ENPE Critiquing
ENPE Critiquing
ENPE Critiquing
This social research focuses on the most common class of psychiatric problems
affecting both children and adults, which is known as anxiety disorders. Since the tools to
effectively monitor and manage anxiety, and comparative research about it are lacking, this
social research studied leveraging passive and unobtrusive data collection from smartphones
because it could be a viable alternative to classical methods, allowing for real-time mental
health surveillance and disease management. The same was made possible through the use of
eWellness, which is an experimental mobile application designed to track a full suite of sensor
and user-log data off an individual’s device in a continuous and passive manner. The metrics of
the mobile application include communication, location, ambient sound, activity and movements,
light, and phone use, wherein these were derived daily aggregated features from these metrics
to define the relationship between smartphone sensors and anxiety symptom severity.
Additionally, this study was scoped to only include individuals without a clinical diagnosis of
Anxiety or Depression. It was reported that the initial pilot study tracking ten people over a
month showed a nearly 76% success rate at predicting daily anxiety levels based solely on the
passively monitored features.
eWellness
- an app designed to capture a broad spectrum of remote monitoring, survey data
acquisition, secure data transmission and management, data analytics, and visualization.
Personally identifiable information (PII)
- any data that could potentially be used to identify a particular person. Examples include
a full name, Social Security number, driver’s license number, bank account number,
passport number, and email address.
Under-sampling technique
- Undersampling refers to a group of techniques designed to balance the class distribution
for a classification dataset that has a skewed class distribution.
The social research was written by Levine et al who are studying in University of
California, Los Angeles. This social research was published on Cornell University website.
Proceeding in analyzing the social research using the “How to Read an Essay” article, The
authors were unbiased and qualified in writing the study because they are graduate researchers
or Phd candidates at UCLA. The essay was not out of date but rather very timely in this
technological era. Looking at the title, “Anxiety Detection Leveraging Mobile Passive Sensing”, it
convinced me to read it and choose it as the social research that I would be critiquing. The title
also helps set up the research’s tone where it is a formal tone and has used narration, definition,
and comparison as the modes of development. The abstract presented in the research paper
helped me to understand their thesis. I did not research the jargon present in the study since it
is already defined and explained in the research itself. The transition devices were used
effectively and have created a smooth flow from idea to idea. I just didn’t understand the graph
since it was complicated for me. The results were presented well because it was not repetitive
nor boring.
Using the article “Seven Sins of Technical Writing”, the authors did not commit sin 1
which is indifference. I am not a psychology/programming major, however, when I was reading
their research, I can comprehend almost all the information exhibited. There was no neglect on
the part of the authors by making sure the readers can really understand what the study is all
about. Assessing how the writer used necessary jargon only, I noticed that Sin 2, fuzziness, was
not committed. The authors used concrete, clear, & and specific words and phrases. Whenever
I encounter jargon on the paper, it is defined in the paper itself. The study also showed no sign
of wordiness, sin 4. The authors also did not use too many complex or abstract words, but
rather they explained it in the fewest words possible without neglecting the content. The authors
did not commit the 5th sin, bad expressions, because they paid attention to the usages of
phrases. There was no sign of awkward and confusing sentence structure. However, sin 6,
deadly passive, is committed by the authors because when I was reading it, it was made
impersonal, eternal, remote, and dead. It is maybe because it is focused on giving information to
the reader that the writer fails to recognize that it will give the readers a hard time retaining the
information given. In some sentences, the subject receives an action, instead of the subject
performing an action. Furthermore, as I’ve carefully read the social research, I can say that the
authors did not commit sin 7, mechanical errors. There were no poorly placed modifiers and the
article was grammatically correct. To sum up, there is only a single sin of writing an essay that
was committed in social research, namely the Deadly Passive.
Meanwhile, I honestly had a hard time understanding the graph presented in my chosen
social research. The confusion matrix really made me confused since it’s my first time
encountering that type of graph. Maybe they assessed that confusion matrix is best to show the
data since it is effective if used for evaluating the performance of a classification model,
however, it will be appreciated if they showcase another type of graph that will be understood by
people who have little to no knowledge about the algorithm and complicated graphs. However, I
acknowledge that this type of graph is useful for those who know how to read it since the matrix
not only gives you insight into the errors being made by your classifier but also the types of
errors that are being made. Additionally, the study aims for an optimal method for the prevention
or care of mental illness through early identification, therefore the predictive data that the graph
shows fit the aim of the researchers. Furthermore, the researchers used the under-sampling
technique because as mentioned by the researchers, they used it to improve the performance of
an imbalanced dataset. The problem with this sampling is that it made the dataset too small to
achieve statistically significant results. Also, I don’t know if my lack of comprehension of the
algorithm is the problem but the statistical analysis for me is lacking since I still didn’t
understand it. However, on the positive side, the interpretation which is in their discussion is
comprehensive and not bias at all. The writing style is appreciable and was able to formally
communicate the information regarding the interpretation of the result (except for the graph & its
analysis). I observed that the research is satisfactory in the data gathering considering the fact
that classification success was achieved by bundling samples across all subjects is remarkable
in its indication that cross-subject learning in this domain could be possible. The 76% success
rate at predicting daily anxiety levels based solely on the passively monitored features is great
enough to be an excellent stepping stone in refining the technology of the anxiety detector app.
Ultimately, the social research entitled ”Anxiety Detection Leveraging Mobile Passive
Sensing” was written satisfactorily as it delivers its data professionally and progressively. There
are parts to be improved, like providing a more comprehensive graph, and statistic analysis for
the people who can’t understand the algorithm. However, considering how the app eWellness
was well researched in a way that it can help for optimal anxiety detection for a substantial
percentage, is wholly impressive.
IV. QUESTIONS
2. Using the technological algorithm, will you be able to connect the eWellness application
to emergency hotlines or mental health professionals when the person is experiencing
severe anxiety?
Sources:
under-sampling technique. 2021. In Machine Learning Mastery. Retrieved May 16, 2021, from
https://machinelearningmastery.com/undersampling-algorithms-for-imbalanced-classification/