Last time I checked in, we were packing for Zanzibar, and I can confidently say it was just as magical as I had hoped. We started in Stone Town, a place that instantly won me over. There’s something about coastal towns that always pulls me in, and Stone Town, with its narrow alleyways, historic charm, and buzzing local life, was no exception.
On day two, we visited the spice market, which felt like a rite of passage. Zanzibar isn’t called “The Spice Island” for nothing, and I made sure to pick up some local spices to bring back. Then it was off to the Jozani Forest, where we saw blue monkeys and rare red colobus monkeys, endemic to the island. That night, we celebrated my birthday on Nungwi Beach with dinner.
The next day brought my favorite adventure yet: snorkeling off the coast of Mnemba Island. I’vealways loved it, but what I didn’t love, until now, was the idea of swimming near dolphins. But when our boat driver suddenly spotted some and told us to jump in, I did. I think the spontaneity helped override my fear of sea creatures larger than fish. And I’m glad I went for it, it ended up being one of the coolest things I’ve ever done. The fish were incredibly colorful, and thedolphins, though intimidating, were mesmerizing to watch in the wild.
On our last day, we visited Prison Island, famous for its giant tortoises, which was a really cool experience. What surprised me even more was learning about the island’s history: despite its name, Prison Island was never actually used as a prison. Instead, it served as a quarantine station for people suspected of carrying Yellow Fever and other contagious diseases before they were allowed to enter Zanzibar.
Before we headed back to Dar, we returned to Stone Town for a rooftop lunch and visited the East African Slave Trade Exhibition. It gave us a deeper understanding of the island’s history, especially Zanzibar’s role in the transatlantic and Indian Ocean slave trades. Seeing the holding chambers and reading personal accounts was heartbreaking, but it also emphasized the significance of remembering and learning from the past.
After such a full weekend, it was time to dive back into my research project. Over the past two weeks, I’ve made significant progress in understanding and visualizing ambient temperature trends for our partner facilities.
One of my first visualizations tracked whether each facility stayed within the recommended 25–35°C range during any given month. The rule was strict: if even one daily minimum temperature reading fell below or above this range, the facility was flagged.
(Facility IDs have been removed from this visualization.)
From there, I moved on to exploring ambient and newborn temperature data together. This integration step took a lot of time and quite a few data revisions. When I finally visualized the combined data, I found something surprising: most cases showed babies with low body temperatures (hypothermia) even when the ward temperatures were within or above the recommended range.
This unexpected result sparked some important conversations. One possibility? The location of ambient monitors. Many are installed near the ceiling to avoid interference, but heat rises, so the baby’s actual environment at bed level could be significantly cooler. That gap might explain the mismatch. It also raises the question of whether we need to make a data adjustment to better reflect the temperatures experienced by the newborns.
There are further limitations in the data: ambient readings include daily minimums, maximums, and averages, while baby temperatures are usually captured as a single lowest reading after admission, with no specific timestamp. To bridge this gap, I used an average of ambient temperatures from the seven days following each baby’s admission based on NEST360’s collected data, which shows that most ward stays across the countries last under a week. While this assumption isn’t perfect, it allows for a more consistent comparison between ambient and newborn temperatures. Working through this has shown me just how much nuance and decision-making goes into data science, and I’ve been so grateful for my mentor John’s guidance throughout.
With a cleaner dataset, I began running correlation analyses and found statistically significant, but weak, relationships between ambient and newborn temperatures. This suggests that while ambient conditions do matter, other factors (like birth weight, season, or facility practices) likely play big roles too. Adding controls for birth weight and seasonal changes strengthened the model, and I’m optimistic that when we introduce satellite-derived data, we’ll see even more robust relationships. I’m also hoping to explore time series models soon, so I’ve been brushing up on my forecasting skills!
Alongside the temperature analysis, I’ve also identified a new need-finding project at DIT. Joel mentioned he’s collected a ton of visitor data for the DIT Design Studio (why people visit, what resources they use) but hasn’t had a good way to process or use it. So I will begin designing a dashboard that automatically updates using a Google Form and summarizes visitor activity. The goal is to help the lab understand how it’s being used and identify areas where students may need more support. If successful, it could also help make the case for additional funding or resources.
And what would a blog be without a glimpse of the next adventure?
As I write this, we’re on our way to Arusha, a gateway to Tanzania’s most famous national parks. Over the next few days, we’ll be visiting Tarangire, Ngorongoro Crater, and the Serengeti. It’s a dream come true, and I already know I’ll have so much to write about in the next post.
Hadi wakati huo!