Forum Replies Created
-
Posts
-
-
Q1. Why was the author interested in investigating the suicide problem in Thailand during that time?
The fact that the number of suicides has consistently ranged between 3,600 and 4,000 annually would draw noticeable attention to normal people. The author, including scholars, was interested in investigating the factors influencing the suicide problem in Thailand during that time since macro-level data on the topic for Thailand is limited and it has different economic and social characteristics than industrialized societies.
Q2. Picks one potential risk factor mentioned in the paper and explains how the variable can contribute to the suicide rate.
In the paper, a theoretical framework, the Utility Maximization Model, suggests that a person’s age and income influence their probability of suicide.
The question is, how is the divorce rate associated with the suicide rate, while the framework mentions only age and income?
One of the five assumptions delivered by the theory (Lester and Yang, 1997) is that divorce leads to the reduction of household income; therefore, such circumstances are more likely to commit suicide.
In the paper, the relationship between divorce rates and suicide rates is found to be statistically significant, with a coefficient of 0.508 showing a positive impact on suicide rates; meaning as the divorce rate increases, the suicide rate also tends to increase.
Q3. How can statistical modeling contribute to investigating the epidemiology and spatial aspects of the Thai suicide problem?
First, the author was interested in investigating the factors influencing the suicide rate in Thailand. So, to mathematically make a valid conclusion, statistical modeling helps to identify and quantify the impact of various factors on suicide rates across different places. By regression analysis, the author explored which socio-economic variables are statistically significant determinants of suicide rates.
To my understanding, spatial epidemiology is the description and analysis of geographic variations in disease with respect to demographic, environmental, behavioral, socioeconomic, genetic, and infectious risk factors (Elliott & Wartenberg, 2004). The paper focuses on a multiple regression analysis to find correlations between these factors, rather than analyzing the geographical distribution or patterns of the disease.
However, in associating spatial aspects of the suicide problem, the study identifies that the northern region of Thailand had the highest suicide rates, while provinces in the northeastern, central, and Bangkok regions had lower rates. This helps to map out and identify specific areas that may require targeted interventions.
Reference
Prasertthai, S., & Panyathorn, J. (2019). Determinants of Suicide Rates in Thailand. *Journal of Community Development Research (Humanities and Social Sciences)*, *12*(1), 15-24. [https://doi.org/10.14456/jcdr.2019.2](https://www.google.com/search?q=https://doi.org/10.14456/jcdr.2019.2)Elliott, P., & Wartenberg, D. (2004). Spatial epidemiology: Current approaches and future challenges. *Environmental Health Perspectives*, *112*(9), 998-1006. [https://doi.org/10.1289/ehp.6720](https://www.google.com/search?q=https://doi.org/10.1289/ehp.6720)
-
Q1 (a). What are possible reasons locations in epidemiological research have not been incorporated as much as other components in epidemiological research?
Answer: The main reason could be the difficulties in correlating with places/locations in epidemiological research because chorographia or geography was not a part of epidemiological research. It is more likely that it was focused on person and time dimensions, for example, gender, age, occupation, socioeconomic level, disease trends based on a specific period, such as weather and climate. Other possible reasons could be a lack of technology support and data acquisition. It would not be convenient to collect data and tag locations on a paper map, and in the end, there would be challenges in performing clustering and correlation geographically. Therefore, locations were not much incorporated in the epidemiological research.
Q1 (b). How can spatial epidemiology be considered as an interdisciplinary science?
In discovering disease distribution patterns and the causative determinants among populations, focusing on locations, epidemiological knowledge is insufficient. Geography plays in understanding information about a place (such as weather, soil, air, and temperature). To infer the correct result from collected data, statistics helps with valid logical reasoning. To develop efficient analysis algorithms and visualization from large datasets, knowledge of computer science is required. Environmental science and public health also help identify environmental risk factors affecting disease patterns. Therefore, spatial epidemiology is inherently interdisciplinary and cannot function as an isolated field.
Q2 (a). Explain why it is widely recognized that the place where an individual lives or works should be considered as a potential disease determinant, and give some examples?
Vector-borne diseases such as malaria and dengue are distributed in spatial patterns. Living in such regions could increase the risk.
People in remote rural areas lack health infrastructure, resulting in poorer health outcomes.
People living near or working in mines, landfills, and industrial complexes have a higher chance of lung cancer.
Therefore, it is widely recognized that the place where an individual lives or works should be considered as a potential disease determinant.
-
What role does privacy play in user evaluations?
Not all users who rated the privacy aspects of the apps were satisfied with them. The majority of users expressed concerns about their privacy. It includes
– Excessive permission requests by software to function
– Concerns of data leakage and personal information security, such as unwanted SMS notifications for refunds on orders that were never placed.
– Unauthorized alteration of vaccine records, such as a child’s designated hospitals inexplicably being changed to another location.In my opinion, these facts require a closer look.
First, unless requested permissions are given, does the software/app still function well? If it works well, it is excessive or suspicious.
Second, if unwanted SMS notifications are due to a technical glitch and happened once, this can be considered an unintentional mistake unless it falls into violation of privacy.
Third, from my perspective, alteration of designated vaccination sites seems acceptable, while it does not seem to be a medical record. It is more of a capacity or management issue.
-
How did government restrictions affect mobility patterns?
The government restrictions include the closure of the friendship bridge, provincial entry and exit restrictions, with the establishment of checkpoints on main roads and a nationwide curfew.
It was introduced during Interval 4: 4 April 2020 to 3 May 2020 (30 days), the period of the most restricted travel.
The study measures the mobility in the radius of gyration (RoG) and observes the greatest reduction in movement, by more than 90% of the typical travel, by a 32% decrease in short trips outside the home village, and by a 70% decrease in long trips. Additionally, the nationwide curfew was effective, and the RoG was remarkably low during those hours.
The implementation of government restrictions during the pandemic led to a significant decline in mobility in their studied population.
-
What factors influenced participant engagement in the study?
According to the paper, engagement from the web-based survey is higher than that from the mobile app. The platform is a major factor influencing the rate of engagement. In detail, responding to the survey from the app involves a series of steps such as downloading, account registration, activation, and so on. As for me, I won’t try spending these steps just for taking a survey, while participating in the survey already feels like a huge burden and a sheer waste of time.
A reminder also plays a crucial role in engagement in the study. Additional email reminders significantly improved survey completion. Psychologically, the reminder emails would remind them how important the survey is, and how important they are for the survey. If I were them, I would do the same, respect and respond to the survey.
The paper concludes that older adults have the lowest response rate among mobile app users, which is attributed to their lower technology literacy and discomfort with mobile apps. However, from the gerontological point of view, older people are not patient with complexity and longer waiting times. Additionally, normal people would not face barriers to the completion of the survey via the app. In my opinion, it is correct to say due to discomfort with mobile apps or user experience (UX) issues.
-
How did apps affect vaccination uptake rates?
9 of the 28 studies assess how apps affect vaccination uptake rates. Only 4 studies reported a significant improvement in vaccination coverage.
Upon checking the apps from those 4 studies, they deal with immunization-related tasks such as immunization records associated with health workers and doctors. One unique app among them is Carrot Rewards, which enables users to earn loyalty points for completion of vaccination-related tasks.
Generally, regardless of these studies correlating with increased vaccination uptake rates, it can be said that apps do not affect the rate. Instead, the vaccination workflow requires such apps to streamline the process of vaccination records. Carrot Rewards is a different and innovative app that utilizes a brain-rewarding system to push users vaccinated.
-
My first reaction to using the virtual space is that it’s something gamers would enjoy doing because it’s not different from a gaming environment. In sharing pictures or presenting in a shared virtual space, normally, we assume that sharing in a classroom or discussion board would draw more attention than virtual space because we have physical presence and human-to-human interaction, but the reality would seem to be the opposite. We rush to finish writing and reply to a discussion, or we could be daydreaming in a class during a lecture or presentation. In these ways, we already lose our attention. In the virtual environment, what we see is unavoidable, meaning that we become immersed in it and draw attention to everything.
In academics or professional settings, there can be situations where a virtual environment is required, for example, visualization of 3D objects – human anatomy, visualization of micro-molecules, flight simulation, and training. Just as a 2D picture is worth a thousand words, a 3D image is worth a million words. It will enhance and expand our understanding of different perspectives and enable us to see a wider picture.
–
–
-
I tried watching VR videos on YouTube and VR apps such as VR Moon and VR Space. Honestly, I found it very hard to enjoy the experience. The stereoscopic vision was uncomfortable. Immediately after putting it on, I got motion sickness. I also noticed that my right and left eyes were trying to calibrate and not working together for some time. I worried that it might lead to cross-eyedness. After adjusting, aligning, and adapting the vision, I watched VR videos for a minute, and then I got pain or an ache around or inside my eyes.
A standout feature, I feel, is that it replicates the maximum experience of realism. Watching YouTube VR videos on “Indigenous people in the Amazon forest” and “New York street walk” feels like I was there walking and going for an adventure, and makes me realize what it’s like to be there. However, my phone does not have a very high pixel density, which makes me feel like watching pixelated videos or a pixelated experience.
I could only find the head gaze combined with the screen tap interaction method in Google Cardboard and YouTube. The VR cardboard seems purposely made to be smaller than the average size of the phone. My phone, for instance, is bigger and has extra screen space. Later, I found that it aims for screen tap interaction with head gaze.
-
Visualization & Experimentation
While the aircon is on, the temperature and humidity remain the same.
After turning it off and opening all windows, the humidity climbed up while the temperature increased gradually.
Then, I tried to test it with all windows closed while the aircon was still off. The result shows that the temperature and humidity stay in balance.
After the aircon is turned on, I noticed that as the temperature decreases, the humidity also decreases.
During this experiment, the circuit board and sensor are stationary.
Challenges & Solutions
After uploading code to the circuit board, I could not see any output in the console. I noticed that it is suggested to use 9600 baud written on the board. After changing the code to
9600 baudand the console to 9600 baud, it becomes visible.Ideas for Improvement & Application
If I had more time, one modification I would make would be to cache failed records and push them if the internet connection is restored. Because while I was testing this experiment, I was using my neighbor’s Wifi (a very, very poor connection).
In this case, the complexity is not as simple as we think. Because, as far as I’ve researched, the ESP 8266 microcontroller does not have an internal clock. So, we have to manage time programmatically again.The trick is to retrieve the current time from the Network Time Protocol (NTP) via Wi-Fi (internet). Once we get the current time, we compute the current time as time goes by in the loop. However, this can drift over time. So, re-adjusting the clock with NTP will be needed once the connection is restored. While the connection is offline, take the current time from the internally running computed time and cache it with temperature and humidity data.
A real-world public health scenario where this system could be useful is monitoring vaccine cold chain temperature. So, if a circuit board is integrated / integrable with a telecom network/telecom module, it could trigger an automatic SMS alert. If we continue to use the ESP 8266 microcontroller, we could use a cloud communication platform such as Twilio to trigger an SMS message via API.
-
I chose an old article, “A Review on Architectures and Communications Technologies for Wearable Health-Monitoring Systems” by Víctor Custodio, Francisco J. Herrera, Gregorio López, and José Ignacio Moreno (2012) to see what an IoT looked like previously.
Summary
What it does—The article details the LOBIN platform, an IoT platform based on e-textiles and Wireless Sensor Networks (WSN) for healthcare monitoring.
Where it was implemented – It was implemented as a pilot scheme in the Cardiology Unit of La Paz Hospital in Madrid, Spain, to monitor psychological parameters and track patient location.
How it generally works– The system utilizes smart shirts (e-textiles) integrated with sensors that transmit data wirelessly through a network infrastructure.
Key points
Objective – The primary goal is to enable non-invasive and pervasive monitoring of patients with cardiac issues, but within hospital settings. It also aims to measure physiological data (ECG, Heart rate, SpO2, etc) and the indoor location of patients in real-time.
Sensors used –
(1) e-textiles electrodes to measure ECG
(2) 3-axis accelerometer to detect body movement and position
(3) thermometer to measure body temperatureChallenges –
(1) ECG signal loss during patient movement
(2) Data packet loss in their Wireless Communications Infrastructure Subsystems (WCIS)
(3) Patient discomfort requiring modifications like adjusting the smart shirt’s tightnessReference:
Custodio, V., Herrera, F. J., López, G., & Moreno, J. I. (2012). A Review on Architectures and Communications Technologies for Wearable Health-Monitoring Systems. *Sensors*, *12*(10), 13907-13946. https://doi.org/10.3390/s121013907
-
If the dashboard’s target audience is policymakers, it will be a bit scientific and filled with non-layman terms and indicators. In my dashboard, I’ll focus on simplicity.
For the dashboard, I chose Covid-19 dummy data and specified locations with states and regions in Myanmar. The disclaimer here is that the data will not make sense.
As for the policy makers viewing the dashboard, I assume that the following indicators will be important
_ incidence rate: the rate at which new cases are positive in a population per 100,000 people
_ mortality rate: the rate of deaths in a population per 100,000 people
_ case fatality rate: the rate of deaths among positive cases in a population per 100,000 people
_ confirmed / positive cases: the number of positive cases
_ deaths: the number of deaths
_ recovered cases: the number of cases being recovered
_ active cases: the number of positive cases right nowTherefore, I put those essential indicators as a scorecard in the dashboard.
Additionally, the map and detailed table are included. The map helps in identifying patterns of disease distribution geographically. The table describes a detailed report of surveillance information by regions across the country.
The Weekly Trend page is aimed at viewing the trend every week, which allows policymakers to view the trend as smoothing out daily fluctuations and providing a clear picture of the epidemic’s direction, trend, and patterns, which inform their decisions on public health measures, resource allocation, policy adjustment, and so on.
Here is my Looker Studio Project: https://lookerstudio.google.com/s/jVsxj2nfPqk
Please feel free to comment.
-
~
~
~
~
Looker Studio -> https://lookerstudio.google.com/s/vSbz9jI5P2k
-
1. Overview: Includes Score Card, Bar Chart, Running Sum and comparison
2. Cases / Deaths / Recovered: Includes Pivot Table, Running Sum and comparison
3. The last 28 Days Report: Includes Time Series
4. Cases Table by Date: Includes Table, Running Delta, Drill Down and Date, Running Sum and comparison
-
I would like to share COVID-19 Dashboard by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University (JHU).
What I like:
– At a glance, all important information is displayed and the site is responsive to any screen/device sizes.
– Color grading is strong but maintains the importance of information from RED for cases/deaths, GREEN for vaccination, creating intuitive associations
– Prominent key metrics at the top catch the eye (enhance iconic memory)
– Information metrics groups are chunked well for better working memory.
– The dark and minimal background avoids distraction patterns.What needs improvement
– Following chunking principles, large numbers are difficult to process such as 676,609,955 and it would be better displayed as 676.6M to reduce cognitive load.
– Red-green color scheme can be problematic for color-blind users
– Crowded country metrics in the left sidebar are unnecessary and do not allow easy comparison. This could be improved by an approach with a shorter list for better visual hierarchy. -
-
-
Very neat and practically clear design which is very friendly with data collector, especially in wide tick box.
In my opinion, I would remove weight in collecting demographic data, which I assume that it cannot be associated with the study.
Physical examination is well categorized including a detail section to provide its abnormal codes.
I am not sure what collected sample type means. If there is ID in the specimen or lab result, we can also add it.
More detail information about administrated vaccine can be added such as site of administration, route of administration, lot no, which would help in tracing back in finding out in manufacturing fault and adverse events.
-
Thank you for sharing your discussion, Than. My curiosity was piqued about whether the finding would differ from other kinds of research designs, such as longitudinal analysis, cohort studies, and time-series analysis, when you mentioned it’s a cross-sectional analysis. Regarding spatial analysis, I am not sure whether adding location data as an independent variable of regression can be said to be spatial analysis. I think your response to the second question seems to be miscopied.
-
As you discussed, I believe that divorce can bring serious mental complications in places where people consider marriage as a higher cultural value and societal expectation, as the paper mentioned. Statistical modeling, to be specific, two models of regression analysis are used for prediction and inference, by the relationship between the dependent variable (suicide rate) and the independent variables.
-
-
-
-
Thank you for sharing your discussion. Regarding databases, to be more specific, historically, they lacked the functions of databases that offer storage, management, integrity, security, data retrieval, etc, for performing spatial epidemiological research. The lack of such technology support acted as a barrier to the continuation of further exploring spatial epidemiology.
-
-
-
The ability to access user data by developers is a significant concern. This is why the critical elements of any software are not its features, but the software firm’s strict adherence to data privacy and compliance standards, which technically prevents in-house engineers or anyone from accessing live data.
-
-
Thanks for sharing your opinion. It’s crucial to maintain a patient-centered, empathetic perspective, even when medical professionals are overwhelmed with their tasks. Instead of rescheduling the patient’s appointment with one click without notice, one call to explain the reasons for the change would be more appropriate and respectful.
-
-
Thank you for sharing your discussion. I haven’t considered such an analysis by comparing travelers and non-travelers to provide insights for the imposition of stricter travel restrictions. As for me, since disease surveillance needs to be timely, I would suggest health screening monitoring, or requesting health declaration forms at points of entry from land, sea, and air, as soon as the initial evidence of a significant new threat, such as positive cases in neighboring countries, is received.
-
It’s interesting to see that night work/life is still functioning (very little), although the night-time curfew led to a significantly decreased RoG value. When “curfew” is mentioned, I thought it’s unnecessarily serious because I am familiar only to “military martial law curfew”. I didn’t realize other kinds of curfew exists, e.g., public health curfew.
-
-
-
For me, it does not make sense if apps affect vaccination uptake rates because those apps are made to support the workflow of vaccination, such as digital records. Even if those apps also include a public health education feature, it’s public health education that increases the uptake, not necessarily the apps.
-
-
-
Your discussion makes me wonder how sensors can get actual values instead of dependent values. In your case, temperature values are not immediately responsive to the environment but are being settled with heat affected on board (itself).
Your observation about the interval is very insightful. I didn’t realize that because the time value comes from App Script, and there’s a significant delay in data transmission, the sensor’s timestamps end up being incorrect. -
When my phone is also not fit with the cardboard, I utilize the extra screen space for touch/click interaction, where a pointer moves with my head on the VR screen (phone screen). So, to select an item, you can tap anywhere on the extra screen space. To navigate back, I rotate 90 degrees as described in the help tutorial on the Google Cardboard app.
-
Thank you for sharing a very insightful article. It is a comprehensive framework for IoT from all dimensions. Although the article does not explicitly mention the details of specific technical, logistical, or ethical challenges, it emphasizes the importance of 100% correct data transmission, identifying faulty sensors, the privacy of patients, and the increasing cost of investment.
-
-
-
I found your dashboard styling very advanced and well organized to fit on one page. On the other hand, there are some areas I would improve in terms of human perception.
_ less red color in the scorecard area on the left side; instead, the text color will be black by default, if serious, it will be red in title or number
_ use text for labelling instead of using scorecard default field name labelling which cannot be modified (in state / region scorecards, beneficiaries count labelling)
-
Hello Wannisa, as always, your presentation is elegant and aesthetic. I notice that controls, charts, and buttons are well organized and placed in the appropriate positions. Conditional formatting styling allows us to focus on the important information. I see that in this dashboard, you’ve followed all the principles learned from this course, for example, using graphic in donut chart.
-
-
-
Hi Kedsarin, your infographic is super creative and includes most of the discussion points in the webinar. If there is a legal case from using a teleconsultation app, not only doctors but also the platform could be held responsible in places where the platform is not required legally to have a physical hospital and services.
-
Hi Wannisa, thanks for sharing a summary that covers everything we’ve learned and discussed this week. I think the challenge of “legal issues when a problem occurs” could be simplified when a teleconsultation service via an app requires physical hospital and healthcare services, as in Thailand. Otherwise, as you said, legal issues could be debatable when a problem occurs.
-
-
-
-
