See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/326265014

Effects of visual signaling in screenshots: an eye tracking study

Article  in  Technical Communication (Washington) · November 2019

CITATIONS READS
4 497

1 author:

Michael Meng
Hochschule Merseburg
33 PUBLICATIONS   691 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Optimizing API documentation View project

All content following this page was uploaded by Michael Meng on 21 November 2019.

The user has requested enhancement of the downloaded file.

 Applied Research

Effects of Visual Signaling in Screenshots:

An Eye Tracking Study
By Michael Meng

Abstract Purpose: Screenshots are an important means of visualization in software

documentation. One question technical communicators need to address when dealing
with screenshots is whether visual signaling elements, such as arrows or frames, should
be added in order to highlight relevant information. This article reports the results of
an experimental study that examined whether signaling elements successfully guide
visual attention of readers to relevant screenshot information as intended. A second
goal was to find out whether visual signaling has a positive impact on how accurate and
fast users execute the tasks which the screenshots support.
Method: Two versions of a software tutorial were constructed that included screenshots
with or without signaling elements. Participants’ eye movements were recorded while
they studied the tutorial and executed the tasks described therein. In addition to eye
movement measures, accuracy of task execution and time to complete the tasks were
determined as measures of overall success on the tasks.
Results: Participants working with tutorials that used visual signaling executed more
tasks correctly. No differences were found regarding the time needed to complete
the tasks. Analysis of the eye tracking data showed that participants fixated relevant
screenshot areas longer and more often if highlighted by signaling elements.
Conclusions: The results provide evidence that adding signaling elements to
screenshots is an effective means to guide the visual attention of users. As predicted by
the Cognitive Theory of Multimedia Learning, visual signaling does not simply increase
interest in pictures but helps users to select relevant information.
Keywords: screenshots, software tutorials, visual signaling, eye tracking

Practitioner’s • Research on factors that modulate • More research is needed to determine

the effects of screenshots on user which signaling techniques are
Takeaway: performance can help technical effective and whether effects of
communicators to make informed signaling interact with other factors
decisions regarding screenshot usage such as task complexity and user
and design. experience. Collecting eye movements
• The article provides empirical while users can freely switch their
evidence that adding signaling attention between reading and acting
elements to screenshots helps the can help to address these questions.
user to select relevant information
from screenshots and improves user
performance.

396 Technical Communication l Volume 66, Number 4, November 2019

 Applied Research

Michael Meng

Introduction elements do not add content. Their main function is

to direct the attention of the user to content that is
Planning and creating graphics that assist the user relevant in the context of a certain task or goal.
in executing a procedure correctly and efficiently is Two types of signaling have been distinguished
a routine task of technical communicators. In the in the literature: verbal signaling and visual signaling
area of software documentation, the type of graphic (Mayer, 2009). Verbal signaling includes cues that are
most commonly used for this purpose is screenshots. added to a text, such as headings, highlighted words, or
Screenshots provide an efficient means to visualize outlines presented before a text. Headings, highlighted
the state of the graphical user interface of a software words, and outlines help users to understand the
system at a certain point, such as the initial state organization of a text. They are added to direct the
from which a procedure starts, the overall goal state, user’s attention to particular terms, concepts, or
or intermediate states that result if the procedure is propositions that the author regards important. In
carried out successfully (Farkas, 1999; van der Meij & technical documentation, verbal signaling is also
Gellevij, 2004). commonly used to indicate the function that a
When adding screenshots to software manuals, particular unit of text serves. For example, numbers or
online help systems or tutorials, technical bullets may be used to identify a text unit as an action
communicators have to address several important step, and check marks could be used to identify a text
questions (van der Meij & Gellevij, 1998). A first set unit as a result statement.
of decisions concerns when screenshots should be used Visual signaling, on the other hand, refers to cues
and which function each individual screenshot should that are added to pictures, such as frames, arrows,
serve. For example, when developing a procedure magnifiers, or distinctive colors. Frames, arrows, and
consisting of several action steps, decisions have to be similar signaling techniques are added to pictures to
made whether none, some, or all action steps should be direct the visual attention of users to specific areas of
supported by a screenshot, and whether the individual the picture containing important information. The goal
screenshots should help the user to locate and identify of the current study is to investigate the effect of visual
user interface elements (e.g., by depicting a menu signaling in screenshots contained in procedures on
entry or to help the user to compare the current state user behavior. More specifically, the study asks whether
of the user interface with the intended goal state, e.g., signaling indeed guides visual attention, and whether
by depicting the screen that results from carrying out this “guiding effect” improves user performance.
an action step). Knowing whether and why users benefit from
Once the decision to use a screenshot at a certain visual signaling in screenshots is not only of interest
point has been made, several additional decisions from a theoretical point of view but also if an applied
follow that relate to screenshot design (van der Meij, perspective is taken (Martin-Michiellot & Mendelsohn,
2000), such as which portion of the user interface 2000). Adding frames, arrows, or similar cues to
the screenshot should depict or where the screenshot screenshots is associated with a specific cost, which
should be positioned. This article focuses on the effects reflects the effort it takes to implement the signaling
of a design decision that is particularly important: the elements. This cost increases considerably if not only
decision of whether signaling elements should be added the initial efforts in designing a screenshot are taken
to a screenshot or not. into account but also subsequent efforts related to
Signaling refers to a set of cueing techniques that documentation maintenance and localization. In order
have been discussed extensively both in the domain to decide whether these efforts are justified, technical
of text design and text comprehension (Lorch, 1989; communicators have to know whether visual signaling
Spyridakis, 1989) as well as in multimedia learning works as intended. The present study is intended to
(Mautone & Mayer, 2001; Koning, Tabbers, Rikers, & provide arguments on which such a decision can be
Paas, 2009; Moreno, 2007; Richter, Scheiter, & Eitel, based. It contributes the results of research that—in
2016). Common to all signaling techniques is that they the sense of Boekelder and Steehouder (1999)—was
are added to material in order to help the user know designed to serve practice by studying the functional
how to process the material (Mayer, 2009). Signaling relation between a design aspect (visual signaling in

Volume 66, Number 4, November 2019 l Technical Communication 397

 Applied Research

Effects of Visual Signaling in Screenshots

screenshots) and reading behavior of documentation by screenshots faster. Gellevij, van der Meij, de Jong,
users. The study asks whether a specific design change and Pieters (2002) investigated whether screenshots
(adding visual signaling) changes reading behavior also support the development of a mental model of
in the intended way, which is that users pay more the user interface, again contrasting a manual version
attention to relevant information on a screenshot and containing screenshots with a text-only version. Users
ignore irrelevant information. In addition, the study working with the screenshot manual completed training
asks whether this change in reading behavior results faster and scored higher on tasks designed to measure
in documentation that enables users to execute a learning effects. Like van der Meij (1996), both studies
procedure more effectively. provide important hints regarding global effects of
By focusing on user behavior, the type of research screenshots on task execution and learning. However,
reported here is closely related to efforts to make since all studies discussed so far directly contrasted
documentation more usable, which is a fundamental a visual manual version with a text-only version,
concern of technical communicators (Guillemette, it remains open why these effects arise and what
1989; Redish, 2010; Alexander, 2013; Arachya, 2017). contributions individual design variables such as visual
However, unlike standard usability testing, the aim signaling make.
is not to find out whether a particular information While the studies discussed so far clearly
product is effective or not or where potential trouble demonstrate that including screenshots in manuals
spots are. Rather, the goal is to reveal more fundamental can have a positive effect on task performance and
principles of communication whose consequences learning, other studies failed to find such effects
can be applied to information products containing (van der Meij & Gellevij, 2002; Nowaczyk & James,
screenshots in general. 1993), suggesting that effects induced by screenshots
are subject to boundary conditions. For example,
Prior Research on Screenshots the results reported in Sweller and Chandler (1994)
The question whether screenshots effectively support provide preliminary evidence that screenshots only
users of software documentation has been addressed support user performance if the task to be solved is
in a series of studies by Hans van der Meij, Mark sufficiently complex. The critical role of design variables
Gellevij, and colleagues conducted in the late 1990s as possible boundary condition for screenshot effects
and early 2000s (see Gellevij & van der Meij, 2004, has been emphasized in a different line of studies
and van der Meij, Karreman, & Steehouder, 2009, for by van der Meij, Gellevij, and colleagues. Gellevij,
overviews). One of the first experiments demonstrating van der Meij, de Jong, and Pieters (1999) addressed
that screenshots may positively affect task execution screenshot coverage as design variable. Their study
was reported in van der Meij (1996). The study compared a text-only version of a manual with two
tested users on two versions of a training manual for types of visual manuals that differed with respect to the
a database application. The study found that users of screenshot area depicted by screenshots: a visual manual
the manual version containing screenshots completed version in which screenshot coverage was restricted
the training tasks significantly faster compared to the to elements relevant in the context of a specific task,
group working with a text-only control version. No and a visual manual version containing screenshots
effect of screenshots on the accuracy of task execution covering the full screen. Overall, the results suggest
was found. that coverage had an effect: Participants working
Evidence confirming this finding was reported with a partial-screenshot manual performed worse on
by several later studies which focused more closely several measures, such as performance on untrained
on individual screenshot functions. For example, tasks, compared to participants working with a manual
Gellevij and van der Meij (2004) examined the effect containing screenshots that captured the full screen.
of screenshots that were intended to support users Interestingly, participants using text-only manuals also
comparing the current state of the software user tended to outperform participants using manuals with
interface with the intended goal state. They report that partial screenshot.
users working with a manual containing screenshots In addition to screenshot coverage, van der
made fewer errors and executed actions steps supported Meij (2000) also tested effects of the positioning

398 Technical Communication l Volume 66, Number 4, November 2019

 Applied Research

Michael Meng

of screenshots relative to related text segments by can take in visual information (Cooke, 2005; Goldberg
comparing different layout variants that arranged & Wichansky, 2003). Saccades refer to the rapid
screenshots and text segments in adjacent columns. The movements which the eyes must perform in order to
study replicated the finding that full-screen screenshots fixate on another position in the visual field. Based
lead to better performance than partial-screen on the observed pattern of fixations and saccades,
screenshots, but only if the instructions were positioned it is possible to infer which information observers
in the left column and the screenshots to the right. attend to (e.g., while reading, examining a webpage,
Similarly, Martin-Michiellot and Mendelsohn (2000) or interacting with a software system). Research using
included a manual version with juxtaposed screenshots eye tracking is based on the general assumption that
and a manual version in which text elements and the information a user visually attends to correlates
corresponding screenshot areas were integrated more highly with the thoughts or mental activities the user
closely by using callout lines. In addition, this study is involved with (“eye-mind hypothesis,” Goldberg &
also manipulated the complexity of the tasks that Wichansky, 2003, p. 507).
subjects had to perform for test purposes. Although Eye tracking has been used successfully for
both types of visual manuals accelerated training time many years to study reading, text comprehension,
compared to a text-only control version, there was no scene perception, or multimedia learning (Rayner &
significant difference between juxtaposed and integrated Pollatsek, 2006; Rayner, 2009; Mayer, 2010). Eye
screenshot conditions on task performance, neither for tracking has also become a prominent method to
simple nor more complex tasks. explore various aspects of Web usability, such as screen
Taken together, prior research clearly backs the use design, navigation architecture, or processes of visual
of screenshots in software documentation. This research search (Cooke, 2005; Cooke et al., 2008; Duchowski,
also points to conditions that screenshots have to 2017; Nielsen & Pernice, 2010). Moreover, eye tracking
meet in order to be effective. However, whether visual can make verbalizations obtained from participants in
signaling contributes to effects of screenshots on user concurrent think-aloud tests more informative (Cooke,
performance is still an open question. 2010; Elling, Lentz, & de Jong, 2012).
The current study uses eye tracking to study the
Background on Eye Tracking impact of a specific design decision (adding signaling
Besides addressing an issue that is still unexplored, the elements to screenshots) on which information users
current study also extends prior research on screenshots attend to when working with technical documentation
by using eye tracking. The eye tracking method makes to solve a problem. At a more general level, the
it possible to determine with a high degree of accuracy study can be viewed as an attempt to extend the
where people look at a certain point in time (Holmqvist range of applications for eye tracking in technical
et al., 2011; Williams et al., 2005). To determine communication research and to explore a new
where people look, special eye tracking hardware is way in which this method can be used to further
used which measures the gaze position of a user in very our understanding of how people read and apply
short intervals continuously over time. The eye tracker technical information.
used in the current study—the RED 250mobile system
from SensoMotoric Instruments (SMI)—recorded Research Objectives and Predictions
the gaze position at a rate of 120 Hz, which means To sum up, the research reported here was designed
that a gaze position measure was taken roughly every to address two objectives. One objective was to test
8 milliseconds. Based on this continuous data stream, whether visual signaling indeed guides visual attention
analysis software can then model various types of eye to relevant information in a screenshot, or whether
movement events. signaling more generally increases interest in pictures.
The eye movement events of relevance to the A second objective was to find out whether visual
current study (and for most studies in the field of signaling has a positive impact on the accuracy and
usability) are fixations and saccades. Fixations are efficiency of task execution.
intervals of 200–500 milliseconds in which the gaze The hypotheses guiding the current study are based
position remains rather stable and in which the eyes on the Cognitive Theory of Multimedia Learning

Volume 66, Number 4, November 2019 l Technical Communication 399

 Applied Research

Effects of Visual Signaling in Screenshots

(CTML; Mayer, 2009; Moreno & Mayer, 2007), the current study, I therefore take CTML to motivate
which attempts to model the process of learning the prediction that visual signaling supports the user
from material containing text and pictures. CTML in allocating more visual attention to relevant areas
distinguishes three types of processes that compete of a screenshot. Hence, this theory leads us to expect
for working memory capacity when learning from that more fixations and longer fixation times should
multimedia material. CTML uses the terms “essential be observed in a relevant screenshot area if signaling is
processing” and “generative processing” to refer to used. Since signaling reduces working memory capacity
processes which are necessary for learning to take place, required for extraneous processing, making this capacity
such as selecting information for processing in working available for other cognitive processes involved in
memory, developing verbal and pictorial representations solving the current task, I also derive the hypothesis
for incoming information, as well as integrating verbal that visual signaling improves overall task performance.
and pictorial representations with information that
already exists in long-term memory. Besides essential Method
and generative processing, working with multimedia
material may also trigger processes which are unrelated
to learning and which result from the way multimedia Materials
material is presented to the learner, such as efforts For the test, I developed a short tutorial that described
to identify relevant information in texts or pictures, how to add a colored picture frame around a digital
or to connect text and picture elements that contain image using GIMP (www.gimp.org), an open source
related information. These processes are referred to as image manipulation program similar to Adobe
“extraneous processing.” Photoshop. The tutorial was created in German, as the
According to CTML, signaling helps the learner participants were native speakers of German (see section
because it reduces extraneous processing (Mayer, 2009, Participants below).
p. 108ff.). By reducing extraneous processing, signaling The tutorial consisted of 12 pages. It started with a
frees up working memory capacity which then becomes short introduction (2 pages) which explained the goal of
available for cognitive processes related to learning. the tutorial, illustrated the initial state and the intended
Several studies have already demonstrated that both result, and described relevant elements of the user
verbal and visual signaling can foster learning, and more interface. The process to create the picture frame was
specifically, can successfully direct the attention of the divided into 8 sub-tasks (procedures) that contained 2
learner to relevant information (see Mayer, 2009, and to 5 action steps each. The tutorial was completed by a
Richter et al., 2016, for reviews). So far, however, this page that congratulated the participants for completing
research has focused on “reading-to-learn” scenarios the tutorial and again displayed the intended result. An
in the sense of Redish (1989), e.g., scenarios in which example illustrating the tasks is given in Figure 1. The
learners study material in order to understand how task describes how to apply one of the filters offered by
lightning is formed or how a pump works. However, GIMP in order to create a blur effect on a layer that
procedures contained in manuals, online help, or will later be part of the picture frame. The four action
tutorials are often not read with the intention to learn steps instruct the reader to select the correct layer from
the series of action steps described there but with the a list, to launch the filter, to add values that control how
goal to directly execute the steps in order to accomplish intense the blur effect will be, and to apply the settings
a certain goal, a scenario which Redish (1989) by clicking OK.
characterizes as “reading-to-do.” In each task, one or two of the action steps were
Only few attempts have been made so far to supported by screenshots. Screenshots were inserted at
investigate whether results obtained by research on points that were judged by the author to potentially
multimedia learning in “reading-to-learn” scenarios lead to errors. Screenshots were intended to help
generalize to procedural text that is “read-to-do,” but at participants to locate and identify user interface
least preliminary evidence is available suggesting that it elements, and to compare the current state of the user
does (Irrazabal, Saux, & Burin, 2016; van Genuchten, interface with the intended goal state. Following design
Hooijdonk, Schüler, & Scheiter, 2014). With respect to recommendations in Gellevij et al. (2002), screenshot

400 Technical Communication l Volume 66, Number 4, November 2019

 Applied Research

Michael Meng

size and coverage were optimized depending on for which values had to be checked or changed, buttons
screenshot function. Each tutorial task was designed to on which participants were to click or elements to select
fit on a single screen page, except one task which had from a list. For example, in the task shown in Figure
to be distributed over two pages due to the size of the 1, an arrow was added to the screenshot following step
screenshots used. 1 to help the user identify which of the two elements
To address the research questions of the current to select from a list of available layers. The screenshot
study, I created two versions of the tutorial that following step 3 uses a frame to emphasize the fields into
contained identical instructions but that differed with which values have to be entered to control how intense
respect to whether the screenshots used visual signaling the blur effect will be. The second tutorial version
elements or not. The first version (condition “signaled”) (condition “nonsignaled”, see Figure 2) contained the
included screenshots to which arrows, frames or same screenshots, but without signaling elements.
magnifiers were added in order to highlight relevant Participants were instructed to work though the
information, such as fields in the GIMP user interface tutorial and to execute the tasks described therein.
Special care was taken to create a task setting that was
as natural as possible and encouraged normal reading.
Therefore, the tutorial was presented on a computer
screen adjacent to the GIMP program (see Figure
3); hence, the tutorial and the GIMP software were
available to the user simultaneously.

Figure 1. Example page of the tutorial in condition “sig-

naled.” The task describes how to apply one of the filters
offered by GIMP in order to create a blur effect. The blur
effect is used on the layer that will later serve as part of the
picture frame. The steps instruct the reader to select the
correct layer from a list, to launch the filter, to add values that
control how intense the blur effect will be, and to apply the Figure 2. Example page of the tutorial in condition
settings by clicking OK. “nonsignaled”

Volume 66, Number 4, November 2019 l Technical Communication 401

 Applied Research

Effects of Visual Signaling in Screenshots

Procedure mode. For the test, the eye tracker was attached to a
The experiment was run in the usability lab of 24-inch monitor that served for stimulus presentation.
Merseburg University of Applied Sciences and Tracking rate was set to 120 Hz, and data were recorded
comprised two parts. Participants first received a short from both eyes simultaneously. The software package
questionnaire that contained four questions regarding ExperimentSuite Scientific 3.6 (also from SMI) was
age of participants, level of computer experience, general used for data recording (ExperimentCenter) and
experience with image manipulation software, and analysis (BeGaze).
specific experience with GIMP. After the questionnaire The tutorial was displayed as a series of JPEG
was completed, participants received instructions for pictures using the Windows 7 Photo Viewer. The
the test. They were then positioned in front of the test Windows 7 Photo Viewer was chosen because it
monitor and a calibration procedure was run to prepare features salient back and forward navigation buttons
the eye movement recording. Participants then started to that are easy to operate and that participants used to
work on the tasks described in the tutorial. Participants navigate through the document. The tutorials were
were instructed to work with the tutorials as they would presented alongside GIMP on the stimulus monitor
normally. No special emphasis was put on reading the using windows of fixed size and position. The tutorial
tutorials. Statement of consent was obtained from the was presented to the right of GIMP and occupied about
participants at the beginning of each session. one third of the monitor area. The remaining monitor
While participants worked through the tutorial, area was occupied by GIMP (see Figure 3).
their eye movements were recorded using an RED During each session, the SMI software recorded a
250mobile eye tracker. The RED 250mobile is a screen video onto which gaze data were mapped later
video-based eye tracker that operates in head-free for analysis purposes.

Figure 3. Positioning of tutorial and GIMP on the stimulus presentation monitor

402 Technical Communication l Volume 66, Number 4, November 2019

 Applied Research

Michael Meng

Participants measures to examine whether or not visual signaling

32 students (14 female, 18 male) were recruited effectively guides visual attention.
from Merseburg University of Applied Sciences as
participants for the experiment. Participants were native Accuracy on tasks
speakers of German and unaware of the purpose of the To determine the accuracy of task execution, I coded for
experiment. All participants had normal or corrected- each of the 8 tasks whether participants had executed
to-normal vision. The participants were randomly the task correctly or not. Task execution was scored as
assigned to one of the two conditions, resulting in correct (coded by value “1”) if participants correctly
16 participants per condition. Mean age was 27.6 executed all action steps required by the task. If one or
years (median= 27, min = 21, max = 42). Mean more of the action steps were not executed as described
experience with image manipulation programs rated in the tutorial, the task was scored as incorrect (coded
on a Likert scale ranging from 1 (“very experienced”) by value “0”). Scoring was done by post-hoc inspection
to 6 (“no experience”) was 2.78 (median=3). Mean of the screen videos produced by the SMI software
experience with GIMP rated on the same scale was during the eye tracking session for each participant. All
5.40 (median=6), and general computer experience coding was done by the author, no additional coders
was 1.75 (median=2). In sum, participants were were involved.
experienced computer users with some experience with
image manipulation programs but no significant prior Time for task completion
exposure to GIMP. Rating data are summarized in Table The screen videos also formed the basis for analyzing
1 individually for the two conditions. Mann-Whitney the time participants needed to complete the tutorial.
U-tests on the rating data confirmed that there were Like task accuracy, time for task completion was
no significant differences between the groups regarding obtained on a per-task basis using the time stamps that
general computer experience (W=136, p=0.76), general were recorded automatically with the screen videos. For
experience with image manipulation programs (W=136, each task, the start time was defined by the time stamp
p=0.76), or experience with GIMP (W=132, p=0.87). of the first video frame on which the tutorial page for
the respective task was open. Likewise, the end time
Table 1. Participants’ self-assessments of general computer was defined as the time stamp of the last video frame,
experience, general experience with image manipulation immediately before participants opened the tutorial
programs and experience with GIMP. Means per experimen- page for the next task. Time for task completion for
tal condition with medians in parentheses. each task, then, is the difference between the time
stamp for end time and start time.
Rating Condition
signaled nonsignaled Fixation times and number of fixations
To determine the number of fixations and fixation times,
General computer experience 1.75 (1.5) 1.75 (2.0) two different types of so-called “areas of interest” (AOIs)
General experience with image 2.75 (2.5) 2.81 (3.0) were defined: “screenshot” and “relevant area.” The AOI
manipulation programs “screenshot” encompassed the screenshots on each page
Experience with GIMP 5.31 (6.0) 5.5 (6.0) of the tutorial. The AOI “relevant area” was embedded
into the AOI “screenshot.” This AOI marked the area
that was relevant in the context of the current task and
Dependent Variables was therefore highlighted in the condition “signaled.”
Four dependent variables were used to examine effects All AOIs were applied to the conditions “signaled” and
of visual signaling. Accuracy of task execution and the “nonsignaled,” and drawn manually using the BeGaze
time participants needed to complete the tasks were AOI editor.
intended as measures to inform about the participants’ Since the pages for tasks 1–8 of the tutorial
overall performance on the tutorial task. Fixation times differed regarding how many screenshots were used,
and number of fixations were selected as eye tracking as well as their size and position, special care had to

Volume 66, Number 4, November 2019 l Technical Communication 403

 Applied Research

Effects of Visual Signaling in Screenshots

be taken to ensure that the AOIs used on the different Mächler, Bolker, & Walker, 2015). Over the last
tutorial pages were identical in size and occupied decade, this technique has become a standard method
identical positions across the conditions “signaled” of analysis in psycholinguistic studies, both for the
and “nonsignaled.” I therefore first defined the AOIs analysis of continuous outcomes (such as fixation times
in the condition “signaled,” then exported the AOI in this study) and for the analysis of categorical data
definitions and afterwards reimported them to the data (such as task accuracy). There are various reasons why
set of condition “nonsignaled.” Positioning of AOIs is mixed-effects modeling has been adopted so widely,
illustrated in Figure 4, which also shows an individual which are discussed in detail in Baayen et al. (2008)
scan path pattern for one of the tutorial pages. and Jaeger (2008), among others (for a nontechnical
For identifying fixations and saccades, BeGaze was introduction, see Balling, 2018). I selected this
used with unmodified default settings for event detection. technique because, in contrast to alternative methods
Prior to analysis, data for fixation times and number of such as analysis of variance (ANOVA), mixed-effects
fixations were inspected visually for outliers. Since no regression modeling makes it possible to control for
clear outliers could be identified, no data were excluded. multiple random effects in a single analysis, such
as random variation that is due to the particular
Data Analysis selection of stimuli, or systematic differences between
All statistical analyses reported in this paper were participants. In the present study, random effects for
conducted using the statistics software R, version 3.4.2 tasks were included besides participants in order to
(R Core Team, 2017). The data were analyzed with capture variance, which is due to individual properties
mixed-effects regression modeling (Baayen, Davidson, of the 8 tutorial tasks and the screenshots they
& Bates, 2008) using the lme4 R package (Bates, contained, such as differences regarding screenshot size,

Figure 4. Positioning of AOIs “screenshot” and “relevant area” and sample scan path

404 Technical Communication l Volume 66, Number 4, November 2019

 Applied Research

Michael Meng

function or positioning as well as size and proportion To analyze the effect of visual signaling on how
of AOIs. A significant effect of a fixed factor in such a accurately participants solved the tutorial tasks, a
model therefore reflects differences that persist after the logistic mixed-effects analysis using the glmer function
variance introduced by the particular set of participants of the lme4 package was conducted, which is suited
and tasks has been identified and controlled for. for the analysis of binary dependent variables (Jaeger,
Regarding the procedure to follow when 2008). Since each participant contributed accuracy
constructing a mixed-effects model for confirmatory values for each of the 8 tutorial tasks, the analysis
hypothesis testing, several approaches have been was based on a total of 256 observations. The mixed-
proposed which differ mainly with respect to the effects model included the factor “screenshot” with
random effects structure that models should include levels “signaled” and “nonsignaled” as fixed effect.
(Barr, Levy, & Scheepers, 2013; Bates, Kliegl, Vasishth, Table 3 reports the parameter estimates of the model,
& Baayen, 2015; Matuschek, Kliegl, Vasishth, Baayen, the standard errors, the resulting z-values, and the
& Bates, 2017). I followed Matuschek et al. (2017) and associated probabilities. The analysis reveals that
Bates, Kliegl et al. (2015) who recommend using the accuracy on tasks was significantly higher for condition
most parsimonious model that can be assumed to have “signaled” compared to condition “nonsignaled.”
generated the observed pattern of results. To identify
the most parsimonious model, I first constructed a Table 3. Results of mixed-effects model for task accuracy
model with the maximum random effects structure Contrast Estimate Std. Error z value Pr(>|z|)
supported by the factorial design (Barr et al., 2013).
The model was then simplified in an iterative manner Formula: score_correct ~ screenshot + (1|task)
by identifying and removing parameters from the (Intercept) 3.0423 0.5915 5.143 < 0.001
random effects structure with the smallest variance “signaled” vs. 2.1184 0.5696 3.719 < 0.001
contribution. This process was continued as long as “nonsignaled”
the simpler model did not differ significantly from the
preceding model in terms of goodness of fit.
In the results tables below, the final models arrived Time for Task Completion
at are provided using the notation of (Bates, Mächler et As described above, time for task completion was
al., 2015) which specifies the dependent variable, the also determined on a by-task basis, again yielding
independent variable(s) entered as fixed effect(s) and— 8 observations for each of the 32 participants. The
using parentheses—the random effect terms. Note mean duration per task for each condition, along with
that the random effect terms may differ from analysis standard errors, is shown in Table 4.
to analysis, which is a consequence of the iterative
approach used here. Table 4. Mean time per task (in milliseconds) to complete the
tasks and standard errors
Results signaled nonsignaled
38864 (1054) 37846 (1441)
Accuracy on Tasks
Mean accuracy scores and standard errors for each of The data for time to task completion reveal a slight
the two tutorial conditions are provided in Table 2. As advantage for condition “nonsignaled” compared to
Table 2 shows, accuracy on tasks was higher in condition condition “signaled.” Task duration data were analyzed
“signaled” compared to condition “nonsignaled.” using the lmer function of the lme4 package, including
factor “screenshot” as fixed effect. Table 5 reports
Table 2. Mean accuracy scores as proportion of tasks solved the parameter estimates, the standard errors, and the
correctly (in %) and standard errors resulting t-values. Because exact probabilities cannot be
signaled nonsignaled computed for such models, I consider contrasts with a
t-value greater 2 as significant (Baayen et al., 2008).
96.88 (1.40) 82.03 (3.60)

Volume 66, Number 4, November 2019 l Technical Communication 405

 Applied Research

Effects of Visual Signaling in Screenshots

Table 5. Results of mixed-effects model for time to task AOI “relevant area” (see Figure 4). By definition, all
completion screenshot information not included within the AOI
Contrast Estimate Std. Error t-value
“relevant area” is considered irrelevant in the context of
the task at hand. Fixation times for irrelevant areas were
Formula: duration ~ screenshot + (1|subject) + (1|task) computed by subtracting all fixations that landed in the
(Intercept) 38355 3062 12.528 AOI “relevant area” from the fixations collected by the
“signaled” vs. -1019 2034 -0.501 AOI “screenshot,” which marked the entire screenshot.
“nonsignaled” Therefore, a 2x2 design resulted for subsequent analyses
including the factor “area” with factor levels “relevant”
and “irrelevant” in addition to the factor “screenshot”
As Table 5 shows, the difference between the means with levels “signaled” and “nonsignaled.”
is not significant. I therefore conclude that the time The mean fixation times for each of the four
participants needed to complete the tasks was not conditions are visualized in Figure 5 and are provided
dependent on whether they worked with a tutorial numerically along with factor means and grand mean as
containing screenshots with or without visual signaling. well as respective standard errors in Table 6.

Fixation Times Table 6. Mean fixation times per task with standard errors in
In order to determine whether signaling techniques, parentheses
such as colored frames or arrows, indeed direct visual screenshot area
attention to screenshot areas that are relevant in the
context of a task, a second factor was defined that relevant irrelevant mean
distinguishes relevant and irrelevant screenshot areas. signaled 1560 (112) 1414 (106) 1487 (77)
Relevant screenshot areas are highlighted by signaling
nonsignaled 828 (81) 1568 (131) 1198 (80)
and contain the information that readers should attend
to. In this study, these areas are designated by the mean 1194 (73) 1491 (84) 1343 (56)

For statistical analysis, the lmer function of

the lme4 package was used. The model included
“screenshot” and “area” as fixed effects. Results of
mixed-effects modeling are provided in Table 7.

Table 7. Mixed-effects model for fixation times

Contrast Estimate Std. Error t-value
Formula: fixation_time ~ area * screenshot + (1|subject)
+ (1 + area||task)
(Intercept) 1342.6 195.0 6.886
area 296.8 276.4 1.074
screenshot -289.0 184.0 -1.570
area:screenshot 887.1 176.4 5.028

As Table 7 shows, main effects were not significant,

but there was a significant interaction between “area”
and “screenshot.” To explore the interaction further,
Figure 5. Mean fixation times per task in relevant or irrele- I performed post-hoc Tukey HSD comparisons using
vant areas of the screenshots depending on the availability of the lsmeans R package (Lenth, 2016). The comparisons
visual signaling. Bars represent 95% confidence intervals. revealed a significant difference between the conditions

406 Technical Communication l Volume 66, Number 4, November 2019

 Applied Research

Michael Meng

“signaled” and “nonsignaled” for relevant screenshot The data resulting from mixed-effects modeling are
areas (1560 ms vs. 828 ms, t(45.21) = 3.59, p < .01), summarized in Table 9. As is evident from Table 9, the
whereas the difference was not significant for irrelevant analysis revealed a very similar pattern, which is not
areas (1414 ms vs. 1568 ms, t(45.21) = -0.757, p = surprising as there is a very strong overall correlation
0.45). I conclude that relevant screenshot areas indeed between fixation times and number of fixations
attract longer fixation times if visual signaling is used (r=0.93). The only difference is that the main effect
compared to when no visual signaling is used. of “area” now reached significance as well, but it is
qualified by a significant interaction.
Number of Fixations
Number of fixations was analyzed using the same 2x2 Table 9. Mixed-effects model for number of fixations
factorial design and the same approach to mixed-effects Contrast Estimate Std. Error t-value
modeling. The respective condition means are shown in
Formula: fixation_count ~ area * screenshot +
Figure 6 and provided along with factor means, grand
(1|subject) + (1 + area||task)
mean, and standard errors in Table 8.
(Intercept) 4.55 0.64 7.166
Table 8. Mean number of fixations per task with standard area 2.16 0.85 2.540
errors in parentheses screenshot -0.19 0.57 -0.328
screenshot area area:screenshot 2.58 0.52 4.916
relevant irrelevant mean Post-hoc comparisons show that the mean number
signaled 4.21 (0.26) 5.09 (0.30) 4.65 (0.20) of fixations is significantly different between the
conditions “signaled” and “nonsignaled” for relevant
nonsignaled 2.73 (0.25) 6.19 (0.47) 4.46 (0.29)
screenshot areas (4.21 vs. 2.73, t(43.79) = 2.34, p <
mean 3.47 (0.19) 5.64 (0.28) 4.55 (0.17) .05) and marginally significant for irrelevant screenshot
areas (5.09 vs. 6.19, t(43.79) = -1.75, p = 0.09). I
therefore derive the additional conclusion that relevant
screenshot areas highlighted by visual signaling are not
only fixated longer but also more often.

Discussion

Main Findings
The research reported here was designed to investigate
whether and how visual signaling affects user
performance in a “reading-to-do” scenario. More
specifically, I wanted to know whether visual signaling
in screenshots indeed directs visual attention of
documentation users to the signaled areas. Related to
that, I also examined whether directing visual attention
to relevant screenshot areas would improve performance
on executing the tutorial tasks.
Regarding overall user performance, I found a
significant effect of visual signaling on overall task
Figure 6. Mean number of fixations per task in relevant accuracy. Participants working with the tutorial
or irrelevant areas of the screenshots depending on containing screenshots with visual signaling made fewer
the availability of visual signaling. Bars represent 95% errors compared to participants in the “nonsignaled”
confidence intervals. condition. No reliable difference between conditions

Volume 66, Number 4, November 2019 l Technical Communication 407

 Applied Research

Effects of Visual Signaling in Screenshots

was found for the time participants needed to execute appropriately support the communicative function of
the tutorial tasks. the screenshot.
Consistent with the effect of visual signaling on Given that implementing and maintaining visual
task accuracy, I found evidence that signaling triggers signaling elements is associated with investments in
users to allocate more visual attention to screenshot terms of time and resources, and given further that
areas that are relevant in the context of a certain task technical communicators typically work under pressure
and therefore highlighted by arrows, frames or similar to reduce costs, the results of this study may provide an
techniques. If relevant screenshot areas are emphasized, additional argument that can be used to justify these
users fixate on them more often and longer compared costs: The costs are justified because they contribute
to the condition in which no signaling was used. At to making users more effective. In this sense, the study
the same time, I observed a tendency that irrelevant attempts to contribute to a line of research that helps
screenshot regions were fixated less often and with to validate design decisions by demonstrating that
overall shorter fixation times. Taken together, the they are grounded in empirical research supporting
results imply that participants do not generally look their supposed effects. This line of research has a long
more often or longer on a screenshot when visual tradition in the field of technical communication
signaling is used, but they look more often and longer (as shown, e.g., by van der Meij et al., 2009), and its
at the right place, as is intended. This pattern of effects relevance from a practitioner’s point of view has been
is consistent with accounts of visual signaling in the confirmed recently by Carliner, Coppola, Grady, and
Cognitive Theory of Multimedia Learning (Mayer, Hayhoe (2011), and St.Amant and Meloncon (2016).
2009). According to this account, visual signaling The effect of visual signaling on the accuracy of task
does not simply increase overall interest in pictures execution was statistically significant, but it was rather
but specifically helps the user to identify and select small numerically. Note that accuracy on tasks was fairly
relevant information from a picture, thereby reducing high in general and reached an almost perfect score in
the amount of working memory capacity that has to be condition “signaled,” which may have reduced the effect
devoted to extraneous processing. size. A follow-up study could use more complex tasks to
check whether the effect size increases with increasing
Practical Implication and Suggestions task complexity.
for Future Research Although the current study demonstrates the
When designing screenshots for software effectiveness of signaling, it does not allow conclusions
documentation, whether or not to use visual signaling regarding which specific signaling techniques are
is one of the important design decisions that technical effective in guiding the visual attention of the user and
communicators have to make. The main practical which aren’t. Another question that future research
implication of the current study is that the results— needs to address is whether the effect of visual signaling
if confirmed by future studies—back the use of is modulated by other factors, such as user experience
visual signaling in screenshots. The results suggest or reading goals of the user. As discussed above, prior
that signaling is a design dimension which supports research on screenshots suggests that such interactions
positive effects of screenshots on user performance. with other factors can play a role. Of course, knowing
Visual signaling elements guide the user‘s visual about relevant factors is of interest both from a
attention to relevant information. They help the user theoretical and from an applied perspective. Note that
to identify relevant information, which improves user the importance of understanding boundary conditions
performance. In light of the findings reported here and for design decisions has also been emphasized in
findings from prior research, practitioners can derive research on multimedia learning (Mayer, 2009; Sweller,
the recommendation that they should consider using Ayres, & Kalyuga, 2011).
screenshots in software documentation containing The eye tracking method and the specific
procedures, but that screenshots should be designed experimental setup developed to study the reading
properly to leverage their full potential. Screenshots behavior of users in a “reading-to-do” scenario have
are more helpful if they are enriched with visual proven very useful, which suggests that eye tracking
signaling elements that relate to the task at hand and can help to address the questions for future research

408 Technical Communication l Volume 66, Number 4, November 2019

 Applied Research

Michael Meng

raised above as well. A key feature of the experimental switching and the coordination between reading and
setup used here is that the software system and the acting, the question arises whether visual signaling
documentation were presented simultaneously on the also helps the user to locate the relevant parts of the
same screen. An advantage of this setup is that remote user interface that are depicted by the screenshot and
eye tracking systems can be used for data collection emphasized by signaling. Eye tracking may pave the
and that techniques for quantitative data analysis, such way for a deeper understanding of attention switching
as the definition of areas of interest and comparisons processes, which in turn could lead to specific design
across areas of interest, can be applied in an efficient recommendations that help add additional value to
way, simply because the screen provides a fixed technical documentation.
point of reference.
This opens new possibilities to leverage the References
potential of eye tracking for determining accurately
where users look while still allowing users to work with Acharya, K. R. (2017). User value and usability in
the software system and the respective documentation technical communication: A value-proposition
simultaneously in a fairly unconstrained way. In design model. Communication Design Quarterly
particular, the setup does not enforce a fixed sequence Review, 4(3), 26–34.
of reading and acting. Consequently, the setup Alexander, K. P. (2013). The usability of print and
described here opens the possibility of applying eye online video instructions. Technical Communication
tracking to new areas of technical communication Quarterly, 22, 237–259.
research that extend the scenarios discussed in Cooke Baayen, R. H., Davidson, D. J., & Bates, D. M.
(2005). One such possibility is to use eye tracking to (2008). Mixed-effects modeling with crossed
study reading strategies and aspects of information random effects for subjects and items. Journal of
selection in manuals beyond visual signaling. For Memory and Language, 59(4), 390–412.
example, eye tracking investigations could reveal which Balling, L. W. (2018). No effect of writing advice on
types of information users access spontaneously in reading comprehension. Journal of Technical Writing
manuals, at which point they access certain information and Communication, 48, 104–122.
(e.g., before, during or after carrying out an action), Barr, D. J., Levy, R., Scheepers, C., & Tily, H. J.
and which variables (e.g., reading goal or level of (2013). Random effects structure for confirmatory
expertise) influence the section process (van der Meij et hypothesis testing: Keep it maximal. Journal of
al., 2009; Ummelen, 1999). Memory and Language, 68(3), 255–278.
Another area in which eye tracking—when used in Bates, D., Kliegl, R., Vasishth, S., & Baayen, H. (2015).
a setup that enables parallel use of documentation and Parsimonious mixed models. arXiv.org. Retrieved
software system—can prove very useful is to study how from https://arxiv.org/pdf/1506.04967.pdf.
users coordinate reading and acting in task execution, Bates, D., Mächler, M., Bolker, B., & Walker, S.
and how they switch attention between documentation (2015). Fitting linear mixed-effects models using
and software system when working with procedures. lme4. Journal of Statistical Software, 67(1), 1–48.
Switching attention is an important process that Boekelder, A., & Steehouder, M. (1998). Selecting
mediates between selecting information contained in and switching: Some advantages of diagrams over
a manual and putting this information to actual use tables and lists for presenting instructions. IEEE
(Boekelder & Steehouder, 1998; van der Meij, 1998; Transactions on Professional Communication, 41,
van der Meij & Gellevij, 1998). An important task for 229–241
future research is to determine which design techniques Boekelder, A., & Steehouder, M. (1999). Switching
support the attention switching process effectively and, from instructions to equipment: the effect of
therefore, contribute to more effective and efficient graphic design. In H. Zwaga, T. Boersema, & H.
task execution. For example, this study has shown Hoonhout (Eds.), Visual information for everyday
that visual signaling in screenshots helps the user to use: Design and research perspectives (pp. 67–73).
identify screenshot information that is relevant in the London, UK: Taylor & Francis.
context of the current task. With respect to attention

Volume 66, Number 4, November 2019 l Technical Communication 409

 Applied Research

Effects of Visual Signaling in Screenshots

Carliner, S., Coppola, N., Grady, H., & Hayhoe, tracking: A comprehensive guide to methods and
G. (2011). What does the transactions publish? measures. Oxford, UK: Oxford University Press.
What do transactions’ readers want to read? IEEE Irrazabal, N., Saux, G., & Burin, D. (2016). Procedural
Transactions on Professional Communication, 54, multimedia presentations: The effects of working
341–359. memory and task complexity on instruction time
Cooke, L. (2005). Eye tracking: How it works and how and assembly accuracy. Applied Cognitive Psychology,
it relates to usability. Technical Communication, 52, 30, 1052–1060.
456–463. Jaeger, T. F. (2008). Categorical data analysis: Away
Cooke, L. (2010). Assessing concurrent think-aloud from ANOVAs (transformation or not) and
protocol as a usability test method: A technical towards logit mixed models. Journal of Memory and
communication approach. IEEE Transactions on Language, 59(4), 434–446.
Professional Communication, 53, 202–215. Koning, B. B. de, Tabbers, H. K., Rikers, R. M., &
Cooke, L., Taylor, A. G., & Canny, J. (2008). How Paas, F. (2009). Towards a framework for attention
do users search Web home pages? Technical cueing in instructional animations: Guidelines for
Communication, 55, 176–194. research and design. Educational Psychology Review,
Duchowski, A. T. (2017). Eye Tracking Methodology (3rd 21(2), 113–140.
ed.). Cham: Springer International Publishing. Lenth, R. V. (2016). Least-Squares Means: The R
Elling, S., Lentz, L., & Jong, M. de. (2012). package lsmeans. Journal of Statistical Software,
Combining concurrent think-aloud protocols 69(1), 1–33.
and eye-tracking observations: An analysis of Lorch, R. F. (1989). Text-signaling devices and
verbalizations and silences. IEEE Transactions on their effects on reading and memory processes.
Professional Communication, 55, 206–220. Educational Psychology Review, 1(3), 209–234.
Farkas, D. K. (1999). The logical and rhetorical Martin-Michiellot, S., & Mendelsohn, P. (2000).
construction of procedural discourse. Technical Cognitive load while learning with a graphical
Communication, 46, 42–43. computer interface. Journal of Computer Assisted
Gellevij, M., & van der Meij, H. (2002). Screen Learning, 16(4), 284–293.
captures to support switching attention. IEEE Matuschek, H., Kliegl, R., Vasishth, S., Baayen, H., &
Transactions on Professional Communication, 45, Bates, D. (2017). Balancing Type I error and power
115–122. in linear mixed models. Journal of Memory and
Gellevij, M., & van der Meij, H. (2004). Empirical Language, 94, 305–315.
proof for presenting screen captures in software Mautone, P. D., & Mayer, R. E. (2001). Signaling as a
documentation. Technical Communication, 51, cognitive guide in multimedia learning. Journal of
224–258. Educational Psychology, 93(2), 377.
Gellevij, M., van der Meij, H., Jong, T. de, & Pieters, Mayer, R. E. (2009). Multimedia learning (2nd ed.).
J. (1999). The effects of screen captures in manuals: Cambridge, UK: Cambridge University Press.
A textual and two visual manuals compared. IEEE Mayer, R. E. (2010). Unique contributions of eye-
Transactions on Professional Communication, 42, tracking research to the study of learning with
77–91. graphics. Learning and Instruction, 20(2), 167–171.
Gellevij, M., van der Meij, H., Jong, T. de, & Pieters, Moreno, R. (2007). Optimising learning from
J. (2002). Multimodal versus unimodal instruction animations by minimising cognitive load: Cognitive
in a complex learning context. The Journal of and affective consequences of signalling and
Experimental Education, 70(3), 215–239. segmentation methods. Applied Cognitive Psychology,
Guillemette, R. A. (1989). Usability in computer 21(6), 765–781.
documentation design: Conceptual and Moreno, R., & Mayer, R. (2007). Interactive
methodological considerations. IEEE Transactions multimodal learning environments. Educational
on Professional Communication, 32, 217–229. Psychology Review, 19(3), 309–326.
Holmqvist, K., Nyström, M., Andersson, R., Dewhurst, Nielsen, J., & Pernice, K. (2010). Eyetracking web
R., Jarodzka, H., & van de Weijer, J. (2011). Eye usability. Berkeley, CA: New Riders.

410 Technical Communication l Volume 66, Number 4, November 2019

 Applied Research

Michael Meng

Nowaczyk, R. H., & James, E. C. (1993). Applying Minimalism beyond the Nurnberg Funnel (pp. 275–
minimal manual principles for documentation 309). Cambridge, Mass: MIT Press.
of graphical user interfaces. Journal of Technical van der Meij, H. (2000). The role and design of screen
Writing and Communication, 23, 379–388. images in software documentation. Journal of
Rayner, K. (2009). Eye movements and attention Computer Assisted Learning, 16, 294–306.
in reading, scene perception, and visual search. van der Meij, H., & Gellevij, M. (1998). Screen
Quarterly Journal of Experimental Psychology (2006), captures in software documentation. Technical
62(8), 1457–1506. Communication, 45, 529–543.
Rayner, K., & Pollatsek, A. (2006). Eye-movement van der Meij, H., & Gellevij, M. (2002). Effects of
control in reading. In M. Traxler & M. pictures, age, and experience on learning to use a
Gernsbacher (Eds.), Handbook of psycholinguistics computer program. Technical Communication, 49,
(pp. 613–657). Cambridge, MA: Academic Press. 330–339.
R Core Team. (2017). R: A Language and Environment van der Meij, H., & Gellevij, M. (2004). The four
for Statistical Computing. R Foundation for components of a procedure. IEEE Transactions on
Statistical Computing, Vienna, Austria. Professional Communication, 47, 5–14.
Redish, J. C. (1989). Reading to learn to do. IEEE van der Meij, H., Karreman, J., & Steehouder, M.
Transactions on Professional Communication, 32, (2009). Three decades of research and professional
289–293. practice on printed software tutorials for novices.
Redish, J. C. (2010). Technical communication Technical Communication, 56, 265–292.
and usability: Intertwined strands and mutual van Genuchten, E., Hooijdonk, C., Schüler, A., &
influences. IEEE Transactions on Professional Scheiter, K. (2014). The role of working memory
Communication, 53, 191–201. when ‘learning how’ with multimedia learning
Richter, J., Scheiter, K., & Eitel, A. (2016). Signaling material. Applied Cognitive Psychology, 28(3),
text-picture relations in multimedia learning: A 327–335.
comprehensive meta-analysis. Educational Research Williams, T. R., Mulligan, C., Koprowicz, K., Miller,
Review, 17, 19–36. J., Reimann, C., & Wang, D.-S. (2005). Does
Spyridakis, J. H. (1989). Signaling effects: A review of isolating a visual element call attention to it? Results
the research—Part I. Journal of Technical Writing of an eye-tracking investigation of the effects of
and Communication, 19, 227–240. isolation on emphasis. Technical Communication,
St.Amant, K., & Meloncon, L. (2016). Reflections on 52, 21–27.
research: Examining practitioner perspectives on
the state of research in technical communication. About the Author
Technical Communication, 63, 346–364.
Sweller, J., Ayres, P., & Kalyuga, S. (Eds.). (2011). Michael Meng is a professor of Applied Linguistics at
Cognitive load theory. New York, Dordrecht, Merseburg University of Applied Sciences (Germany)
Heidelberg, London: Springer. where he teaches text production, research methods,
Sweller, J., & Chandler, P. (1994). Why some material and usability. Before joining the university faculty, he
is difficult to learn. Cognition and Instruction, 12(3), worked as a technical writer and localization specialist
185–233. for an international software company. His research
Ummelen, N. (1999). Studying the process of focuses on using empirical methods to study the effects
information selection in manuals: a review of four of linguistic and design variables on the usability of
instruments. Document Design, 1(2), 119–130. information products in technical communication.
van der Meij, H. (1996). A closer look at visual Contact: [email protected].
manuals. Journal of Technical Writing and
Communication, 26, 371–383.
van der Meij, H. (1998). Optimizing the joint handling Manuscript received 12 March 2018, revised 29 May 2018;
of manual and screen. In J. M. Carroll (Ed.), accepted 3 July 2018.

Volume 66, Number 4, November 2019 l Technical Communication 411

View publication stats