ArticlePDF Available

An Efficiency Comparison of Document Preparation Systems Used in Academic Research and Development

Authors:

Abstract and Figures

The choice of an efficient document preparation system is an important decision for any academic researcher. To assist the research community, we report a software usability study in which 40 researchers across different disciplines prepared scholarly texts with either Microsoft Word or LaTeX. The probe texts included simple continuous text, text with tables and subheadings, and complex text with several mathematical equations. We show that LaTeX users were slower than Word users, wrote less text in the same amount of time, and produced more typesetting, orthographical, grammatical, and formatting errors. On most measures, expert LaTeX users performed even worse than novice Word users. LaTeX users, however, more often report enjoying using their respective software. We conclude that even experienced LaTeX users may suffer a loss in productivity when LaTeX is used, relative to other document preparation systems. Individuals, institutions, and journals should carefully consider the ramifications of this finding when choosing document preparation strategies, or requiring them of authors.
Content may be subject to copyright.
RESEARCH ARTICLE
An Efficiency Comparison of Document
Preparation Systems Used in Academic
Research and Development
Markus Knauff*, Jelica Nejasmic
Department of Psychology, Experimental Psychology and Cognitive Science, University of Giessen, Giessen,
Germany
*markus.knauff@psychol.uni-giessen.de
Abstract
The choice of an efficient document preparation system is an important decision for
any academic researcher. To assist the research community, we report a software
usability study in which 40 researchers across different disciplines prepared
scholarly texts with either Microsoft Word or LaTeX. The probe texts included
simple continuous text, text with tables and subheadings, and complex text with
several mathematical equations. We show that LaTeX users were slower than Word
users, wrote less text in the same amount of time, and produced more typesetting,
orthographical, grammatical, and formatting errors. On most measures, expert
LaTeX users performed even worse than novice Word users. LaTeX users,
however, more often report enjoying using their respective software. We conclude
that even experienced LaTeX users may suffer a loss in productivity when LaTeX is
used, relative to other document preparation systems. Individuals, institutions, and
journals should carefully consider the ramifications of this finding when choosing
document preparation strategies, or requiring them of authors.
Introduction
The key communication of academic research and development is through diverse
forms of publications. Most scholars spend many hours writing journal articles,
books, or other forms of scholarly text. Virtually all researchers use one of two
document preparation systems: Microsoft Word or LaTeX. Publishers often
accept just one of the two text file formats [1]. Microsoft Word is based on a
principle called ‘‘What you see is what you get’’ (WYSIWYG), which means that
the user immediately sees the document on the screen as it will appear on the
OPEN ACCESS
Citation: Knauff M, Nejasmic J (2014) An
Efficiency Comparison of Document Preparation
Systems Used in Academic Research and
Development. PLoS ONE 9(12): e115069. doi:10.
1371/journal.pone.0115069
Editor: Cynthia Gibas, University of North Carolina
at Charlotte, United States of America
Received: July 29, 2014
Accepted: November 18, 2014
Published: December 19, 2014
Copyright: ß2014 Knauff, Nejasmic. This is an
open-access article distributed under the terms of
the Creative Commons Attribution License, which
permits unrestricted use, distribution, and repro-
duction in any medium, provided the original author
and source are credited.
Data Availability: The authors confirm that all data
underlying the findings are fully available without
restriction. All relevant data and information about
how the data were analyzed are included in the
Supporting Information files.
Funding: The study was financially supported by
the University of Gießen. The funder had no role in
study design, data collection and analysis, decision
to publish, or preparation of the manuscript.
Competing Interests: The text processing soft-
ware Microsoft Word is licensed and traded by the
Microsoft Corporation. The authors do not have
any connection to this company, and the Microsoft
Corporation had no role in the study design, data
collection, data analysis, decision to publish, or
preparation of the manuscript.
PLOS ONE | DOI:10.1371/journal.pone.0115069 December 19, 2014 1/14
printed page. LaTeX, in contrast, embodies the principle of ‘‘What you get is what
you mean’’ (WYGIWYM), which implies that the document is not directly
displayed on the screen and changes, such as format settings, are not immediately
visible. Microsoft Word requires little start-up time and provides easy and
instantaneous control of textual input and output. Microsoft Word is the
predominant document preparation system across many disciplines, including
medicine, law, business, and the life sciences, and is also the dominant document
preparation system for professional communications. LaTeX, in contrast, is a
programming language that requires the use of an external editing interface to
produce documents. LaTeX is frequently used in mathematics, physics, computer
science, and engineering because it provides the user unlimited flexibility and is
particularly useful if the user needs to set complex mathematic equations in a
professional layout. LaTeX is freely available as open-source software. In contrast,
Microsoft Word is a commercial product licensed by the Microsoft Corporation.
In the ‘‘publish or perish’’ age of academic research, many senior researchers
advise their students and junior researchers about how to create professional
document layouts, which software system to use, and which system is more
efficient or user-friendly. Many of these senior researchers will attempt to
convince their students and junior researchers that one system is ‘‘better’’, ‘‘more
elegant’’ ‘‘simpler’’, or ‘‘more flexible’’ than the other system. There are very few
researchers, however, who can confirm empirically how one system is superior to
the other and on what basis they have drawn this conclusion. To date, no
empirical studies exist to identify which system is more efficient. The preference
toward a particular document preparation system can be particularly obstructive
to the progress of research if the research question requires interdisciplinary
teams. For example, a brain computer interface project may require collaborations
between medical scientists, psychologists, computer scientists, biologists, physi-
cists, and engineers. Any researcher who has ever collaborated on such large
interdisciplinary projects has experienced the difficulty with reaching a consensus
about which document preparation system to use. Discussions about document
preparation systems are often unproductive and driven by preconceived opinions,
individual biases, and disciplinary traditions. A fair comparison of the efficiency
and usability of the different document preparation systems based on empirical
evidence rather than individual habits and biases may facilitate such discussions.
Participants, Materials and Methods
To assist the academic research community in the choice of an efficient document
preparation system, we empirically compared the usability of LaTeX and Word
under highly realistic working conditions. The volunteers for this study included
40 researchers and advanced graduate students from six German universities who
wrote scholarly texts in either Microsoft Word or LaTeX (mean age 25.4 years; 14
female; Physics: 12; Psychology: 5; Computer Science: 4; Mathematics: 4; Electrical
engineering: 3; MBA: 3; Sport Science: 4; others: 5). They were recruited from
Document Preparation in Academic Research and Development
PLOS ONE | DOI:10.1371/journal.pone.0115069 December 19, 2014 2/14
newsgroups, mailing lists, blogs, and other sources. Most participants were tested
in their personal office setting, and all participants used their own computer,
which ran either the Windows or Linux operating system. They were informed
that the purpose of the study was to evaluate the quality of their document
preparation system they use in their daily work.
All participants were properly instructed and have indicated that they consent
to participate by signing the informed consent paperwork. The study has been
conducted according to the principles expressed in the Declaration of Helsinki;
the risks of the study were no higher than those experienced by people using their
respective software (Word or LaTeX) on a day-to-day basis, participants could
withdraw from the task at any time, and no identifiable data will be released about
participants. For such studies the ethical guidelines of the Deutsche Gesellschaft
fu
¨r Psychologie (German Psychological Society, DGPs) and the Bund Deutscher
Psychologen (German Psychological Association, BDP) revised on June 28, 2004
specify that approval from an Ethics Committee can be waived "if it can
reasonably be assumed that participation in the research produces no damage or
no discomfort that go beyond everyday experience, and if the research (a) refers to
common education methods, curricula or teaching methods in education; … or
(c) refers to factors that affect work and organizational efficiency in organizations
whose investigation can have no occupational disadvantages for individuals and
for which confidentiality is guaranteed (p. 2, paragraph 6)’’. The present research
belongs to this class of studies; thus, no further approval from an ethics
committee was required.
The participants were divided into 4 groups with 10 participants in each group:
Word novices, Word experts, LaTeX novices, and LaTeX experts. Participants
were classified as ‘‘novices’’ if they had less than 500 hours of experience with the
respective program and ‘‘experts’’ if they had more than 1000 hours of experience
with the respective program. In the resulting groups, participants who were
classified as ‘‘novices’’ had on average 234 hours (SD5153) experience with the
respective program, whereas ‘‘experts’’ had on average 1909 hours experience
with the respective program (SD5211).
The probe texts included three different text structures: (1) simple continuous
text; (2) text with tables; and (3) mathematical text with several equations. The
texts were selected based on a pilot study so that an expert could reproduce
around 90% of the text in thirty minutes. All texts came from the Journal
‘‘Kognitionswissenschaft’’ which was the official Journal of the German Cognitive
Science Society until the year 2002. The selected texts are presented in Fig. 1. The
continuous text consisted of a headline and headings with different font sizes, four
paragraphs, and two footnotes (Fig. 1). The table text consisted of a headline, two
paragraphs, and a table that was divided into several segments and surrounded by
text (Fig. 2). The equation text consisted of a headline, four paragraphs, and six
equations (Fig. 3). Participants were allowed to use all tools, editors, plug-ins, and
add-ons that they were accustomed to using with their respective software. For
example, many LaTeX users produce documents with external text editors such as
TeXnicCenter, LaTeX Editor, Kile, or WinEdit because LaTeX does not offer an
Document Preparation in Academic Research and Development
PLOS ONE | DOI:10.1371/journal.pone.0115069 December 19, 2014 3/14
Document Preparation in Academic Research and Development
PLOS ONE | DOI:10.1371/journal.pone.0115069 December 19, 2014 4/14
internal text editor. All participants already had some experience with formatting
tables and equations and were tested in the presence of the experimenter. The
three text types were presented in a random order to each participant. The
participants were instructed to reproduce the source text within thirty minutes.
Each participant was given five minutes to familiarize themselves with the text.
The performance of each participant was measured for each text sample by three
variables: (1) the number of orthographic and grammatical mistakes; (2) the
number of formatting errors and typos; and (3) the amount of written text (in
symbols and words) produced within 30 minutes. Table 1 provides an overview of
all possible errors in the three probe texts. To measure the user’s opinions and
satisfaction with their software system, each participant also completed an
international standard questionnaire (ISO 9241-10) about usability engineering.
To motivate the participants, the best three performers from each group received
a monetary prize of 150, 100, or 50 euros, respectively. In the following section, we
report the performance of the four groups of participants for the three types of
probe text. Then, we report the results of the ISO 9241-10 questionnaire, which
examines how well each document preparation system fulfilled the general
ergonomic principles that apply to the design of dialogues between humans and
information systems. In the final part of the article, we present some psychological
explanations for the reported results and discuss some implications for academic
research and development.
Results
The performance of the four experimental groups (Word novices, Word experts,
LaTeX novices, and LaTeX experts) on the three probe texts (continuous text,
table text, and equation text) is summarized in Table 2 and Figs. 2,3, and 4. The
results of the usability questionnaire are presented in Table 3.
Continuous text
As shown in Table 2a and Fig. 4, Word users (both novices and experts) made
fewer formatting mistakes (t(37.97)525.94, p,.001) and wrote significantly
more text within 30 minutes (t(38)53.10, p,.01) compared with LaTeX novice
and expert users. The number of orthographic and grammatical errors did not
differ significantly between Word and LaTeX users (t(38)521.02, p5.31).
However, Word experts made significantly fewer formatting mistakes than LaTeX
experts (t(18)524.15, p,.01) and Word novices made significantly fewer
formatting mistakes than LaTeX novices (t(17.92)524.05, p,.01). Interestingly,
Word novices also made significantly fewer formatting mistakes than LaTeX
Fig. 1. The continuous text used in the present study. From: Jameson, A. & Buchholz, K. (1998). Einleitung zum Themenheft ‘‘Ressourcenadaptive
kognitive Prozesse’’, Kognitionswissenschaft, 7, 95.
doi:10.1371/journal.pone.0115069.g001
Document Preparation in Academic Research and Development
PLOS ONE | DOI:10.1371/journal.pone.0115069 December 19, 2014 5/14
Document Preparation in Academic Research and Development
PLOS ONE | DOI:10.1371/journal.pone.0115069 December 19, 2014 6/14
experts (t(17.98)523.84, p,.01). Word experts wrote significantly more text
than LaTeX experts (t(18)52.24, p,.05) and Word novices wrote significantly
more text than LaTeX novices (t(18)52.31, p,.05).
Table text
As shown in Table 2b and Fig. 5, Word users (both novices and experts) made
significantly fewer formatting mistakes (t(36.78)526.72, p,.001) and wrote
more text within 30 minutes (t(31.73)54.31, p,.001) compared with LaTeX
novice and expert users. Word experts made significantly fewer formatting
mistakes than LaTeX experts (t(16)524.40, p,.001) and Word novices made
significantly fewer mistakes than LaTeX novices (t(18)524.98, p,.001). Word
experts wrote significantly more text than LaTeX experts (t(14.19)52.68, p,.05)
and Word novices wrote significantly more text than LaTeX novices (t(15.99)5
3.52, p,.01). Interestingly, Word novices made significantly fewer formatting
Fig. 2. The table text used in the present study. From: Mu¨ller, B. (1998). Kompositionsbildung bei Symbolfolgen und Bediensequenzen: Empirische
Befunde und die Theorie des ‘‘Competitive Chunking’’, Kognitioswissenschaft, 7, 85.
doi:10.1371/journal.pone.0115069.g002
Fig. 3. The equation text used in the present study. From: Spies, M. (1999). Das Langzeitgeda¨ chtnis als Boltzmann-Maschine, Kognitionswissenschaft,
8, 71.
doi:10.1371/journal.pone.0115069.g003
Document Preparation in Academic Research and Development
PLOS ONE | DOI:10.1371/journal.pone.0115069 December 19, 2014 7/14
mistakes (t(17.33)524.78, p,.001) and produced more text than LaTeX experts
(t(15.99)53.52, p,.01).
Equation text
As shown in Table 2c and Fig. 6, LaTeX users (both novices and experts) made
significantly fewer formatting mistakes (t(38)53.35, p,.01) and wrote more text
within 30 minutes (t(38)522.96, p,.001) compared with Word novices and
experts. However, LaTeX users made significantly more orthographic and
grammatical errors than Word users (t(38)522.96, p,.01). LaTeX novices
made significantly fewer formatting mistakes (t(17.61)53.57, p,.01) and also
wrote more text (t(18)524.30, p,.001) than Word novices. Overall, however,
the performance of LaTeX experts and Word experts did not differ significantly.
Usability questionnaire
The international standard questionnaire ISO 9241-10 measures user’s opinions
and satisfaction with their software system. The questionnaire addresses general
ergonomic principles that apply to the design of dialogues between humans and
information systems, including suitability for the task, suitability for learning,
suitability for individualization, conformity with user expectations, self-descrip-
tiveness, controllability, and error tolerance. Furthermore, we asked whether users
perceived their work with the respective software as tiresome, frustrating, or
delightful. Participants rated their software on a seven-point scale from very bad
(-3) to very good (3). As shown in Table 3, Word users rated their respective
software as less efficient than LaTeX users (t(35.6)522.80, p,.01), but LaTeX
users rated the learnability of their respective software as poorer than Word users
Table 1. Overview of possible mistakes in the three probe texts.
Continuous text Table text Equation text
Orthographic and grammatical mistakes in words X X X
in formulas X
Formatting errors and typos header X X X
headline X — —
paragraph X X X
spacing X X X
font X X X
footnote X — —
columns X X X
lines X —
justified text X X X
Amount of written text missing words X X X
missing signs X X X
Note: X 5possible; — 5Not possible.
doi:10.1371/journal.pone.0115069.t001
Document Preparation in Academic Research and Development
PLOS ONE | DOI:10.1371/journal.pone.0115069 December 19, 2014 8/14
(t(33.6)52.13, p,.05). However, LaTeX users assessed their work as less tiresome
(t(35.38)52.16, p,.05) and less frustrating than Word users (t(38)52.27,
p,.05). LaTeX users significantly more often reported to enjoy their work with
their respective software than Word users (t(36.27)523.23, p,.01).
Discussion
Many academic authors believe that they have the right to design documents
themselves and that each researcher should have the freedom to choose the
software that he or she prefers. In fact, our study shows that each document
preparation system has unique advantages and disadvantages, and there might be
no ‘‘best’’ tool for all aspects of a highly complex task such as producing diverse
scientific publications. For example, LaTeX users in our study attained better
performance in the typesetting of mathematical equations, and it is not surprising
that LaTeX users are typically in disciplines where mathematical formulas are
frequent (e.g., mathematics, engineering, or computer science). Indeed these
Table 2. Mean absolute frequencies of orthographic and grammatical mistakes, formatting errors and typos, and the amount of written text (i.e., number of
words) across all four groups for the continuous text (a), the table text (b), and the equation text (c).
a. Continuous text
Word LaTeX
Novices Experts Overall Novices Experts Overall
M SD M SD M SD M SD M SD M SD
Orthographic and grammatical mistakes 5.9 3.5 7.9 6.7 6.9 5.3 7.0 6.6 11.3 9.5 9.2 8.2
Formatting errors and typos 10.0 3.9 9.3 4.1 9.7 3.9 17.3 4.1 16.1 4.0 17.1 4.0
Amount of written text 331 49.1 379 11,7 355 42.4 250 104 308 99.4 279 103.3
b. Table text
Word LaTeX
Novices Experts Overall Novices Experts Overall
MSDMSDMSDMSDMSD MSD
Orthographic and grammatical mistakes 9.9 7.8 7.1 4.4 8.5 6.4 9.7 10.5 7.8 4.5 8.8 7.9
Formatting errors and typos 12.0 3.7 11.3 4.1 11.4 3.9 19.5 3.6 18.7 3.0 19.1 3.3
Amount of written text 353 82.9 395 78.7 374 81.6 191 118 260 137.8 226 130
c. Equation text
Word LaTeX
Novices Experts Overall Novices Experts Overall
MSDMSDMSDMSDMSDMSD
Orthographic and grammatical mistakes 5.2 4.1 3.9 3.5 4.6 3.8 11.4 8.2 9.3 7.9 10.4 7.9
Formatting errors and typos 24.4 6.4 19.3 11.8 21.9 9.6 14.9 5.5 12.5 4.9 13.7 5.2
Amount of written text 231 57.4 270 67.3 250 64.1 314 16.7 312 24.6 313 20.5
Note - Orthographic and grammatical mistakes were counted as one mistake per word, even if a participant made more than one mistakes in a word. Each
formatting error and each typo was counted as one mistake. For instance, if a text contains three different font sizes each wrong formatted text section was
counted as one mistake.
doi:10.1371/journal.pone.0115069.t002
Document Preparation in Academic Research and Development
PLOS ONE | DOI:10.1371/journal.pone.0115069 December 19, 2014 9/14
disciplinary preferences fit with the original motivation for the development of
TeX (the basis of LaTeX) in the 1970s, which was to provide a powerful means to
typeset complex mathematical formulas [2]. Some computer scientists may
therefore think that mastering LaTeX is a ‘‘must’’ for any ‘‘true’’ expert in their
discipline and that someone who already invested significant time and effort in
learning LaTeX may not want to re-learn another tool. One may also argue that
given a well-designed LaTeX document class file, document development speed
and text and formatting accuracy are significantly improved. Another character-
istic of our study is that it is practically impossible to evaluate LaTeX without also
Fig. 4. Mean amount of text written within 30 minutes and the overall number of mistakes for the
continuous text for the four groups of participants (Word experts, Word novices, LaTeX experts, and
LaTeX novices). Error bars represent the standard error.
doi:10.1371/journal.pone.0115069.g004
Table 3. Results from the usability questionnaire ISO 9241-10.
Software
Word LaTeX
Usability questionnaire MSDMSD
Tiredness 3.4 1.9 2.2 1.4
Frustration 3.3 2.0 2.1 1.5
Enjoyment 3.6 1.7 5.2 1.4
Suitability for the task 0.6 1.1 1.4 0.8
Self-descriptiveness 20.2 0.9 20.3 1.2
Controllability 1.6 1.0 1.7 0.9
Conformity with user expectations 1.3 0.7 1.3 0.9
Error tolerance 0.3 1.1 20.6 1.2
Suitability for individualization 0.2 1.1 0.7 1.1
Suitability for learning 0.4 1.1 20.3 0.8
doi:10.1371/journal.pone.0115069.t003
Document Preparation in Academic Research and Development
PLOS ONE | DOI:10.1371/journal.pone.0115069 December 19, 2014 10 / 14
evaluating the used editors. In fact, our research measured the efficiency of Word
against LaTeX in combination with some editor interfaces. However, recent
research shows that it is possible to improve the interfaces to LaTeX by making
Fig. 5. Mean amount of text written within 30 minutes and the overall number of mistakes for the, table
text for the four groups of participants (Word experts, Word novices, LaTeX experts, and LaTeX
novices). Error bars represent the standard error.
doi:10.1371/journal.pone.0115069.g005
Fig. 6. Mean amount of text written within 30 minutes and the overall number of mistakes for the
equation text for the four groups of participants (Word experts, Word novices, LaTeX experts, and
LaTeX novices). Error bars represent the standard error.
doi:10.1371/journal.pone.0115069.g006
Document Preparation in Academic Research and Development
PLOS ONE | DOI:10.1371/journal.pone.0115069 December 19, 2014 11 / 1 4
them do more what the authors expect instead of what the programmers imagined
[3].
However, our study suggests that LaTeX should be used as a document
preparation system only in cases in which a document is heavily loaded with
mathematical equations. For all other types of documents, our results suggest that
LaTeX reduces the user’s productivity and results in more orthographical,
grammatical, and formatting errors, more typos, and less written text than
Microsoft Word over the same duration of time. LaTeX users may argue that the
overall quality of the text that is created with LaTeX is better than the text that is
created with Microsoft Word. Although this argument may be true, the
differences between text produced in more recent editions of Microsoft Word and
text produced in LaTeX may be less obvious than it was in the past. Moreover, we
believe that the appearance of text matters less than the scientific content and
impact to the field. In particular, LaTeX is also used frequently for text that does
not contain a significant amount of mathematical symbols and formula. We
believe that the use of LaTeX under these circumstances is highly problematic and
that researchers should reflect on the criteria that drive their preferences to use
LaTeX over Microsoft Word for text that does not require significant
mathematical representations.
One decision criterion that factors into the choice to use a particular software
system is the usability of the available systems for the given task. The usability of a
software system is a measure of how easy it is to use the program to carry out a
prescribed task. In human-computer interaction and cognitive ergonomics, the
most central aspects of usability include the ‘‘efficiency’’ of the system (which
refers to how quickly users can perform tasks once they have learned the design),
‘‘errors’’ (which refers to how many errors users make, the severity of these errors,
and how easily users can recover from these errors), and ‘‘user satisfaction’’ (the
overall pleasantness and feasibility of the design) [4,5]. Based on these criteria,
our results show that no reasons exist to use LaTeX for documents that do not
contain complex mathematical formula.
A second decision criterion that factors into the choice to use a particular
software system is reflection about what drives certain preferences. A striking
result of our study is that LaTeX users are highly satisfied with their system despite
reduced usability and productivity. From a psychological perspective, this finding
may be related to motivational factors, i.e., the driving forces that compel or
reinforce individuals to act in a certain way to achieve a desired goal. A vital
motivational factor is the tendency to reduce cognitive dissonance. According to
the theory of cognitive dissonance, each individual has a motivational drive to
seek consonance between their beliefs and their actual actions. If a belief set does
not concur with the individual’s actual behavior, then it is usually easier to change
the belief rather than the behavior [6]. The results from many psychological
studies in which people have been asked to choose between one of two items (e.g.,
products, objects, gifts, etc.) and then asked to rate the desirability, value,
attractiveness, or usefulness of their choice, report that participants often reduce
unpleasant feelings of cognitive dissonance by rationalizing the chosen alternative
Document Preparation in Academic Research and Development
PLOS ONE | DOI:10.1371/journal.pone.0115069 December 19, 2014 12 / 14
as more desirable than the unchosen alternative [6,7]. This bias is usually
unconscious and becomes stronger as the effort to reject the chosen alternative
increases, which is similar in nature to the case of learning and using LaTeX.
A third decision criterion that should factor into a researcher’s choice of a
document preparation system is the cost of research and development to the
public or industry. Researchers have a responsibility to act economically and
efficiently to create new technologies and theories that benefit society, especially in
cases in which research is publicly funded. In 2010, the 27 countries of the
European Union invested approximately 247 billion euros into research and
development, which represents approximately 1.9 percent of the EU’s gross
domestic product. In the same year, Germany invested 2.9% (75 billion euros)
and the US invested 2.8% (370 billion dollars) of its gross domestic expenditures
into research and development. A significant portion of these budgets is allocated
to the salaries of researchers. According to the American Association for the
Advancement of Science (AAAS), there are approximately 5.8 million science and
engineering researchers worldwide [8]. No reliable data is available about how
many of these researchers use LaTeX (or MS Word). However, a google search for
LaTeX (together with TeX to avoid ambiguities) results in approximately 18
million hits. Brischoux and Legagneux found that approximately 26% of
submissions to 54 randomly selected scholarly journals from 15 different scientific
disciplines were written in LaTeX, with a significant difference between LaTeX-
using and non-LaTeX-using disciplines [1]. We can only roughly estimate the
average number of hours per day that a researcher spends on writing scholarly
texts, such as internal technical reports, journal articles, and book publications.
For researchers in the field of cognitive and brain science, researchers may spend
approximately 10 to 30 percent of their time engaged in writing.
Given these numbers it remains an open question to determine the amount of
taxpayer money that is spent worldwide for researchers to use LaTeX over a more
efficient document preparation system, which would free up their time to advance
their respective field. Some publishers may save a significant amount of money by
requesting or allowing LaTeX submissions because a well-formed LaTeX
document complying with a well-designed class file (template) is much easier to
bring into their publication workflow. However, this is at the expense of the
researchers’ labor time and effort. We therefore suggest that leading scientific
journals should consider accepting submissions in LaTeX only if this is justified by
the level of mathematics presented in the paper. In all other cases, we think that
scholarly journals should request authors to submit their documents in Word or
PDF format. We believe that this would be a good policy for two reasons. First, we
think that the appearance of the text is secondary to the scientific merit of an
article and its impact to the field. And, second, preventing researchers from
producing documents in LaTeX would save time and money to maximize the
benefit of research and development for both the research team and the public.
Document Preparation in Academic Research and Development
PLOS ONE | DOI:10.1371/journal.pone.0115069 December 19, 2014 13 / 14
Supporting Information
S1 Materials. Data File of the study. Explanations are provided in S2 Materials.
doi:10.1371/journal.pone.0115069.s001 (XLSX)
S2 Materials. Explanations of data file including variable definitions.
doi:10.1371/journal.pone.0115069.s002 (TXT)
S3 Materials. Additional information.
doi:10.1371/journal.pone.0115069.s003 (PDF)
Acknowledgments
The text processing software Microsoft Word is licensed and traded by the
Microsoft Corporation. The authors do not have any connection to this company,
and the Microsoft Corporation had no role in the study design, data collection,
data analysis, decision to publish, or preparation of the manuscript. We thank all
participants for contributing to this study and Peter Flynn and Gerhard Strube for
many helpful suggestions for improving the revision of this article.
Author Contributions
Conceived and designed the experiments: MK JN. Performed the experiments: JN.
Analyzed the data: JN. Contributed reagents/materials/analysis tools: JN. Wrote
the paper: MK.
References
1. Brischoux F Legagneux P (2009) Don’t Format Manuscripts. The Scientist, 23: 24–24.
2. Knuth DE (1996) The TeXBook. Mu¨ nchen: Addison-Wesley.
3. Flynn P (2014) Human Interfaces to Structured Documents: The usability of software for authoring and
editing". PhD Thesis, University College Cork, Ireland. Available: http://epu.ucc.ie/theses/pflynn/
Accessed 2014 Nov 20.
4. Nielsen J (1993) Usability Engineering. San Diego: Academic Press.
5. Rubin J (1994) Handbook of Usability Testing. New York: John Wiley & Sons.
6. Festinger L (1985) A theory of cognitive dissonance. Stanford, CA: Stanford University Press (first
published 1957).
7. Brehm JW (1956) Postdecision changes in the desirability of alternatives. J Abnorm Soc Psych 52:
384–389.
8. OECD (2013) Gross domestic expenditure on R&D, Science and Technology. Available: DOI: 10.1787/
rdxp-table-2013-1-en. Calculated using full-time equivalents. Accessed 2014 Nov 20.
Document Preparation in Academic Research and Development
PLOS ONE | DOI:10.1371/journal.pone.0115069 December 19, 2014 14 / 14
... • tips for scientific writing [11][12][13][14][15][16][17] and collaboration tools [18][19][20][21]; and ...
... Solitary and collaborative [18,20] writing may use a different format/platform than the one used for formatting and finalizing the submission, e.g., one may collaborate via a Google Doc or via Markdown files in a Git repository followed by finalizing the layout in LibreOffice Writer or LaTeX. Moreover, input from one or different authors needs to be unified also on a technical level, regardless of the technical platform used. ...
... Readers should understand that writing a manuscript using literate programming tools is more challenging and may be somewhat tedious in comparison to using GUI wordprocessing and/or statistical programs. Moreover, the process is probably slower [35]. Not all scientists can or should be expected to become proficient in the use of Markdown, R, or Python in order to share their scientific findings. ...
Article
Full-text available
Introduction With the rising complexity of modern multimarker analytical techniques and notable scientific publication retractions required for erroneous statistical analysis, there is increasing awareness of the importance of research transparency and reproducibility. The development of mature open-source tools for literate programming in multiple langauge paradigms has made fully-reproducible authorship possible. Objectives We describe the procedure for manuscript preparation using RMarkdown and the R statistical programming language with application to JMSACL or any other Elsevier journal. Methods An instructional manuscript has been prepared in the RMarkdown markup language with stepwise directions on preparing sections, subsections, lists, tables, figures and reference management in an entirely reproducible format. Results From RMarkdown code, a submission-ready PDF is generated and JMSACL-compatible LaTeX code is generated. These can be uploaded to the Editorial Manager. Conclusion A completely reproducible manuscript preparation pipeline using the R and RMarkdown is described.
... Similarly, code-based programming has shown to be challenging and time-consuming since it has a steep learning curve. In a recent study, Knauff and Nejasmic [81] have suggested that scientific journal submission using Latex should be limited only to articles with a considerable amount of mathematics. ...
Article
This paper is an adaptation of the article "Which Are the Tools Available for Scholars? A Review of Assisting Software for Authors during Peer Reviewing Process", which provides a list of 220 software tools useful for academics during the process of writing, editing, publishing, and reviewing scientific articles. The present adaptation was made with the intention of making the list available to Spanish-speaking authors in order to enrich the experience of researchers in preparing their manuscripts for publication, saving them time and improving the quality of their work. The authors of the original work into the following categories divided the tools mentioned above: (I) Identification and social media, (II) Academic search engines, (III) Journal- abstract matchmakers, (IV) Collaborative text editors, (V) Data visualization and analysis tools, (VI) Reference management, (VII) Proofreading and plagiarism detection, (VIII) Data archiving, and (IX) Scientometrics and Altmetrics. The methodology for the collection of these tools can be found in the original work.
Article
Astronomy papers are most frequently composed and submitted in either Microsoft Word or LaTeX format. While most journals accept submissions in Word format (and some, such as Meteoritics and Planetary Science, until recently required Word format), the majority of astronomy papers posted to the arXiV are submitted in LaTeX format. The wisdom of using LaTeX is sometimes questioned: for instance, a recent study demonstrated that a set of 40 researchers and graduate students were able to transcribe text and tables more quickly and with fewer errors using Word rather than LaTeX, although LaTeX proved similarly efficient when it came to transcribing equations and was more enjoyable to use. However, the study had all participants reproduce a given source text. We suspect that the use of Word may influence an author to include fewer equations when creating an original work. This is a difficult hypothesis to test directly, so we instead use a subset of arXiV submissions to probe whether the use of LaTeX is correlated with the number of standalone equations in a manuscript. We find that PDF-only manuscripts on the arXiV have roughly half as many equations as those with LaTeX source files.
Chapter
Typesetting 2-dimensional mathematical notation can present challenges to users who rely upon WYSIWYG (what you see is what you get) word processing editors, which allow for direct manipulation of text. These editors use various models to represent 2-dimensional mathematical structure within the 1-dimensional word processing environment. The 2-dimensional nature of mathematical notations manifests itself in a variety of ways, and we hypothesize that two distinct models, structure-based (Microsoft Word Equation Editor) and free-form (MC2: Mathematics Classroom Collaborator) handle different types of mathematical structure with varying degrees of success. To test this hypothesis, an eye-tracking study was conducted to compare how these two models affect task efficiency for mathematical “expert” and “novice” users, as well as working memory interference, and cognitive load. The study required users to transcribe mathematical expressions containing three types of structure: linear (1-dimensional), exponential, and rational (fractions). Handwriting was used as a control. Results showed superior performance by the structure-based model for the transcription of fractions, while the free-form model displayed ameliorated performance for the transcription of exponents. Handwriting was found to be significantly more efficient, but cognitive effects were inconclusive. Few differences were found with respect to user mathematical experience level. These findings show evidence that neither of these models is superior for the typesetting of all mathematics, but rather that features of each model are better equipped to handle different mathematical structures. Therefore, word processing editors can improve the facility of typesetting 2-dimensional mathematics by incorporating elements of both tested models to improve the overall user experience. Such optimization will ultimately facilitate the digital learning and communication of mathematical content.
Article
Full-text available
Deep learning is transforming most areas of science and technology, including electron microscopy. This review paper offers a practical perspective aimed at developers with limited familiarity. For context, we review popular applications of deep learning in electron microscopy. Following, we discuss hardware and software needed to get started with deep learning and interface with electron microscopes. We then review neural network components, popular architectures, and their optimization. Finally, we discuss future directions of deep learning in electron microscopy.
Preprint
Full-text available
Deep learning is transforming most areas of science and technology, including electron microscopy. This review paper offers a practical perspective aimed at developers with limited familiarity. For context, we review popular applications of deep learning in electron microscopy. Following, we discuss hardware and software needed to get started with deep learning and interface with electron microscopes. We then review neural network components, popular architectures, and their optimization. Finally, we discuss future directions of deep learning in electron microscopy.
Preprint
Full-text available
Deep learning is transforming most areas of science and technology, including electron microscopy. This review paper offers a practical perspective aimed at developers with limited familiarity. For context, we review popular applications of deep learning in electron microscopy. Following, we discuss hardware and software needed to get started with deep learning and interface with electron microscopes. We then review neural network components, popular architectures, and their optimization. Finally, we discuss future directions of deep learning in electron microscopy.
Article
Female Ss were asked to rate each of eight articles on desirability, choose between two of them and rate each of the articles again. In addition, some Ss were exposed to a mixture of good and bad information about the choice alternatives after the choice was made. The results support a prediction that choosing between alternatives would create dissonance and attempts to reduce it by making the chosen alternative more desirable and the unchosen alternative less desirable. A second prediction, that dissonance and consequent attempts to reduce it would be greater, the more closely the alternatives approached equality, also received support.
Human Interfaces to Structured Documents: The usability of software for authoring and editing
  • P Flynn
Flynn P (2014) Human Interfaces to Structured Documents: The usability of software for authoring and editing". PhD Thesis, University College Cork, Ireland. Available: http://epu.ucc.ie/theses/pflynn/ Accessed 2014 Nov 20.