Roman for text and Arial or Verdana for graphics; these fonts are available on every computer. Cambria Math and Symbol are sometimes acceptable for special characters. Never use non-English fonts such as SimSun and Minchin; they can make it difficult or impossible for reviewers to read your manuscript. It is sometimes acceptable to insert special characters using the Symbol font, but most of the common special characters can be typed in Times New Roman or Arial using the following keyboard shortcuts:
  - Microsoft Windows shortcuts
  - Apple Macintosh shortcuts
• Other symbols: If the symbol you need is not present in the two shortcut documents, most software has a feature that lets you choose symbols from a visual list. In Microsoft Word, this is in the Insert tab of the ribbon or the Insert menu, and the feature is called “Symbol”. To insert the symbol from this dialog box: (1) Select the correct font (one of the options listed above). (2) Select the correct symbol from the list of available symbols.
• Numbering: Add continuous line numbers. This makes it easy for reviewers to describe the location of their changes, and easy for you to describe the location of your reponses. It also eliminates the need to type the page number. If you restart line numbering on each page, ensure that each page is numbered. To add line numbers, select the Layout tab of the ribbon and open the menu for “Line Numbers”.
• Line spacing: Many journals still require that the line spacing be set to double-spaced. Because most reviews are now done on the computer screen, this is less necessary than it used to be, but you should still use double-spacing if the journal guidelines do not specify single spacing. The best way to apply this format is to define this spacing in the paragraph styles; a simpler way is to select all of the text and apply the spacing manually:
  - Microsoft Windows computers: Press Control+A, then press Control+2.
  - Apple Macintosh computers: Press Command+A, then press Command+2.

Author information

• If your family name is not English, capitalize all letters in the name (e.g., HART, not Hart) or provide a list of family names below the title. Without this help, English journals often choose the wrong name. If your name is incorrectly recorded in the list of authors, future researchers will have difficulty finding your paper in a literature search.
• Clearly indicate the corresponding author (usually with an *), and provide their complete mailing address. Do NOT use the word processor’s footnote feature. Type this contact information on the first line after the list of author names and addresses.
• Provide your telephone and fax numbers. If they are the same number, use “Tel./Fax:” to make this clear.

Abstract

• Most journals require an Abstract shorter than 200 or 250 words. Confirm the length limit for the journal that will review your paper, and try to reduce the length to meet that restriction. (Most word processors offer a “word count” feature that lets you select (highlight) the text in the Abstract and determine its length.)
• Remove all literature citations.
• Only present numerical results in the Abstract if you have presented them in the Results or Discussion.
• Whenever you change a number or conclusion in the main text, make the same change in the Abstract.

Introduction

• Start by describing the problem you were trying to solve.
• Describe how previous researchers tried to solve that problem, and how you are building on their research to solve any problems they could not solve.
• Remember that readers of English-language journals are able to read English. You cannot know what other languages they can read. Thus, always cite English references to describe general principles or methods.
• It’s acceptable to cite non-English references for research that is not available in English (e.g., results for a non-English country), particularly to support a study that you performed in a non-English country. However, if it is likely that most readers will be unable to read and understand the papers you are citing, concisely summarize their important points (e.g., methods, results) in English.
• End the Introduction with a list of your research objectives and the research hypotheses that your study tested. Clearly explain what is new about your research.

Methods and materials

• The goal of this section: to let me (or any other reader) repeat your research exactly and achieve the same results (i.e., replication). Always ask someone who did not participate in the research to read this section and tell you what additional things they need to know to achieve that goal.
• Provide English literature citations for standard methods. Non-English literature citations are acceptable if there are no English descriptions available. However, then you must then provide enough details that readers can fully understand how you performed your research.
• For models (equations and computer programs): Name variables so that they resemble the words they will represent (e.g., m = mass, not b = mass; BAI = basal area increment, not I). This makes the meaning of the names easier to remember, and reduces the risk of using the wrong variable in your model. Never reuse a symbol or name for two different variables. This can lead to serious errors.
• Mathematicians and computer programmers are trained professionals; all have at least 1 year of formal training, and most have a 3-year or 4-year degree in their subject. You CANNOT acquire the same level of expertise without a similar investment of time. Always ask a professional to review your work.

Numbers, variables, and equations

• Spacing: Always add a space between numbers and units of measurement (except for %).
• Hyphenating numbers: When a number and a unit of measurement work together to describe a noun, use a hyphen instead of a space. For example: “the tree was 1[space]m tall” versus “we studied a 1-m tree”.
• Format: Italicize simple variables; boldface matrices and vectors. Letters that are not variables should not be italicized or boldfaced. For example, if the variable name is V and you measure that variable before (b) and after (a) a treatment, the resulting variable names would be V_b and V_a. Note that "a" and "b" are not italicized because they are not variables.
• Greek: Greek letters (α, β, γ, δ, etc.) are traditionally not italicized.
• Equations: For all simple variables and equations, type them directly from the keyboard. Do not insert them as graphics, and only use an equation editor if the journal guidelines state that this is acceptable.
• Statistics: For all correlations (r values) and all regression goodness of fit values (R²), use the correct name (r or R²); the two parameters have different meanings. Also, provide the p value to indicate whether the correlation or regression is statistically significant.

Results

• Unless you have combined the Results and Discussion into one section, the Results section should NOT include explanations or interpretations of your results. Move that information to the Discussion.
• The Results should not include literature citations. The goal is to present your results, NOT to compare them with other research results.

Discussion

• Interpret your results and compare them with previous research. Provided citations for that research. If there are contradictions, explain those contradictions based on your own data and based on research from elsewhere in the world.
• Compare your results with studies that are as similar as possible to your study conditions. If you are studying trees in the boreal forest, do NOT compare your results with results for herbs growing in tropical grasslands unless your goal is to compare different systems. Then, clearly indicate the nature of the contrast, and describe the different conditions.
• Conclude by identifying any limitations of your study that must be solved in future research.

Conclusions

• The Conclusions section is similar to the Abstract, but there are two main differences: (1) Describe only broad overall findings rather than presenting numerical data you have already presented at least twice (in the Results and the Discussion). (2) Describe how those results relate to your study hypotheses and goals.
• Focus on your results. Thus, do not include citations for research by other researchers.

References and literature citations

• Citations: To ensure that all references have been cited correctly, copy all the references into a new file, and display that file in a document window beside the window that contains the manuscript. As you write and revise the paper, you can quickly check that each citation is present in the list of references. Confirm that the author name (or names) and the year of publication for each citation is correct. Using reference-management software greatly facilitates this process. If you are NOT using reference-management software, and the journal that will review your paper requires numbered citations (e.g., [1, 2, 7]), use the author/date (e.g., Hart 2018) citation method until you finish revising your paper and reviewing my edits. Then replace the name/date citations with the correct numbers. This eliminates the need to repeatedly renumber all your citations and reduces the frequency of citation errors. If you do use reference-management software, use it to insert ALL citations, and confirm that you selected the correct reference for each citation.
• et al.: Note that “et al.” means “and colleagues”. Do NOT use these words for papers with only 2 authors. If there are only 2 authors, type both names (e.g., Able and Baker, not Able et al.).
• Avoiding mistakes: Copy the author and title information from the journal’s Web site or from a library database. Paste the copied information into your References list. You will need to reformat that information (see the next point), but this is faster and more accurate than typing the information manually. If there are footnote letters (a, 1, †) after the author names, delete those symbols.
• Formatting: Consult a recent issue of the journal to learn their formatting guidelines. The only way to correctly format all references is to work on one characteristic of the references at a time: (1) Confirm that all names are correctly formatted. (2) Confirm that all the years are in the correct position (e.g., after the author names or after the journal name) and are correctly formatted (e.g., with or without brackets). (3) Ensure that the titles are correctly formatted. (4) Ensure that all journal names are correctly formatted (e.g., abbreviated versus full words, italics versus normal type). (5) Ensure that the punctuation for the volume number and page range are correct (e.g., using a “:” instead of a “,” before the page range). (6) For book chapters or symposium papers, provide the names of the editor or editors (followed by “ed.” for 1 editor or “eds.” for 2 or more editors), and ensure that you have provided the page range, and the name and city of the publisher of the proceedings.

Creating figures

Every journal provides guidelines for preparing figures. Read them carefully and follow the instructions. In addition:

• Software: Most journals do not accept graphics created in Microsoft Word or Powerpoint, but you can use these programs if you save the figures in one of the formats that the journal accepts (PDF is a good choice) and then add the resulting file to your manuscript. Consult the author guidelines to learn which formats are acceptable. If you cannot afford expensive software such as Adobe Illustrator, Adobe Photoshop, or CorelDraw, I have provided a list of less expensive or free alternatives.
• Figure size: Most journals still publish a printed version, and figures must be created at a size that will fit within the page of the printed journal. ALWAYS create your figures at their final size, so that they will fit within 1 or 2 columns of text in the printed version of the journal without having to be enlarged or reduced in size. Then print a copy. If the text is even slightly difficult to read, increase the size until it is easy to read. If any colors, symbols, or patterns are difficult to distinguish, replace them with clearer alternatives.(Where a larger size is necessary, the maximum size should be about 1 cm less than the dimensions of the printed page.) If you create figures that are larger than this size, the journal will reduce their size and parts of the figure may become unreadable. Consider breaking the figure into smaller sub-figures, or present it as a supplemental figure that will only appear on the journal's Web site.
• Text size: Text in figures can be smaller than text in the manuscript, but it should not be so small that it is difficult to read. Print a copy of your figures at their final size (to fit within 1 or 2 columns of text in the printed journal), and hold them at a typical reading distance. If any of the text is not easy to read, increase its size. If any of the symbols or colors or patterns are unclear or cannot be easily distinguished from other symbols or colors or patterns, change them. Print the figure again to confirm that you have solved the problem. Remember: many researchers have vision problems, and it is better to increase the size than to risk creating a figure that only you can read.
• Line thickness: Lines must be at least 0.25 points in thickness, and up to 1 point in thickness to ensure that they will be visible when the graphic is printed or viewed on a small screen such as a smartphone or tablet computer.
• Key/legend: It is always clearer to define symbols, colors, and patterns in the key or legend of the graph. Never describe the symbols, colors, and patterns in the figure caption. Although it is possible to insert symbols in the figure caption, the journal’s automatic processing software may change them to a different font, and they will no longer correspond to the symbols in the graphic. Including this information in the graphic ensures that readers will be able to see the symbols exactly as they appear in the graphic.
• Color: Wherever possible, use solid colors (white, grey, and black) instead of patterns (\\\, ///, |||). Patterns often print unclearly; small areas of a figure labeled with these patterns can become difficult or impossible to distinguish. Try to avoid using yellow; in print, the color may be so light that it becomes invisible, and even on the screen, it can be very difficult to read. Color is very expensive to publish, and most journals ask the author to pay this cost. Look for ways to redraw your figures in black, white, and shades of grey. If color is necessary, consider presenting your figures as online supplemental figures that will only appear on the journal’s Web site; there is usually no additional cost for this. If you must use color for the printed journal, create your graphics in software that supports the “CMYK” color format, which is the format used for color printing. The “RGB” color format is only acceptable for graphics that will appear on the computer screen and that will not be printed.
• 3D (three dimensions): Avoid using 3D graphics for data that has only 2 dimensions. This makes it difficult to reliably estimate the value of a parameter and difficult to accurately compare values, and can lead to serious interpretation errors.
• Axes and axis scales: Add a space before brackets that contain units of measurement. English numerical intervals are usually multiples of 1, 5, or 10; Asian intervals are often multiples of 2, 3, or 4, but these intervals should not be used for an English journal. Change the settings of your software to ensure that ticks on the axis (– and |) are beside the numbers on the axis and outside the area of the graph that contains the data. For two graphs that show values of the same parameter, use the same scale (range of values) on the axis. At least 1 time per month, one of my authors misinterprets their own data and reaches the wrong conclusion because they used different scales for the same parameter. I assume that readers make such mistakes even more frequently because they are unfamiliar with your data and will not examine your graphs closely. Where two treatments, locations, or other conditions have very different values for a parameter, present them in the same graph so readers can easily see that they have different ranges of values.
• Panels (parts) of a graphic: Where a figure consists of 2 or more panels (e.g., 2 graphs, or 1 graph and 2 photos), add the letters (a, b, c... or A, B, C...) that identify the panels in your graphics software, not in your word processor. In English, these should appear at the top left side of the figure. In addition, submit the figure to the journal as a single graphic file, not as multiple graphics files. This is the only way to ensure that the letters are not lost and that the panels appear in the right location; journal production staff often lack sufficient time to carefully examine your graphics to ensure they are present in the right positions, and many errors occur when a figure is submitted as multiple files.
• Position: Most journals ask authors to present their figures at the end of manuscript, after the References section. Only include figures in the body of the text if the journal guidelines require this; in that case, insert them immediately after the paragraph where they are cited. If you include the figures in the body of the manuscript instead of as separate files, insert them in the line above the figure’s caption (title). Use the “inline” format so that the text does not “wrap” (flow) around the figure; that is, turn off the “text wrap” property.
• Abbreviations: Define all abbreviations that you used in the figure in its title. Never ask readers to search through your manuscript to find the meanings of these abbreviations.
• Axis text: Present simplified descriptions of variables on the axes of the graph. For example, if you define the variable “net photosynthesis (Pn)” in the caption, then use Pn instead of the full words on the axis. Only use the “/” symbol for division (e.g., m/sec). Use brackets, not /, around units of measurement.

Creating tables

I have provided some basic instructions on how to use Microsoft Word's table features in case you are not already familiar with these features. In any word processor:

• Format: Tables must be submitted to the journal in editable format, not as graphics. This is the only way that the journal can apply its own text and line formats when it prepares your tables for publication. Tables should be created using the table features of your word processor, or using a spreadsheet such as Microsoft Excel. They should not be “embedded” in your word processor file with links to an external file; this can cause serious problems if the external file is updated but the word processor document does not account for these changes.
• Position: Most journals ask authors to present their tables at the end of manuscript, after the References section. Only include tables in the body of the text if the journal guidelines require this; if they do, insert them immediately after the paragraph where they are cited. If you include the tables in the body of the manuscript instead of as separate files, insert them in the line below the table’s caption (its title). Use the “inline” format so that the text does not “wrap” (flow) around the table; that is, turn off the “text wrap” property.
• Size: Tables that will appear in the printed version of the journal should be designed to fit within 1 or 2 columns of text in the printed journal. If you need more space, the maximum size is about 1 cm smaller than the dimensions of the journal’s printed pages. When you prepare your tables, never define the table size numerically. The journal will make the necessary corrections to ensure that the table fits within their printed page. If a table contains so much information that it cannot easily fit within 1 printed page, it is ineffective and should be redesigned to present only the parameters and values that you will be discussing and comparing simultaneously in the same table; that is, create new tables for different comparisons of subsets of your data.
• Large tables: Tables should summarize your data, not present all of your data. Databases and large spreadsheets should be provided as Online Supplemental Material.
• Alignment: There are different preferences for where information should appear in the cells of a table. However, it is always acceptable to align all text in a table with the top left corner of the cell that contains the text. Numbers should be aligned based on the decimal point. Based on my experience with reviewer comments, aligning text with the top of the cell that contains it eliminates misunderstandings that can lead to criticism of your tables.
• Abbreviations: Define all abbreviations used in the table in the title or as a footnote below the tables. Never ask readers to search through your manuscript to find the meanings of these abbreviations.
• Units of measurement: These must appear in the column titles, not in the table title. Only use the “/” symbol for division (e.g., m/sec). Use brackets, not /, around units of measurement.

Final advice

• Before you send me your manuscript, use the spellchecker. If you are not certain that a word is wrong, leave it for me to fix. The spellchecker will help you find the most common spelling errors, but its advice is often poor for complex words and phrases.
• The best thing you can do to improve the quality of your manuscript, and decrease the editing cost, is to ask one or more of your colleagues to review the manuscript. Anything that they do not understand is something that I will not understand, and is something that you should improve before you send me the manuscript. It can be embarrassing when your colleagues cannot understand your paper, but this is less embarrassing than having the journal editor or reviewers criticize the clarity or logic of your paper.

©2004–2024 Geoffrey Hart. All rights reserved.