metrics are numbered with a double ”M” to differen-
tiate them from the corresponding metrics.
Video Transmission Performance - MM04: For
mobile applications, there is a higher concern
over transmission performance since the internet
connection is not stable. If the transmission is not
satisfactory, the video will demand time to begin and
load repeatedly, which affects the user experience
(Hussain et al., 2016).
Proper Sizes - MM06: Mobile devices have smaller
screens, and usually, the interaction is touch-based,
which needs adaptation for the size of buttons, text,
and images (Budiu, 2015).
Adjust Video Quality - MM25: Since mobile de-
vices have a limitation regarding connectivity, it is
necessary to allow users to adapt the video quality
according to their preferences and capacity (Hussain
et al., 2016).
In addition to structuring the metrics, the three us-
ability criteria (effectiveness, efficiency, and user sat-
isfaction) also supported the definition of a priority
system. Each metric was classified regarding usabil-
ity criteria affected by not attending to it. We establish
three priority levels: high, medium, and low. Tables 1
to 3 show the metrics grouped by priorities according
to criteria. This priority system allows the use of these
metrics as requirements for software development ac-
cordingly.
Table 1 includes high-priority metrics whose fail-
ure to meet compromises effectiveness and efficiency
or both effectiveness and user satisfaction. Table 2 in-
cludes medium-priority metrics whose failure to meet
compromises effectiveness or both efficiency and user
satisfaction. Finally, Table 3 includes low-priority
metrics whose failure to meet only compromises ef-
ficiency or user satisfaction.
4 EVALUATION AND RESULTS
In this study, the five web platforms (ScienceTalks,
JoVE, Scivpro, Latest Thinking e STEMcognito)
and five mobile applications (WonderScience, TED,
SciShow, NewScientist, NASA) were evaluated ac-
cording to the 25 metrics presented previously. The
degree of adequacy of each interface was assessed
through the proposed prioritization of metrics to help
obtain a more structured view of the contribution of
platforms and applications to scientific communica-
tion through video articles.
Similar to Eliseo at al. (2017), the evaluation pro-
cedure was based on a heuristic evaluation, using the
metrics defined as a guide. A pre-defined list of tasks
was followed in each interface evaluation while ob-
serving the overall experience with the interaction to
determine whether the metrics had been completely
(green), partially (yellow), or not achieved (red), as
shown in Table 4. Note that when is not possible
to analyze a metric, it is indicated with ”NA”, what
means ”Not Applicable”.
4.1 Web Platforms
Table 4 shows for the high-priority metrics, despite
the significance of search engines in systems with
video libraries, that the Scivpro and STEMcognito
platforms performed poorly in M11 (Search). M16
(Subtitle and/or transcription) and M23 (Text alter-
native), two medium-priority metrics, are particu-
larly noteworthy as they encountered issues on four
out of five platforms. Concerning the low-priority
metrics, metric M10 (Help in Context) stands out
because none of the platforms completely cover it.
While Scivpro and STEMcognito did not fully meet
M14 (Use of design principles), the ScienceTalks and
STEMcognito platforms did not fully meet metric
M08 (Consistency between pages). It suggests that
the design of these interfaces needs attention.
Regardless of the priority category, all platforms
fully satisfied the metrics directly related to the videos
(M17, M18, M19, M20, M21, M22), according to the
analysis done using the platforms, which illustrates
the significance of guaranteeing high-quality interac-
tion with the video content offered by these platforms.
After the empirical evaluation, we analyzed the
resulting data, considering the priorities assigned to
each metric. Thus, for each priority, weights were de-
fined on a scale of 1 to 6, as follows: low priority
– weight 1 if the metric was partially achieved (yel-
low) and weight 2 if it was not (red); medium priority
– weight 3 if the metric was partially achieved (yel-
low) and weight 4 if it was not (red); high priority –
weight 5 if the metric was partially achieved (yellow)
and weight 6 if it was not (red). to
Table 5 shows the number of metrics for each web
platform evaluated that received each of the estab-
lished weights. Furthermore, it shows the total score,
which is calculated by multiplying the number of met-
rics inside the cell by the weight associated with pri-
ority. The lower the total score obtained, the more ap-
propriate the interface is for usability issues. There-
fore, the positive highlight is for the JoVE platform
with the most appropriate interface, and the negative
highlight is for the Scivpro platform, which presented
the most usability problems in the evaluation carried
out. Although Latest Thinking and STEMcognito
Enhancing Scientific Communication: Prioritizing User Experience in Audiovisual Interfaces
373