Stanford Earth Undergraduate Research Symposium
Summer research projects from the 2020 SESUR cohort
Natalie Cross, Dr. Nik Sawe, Dr. Tierney Thys, Dr. Brian Knutson
With over 145 million followers, National Geographic’s Instagram page is one of the most popular pages on Instagram. This study analyzed a set of 890 photographs posted on National Geographic’s Instagram page over a 3-month time period in order to answer the question, which elements lead to an impactful nature photograph? For the purposes of this report, impact is measured by two variables: (1) engagement and (2) positive vs. negative emotional arousal. Engagement refers to the interaction that Instagram viewers had with the photographs and was measured by finding the proportion between the total number of likes and comments on each photograph and the account’s total number of followers at the time of measurement. Arousal refers to the strength of emotion felt when viewing each photograph, with a high positive arousal (PA) correlated with strong positive emotion and a high negative arousal (NA) correlated with strong negative emotion. PA and NA were calculated using valence (positive vs. negative emotion) and arousal (strength of emotion) ratings made by research assistants for each image. Using Pearson’s correlation coefficient, we found that there is a positive correlation between both engagement and PA and the presence of nature and animals in a photograph and a negative correlation between both engagement and PA and photographs containing humans and/or urban settings. There was no strong correlation between NA and the presence of nature/animals vs urban/humans. In addition, engagement was also correlated with the biome region depicted in the photograph; mountain and ocean biomes were positively correlated with engagement while the desert biome was not correlated with engagement. There was not a strong correlation between PA and different biomes. Interestingly, there was a strong correlation between NA and photographs taken in ocean biomes and/or underwater, however mountain and desert biomes did not share this high level of NA. Lastly, there was also a positive correlation between both NA and engagement and the presence of predators/threats in a photograph. The correlations described above are all statistically significant with a p value > 0.001. The results outlined in this study are a preliminary portion of a much larger project. The next stage of the project will be to collect external data through a nationally representative survey. This survey will ask participants about how these images influence not only their engagement and emotional response (both in terms of valence and arousal and discrete emotions), but also their hypothetical willingness to donate to protect the species/region depicted in each image, and perceptions of depicted species (e.g., endangered status, familiarity). After this survey is conducted within the United States, we plan to expand it to both India and Indonesia in order to see if cultural differences have an influence on the impact of nature imagery.
https://openmappr.com/play/NatGeoInstagram
Hello Natalie!
I really loved the work that you did! I was very intrigued by the findings, especially the one where you showed that there tends to be more engagement with the photo when there are natural elements compared to when there are more humans. I have a question: do you think it is a finding that we can generalize for Americans (Since you also talked about doing the study across different countries like India and Indonesia) or do you think that there might have been the fact that people who follow National Geographic may tend to be biased in their engagement with photographs portraying natural environments or in their arousals due to possibly some pre-existing concerns, awareness and/or appreciation of nature, or because of pre-existing expectations on what they want to see from National Geographic?
Hi Nuzhah! That’s a really great question. All the data that I analyzed for this presentation were directly from National Geographic’s instagram page, so yes I think you’re right and there is definitely a good chance that people’s engagement was somewhat biased towards animals/nature considering that had already chosen to follow/view National Geographic’s instagram page. The next phase of this project, a nationally-representative survey in the United States, will hopefully be able to give us some more insight on if this correlation between nature and engagement/arousal continues for Americans who might not have a pre-existing affinity for nature. In addition to having questions about peoples’ feelings towards the images, we also have questions that will ask about their environmental attitudes, experiences and behaviors and hopefully results from this survey will give us a better answer to your question!
That’s great! It’s been great learning about your work, and I am very excited to learn about the results of the next phases of this research!
Hi Natalie,
Loved your work! Your research was so interesting, and your presentation and slides are very visually appealing.
I know there has been some work showing that people can purposely post on Instagram during certain times/days to get higher engagement, was timing of posting controlled for at all? I also wonder if there could be any correlation between low engagement on photos and major US news events or natural disasters (i.e. moments when people may be preoccupied with the news/life and less engaged on Instagram than normal). When was the 3-month period of the study?
I also found your findings super interesting! It’s exciting to hear that you’ll be able to continue working on – I’m excited to see how the cross cultural component of things turn out!
Hi Bianca!
That’s actually a really interesting question and something that I hadn’t thought about at all. When I was doing my analysis I didn’t really look into the time of day/date that the photos were posted. However, I do believe we have that information as a part of the metadata for each image and I’ll definitely want to try and look into that more because I think you might be right and there might be some interesting correlations between real-world events and rates of engagement with the photographs. Thank you for bringing that up, I’m excited now to see what I can find out!
Yay! Glad to be useful – sounds like exciting stuff!! 🙂
Hi Bianca! One of Natalie’s mentors on the project here – re: the 3 month window, the data was around the end of 2017, so we didn’t have big (pandemic-level) confounding events in the midst, thankfully. However, we have wondered about time of posting influencing things – we’d love to dig into that in a future dataset because naturally, on these platforms it can matter a lot!
Great work Natalie! Really polished presentation, you did a wonderful job with both the slides and the delivery. We can chat a bit later on about tweaks to the graphs/results to get into the specifics without being overwhelming (it can be a fine line), but this was a wonderful summary of some of the trends in the data. Can’t wait to see what we discover in the upcoming survey work!
Hi Nik!
Thanks for all your help this summer 🙂 I’m also really excited to see what we discover next!
Hi Natalie! This is a really cool study and it’s great to see how our projects could overlap within the Social Ecology lab. I would be interested to learn more about the coding techniques you used to measure the strength of emotions – were you looking at both captions and the images? I was also wondering if you think the results may have been largely due to the fact that followers of National Geographic are personally invested in protecting the natural world, but it seems that Nuzhah beat me too it and it’s great to see how the expansion of this project will help remove these biases. The next steps to understand how the impact of a photograph can lead people to donate (as a measure of taking positive action) should provide some great insights into how we can use social media as a motivator going forward!
Hi Fenella,
Thanks for your feedback! The initial coding that we did was just from some research assistants in the lab doing a simple ranking of their emotional responses to the images. This is definitely not the most substantial coding – but we have built in quite a few survey questions in our upcoming survey that will give us a much better coding of the images. I’m also super excited about the overlap between our two projects and think the results that come up in the next few months will be really interesting!
Hi Natalie,
Your project on nature photography’s impact was very interesting! I don’t follow national geographic on Instagram, but I know many people who share their pics on their stories. Your data on urban vs nature and positive arousal cool to hear about.
Hi Naomi,
I’m glad you enjoyed my project! I’d definitely recommend National Geographic’s instagram page if you’re looking for someone new to follow 🙂