Instagram behavior for teens different than adults

"We found that teens have much higher levels of self-disclosure on the Internet," said Patrick Shih, a research associate at the College of IST.

Shih, along with Kyungsik "Keith" Han, a doctoral student at the College of IST; Dongwon Lee, an IST professor; and Jin Yea Jang, an IST master's student; conducted a comparative study of 27,000 teens and adults using Instagram, a popular photo-sharing mobile app. In their research, they show the possibility of detecting age information in user profiles by using a combination of textual and facial recognition methods, and of using that information to explore how teens use and engage Instagram compared to adults. They presented the results of their study in their paper, "Generation Like: Characteristics in Instagram," at CHI 2015, a leading conference for research in Human-Computer Interaction, in April, in Seoul, South Korea.

"We are able to show with real data how teens behave on social media," Shih said.

The researchers chose Instagram for their study because more than 90 percent of Instagram users are younger than 35. They defined teens as people between 13 and 19 and adults as those between 25 and 39. They used an application programming interface (API) -- a set of programming instructions guiding interaction with a particular app -- to extract users' data, and data collection was done between April and May 2014.

In their paper, the researchers propose a novel method that leverages online biographies and profile images with existing APIs. First, they applied textual pattern recognition algorithms to parse a list of patterns that specifically describe a user's age (e.g. "I am a teenager"). Second, they used an online tool called Face++, which was designed to detect ages and genders of people depicted in photos.

According to Shih and Han, the results of the study yielded some surprising results. For example, teens were found to post fewer photos on Instagram than adults. The researchers hypothesized that teens may have limited resources to explore environments outside their daily activities, and therefore not have the opportunity to take the diverse array of photos that many adults do.

The researchers also found a clear difference between the two groups in terms of the types of topics they engage. For teens, more than half of photos posted fall under "mood/emotion" and "follow/like" topics. Those topics are not necessarily tied to the content of photos but rather describe one's emotional status or the intention to have more followers. On the other hand, adults show a higher ratio of posts falling under more diverse topics, including "arts/photos/design," "locations," "nature" and "social/people."

"This may be due to the fact that teens are financially and culturally dependent on their parents to venture outside of their daily activities compared to adults," the researchers wrote in their paper.

While teens may not post as many photos as adults do on Instagram, Shih and Han said, they appear to be much more conscientious about how they portray themselves. Through photo content analysis and calculating the number of photos with tags and those with "selfie-tags," the researchers found that teens post more selfies than adults do. In addition, teens tend to manipulate their photo content to receive as many "likes" as possible, and remove photos with too few likes.

Teens also tend to be more verbose in their bios, the researchers reported, and to actively advertise for others to follow them. Teens tend to have more likes, tags and comments and to be more expressive about themselves and their photos. According to the researchers, the results support the idea that teens see social media as a place for self-representation.

These results, Shih and Han said, provide an opportunity to extend their study to other social media sites to validate their method and compare results, since many Instagram users provide additional social media links in their profile. Another possible direction would be to examine generational differences. For example, Shih said, do members of Generation X behave differently on social media than Millennials do?

The researchers are also interested in examining whether, for example, Millennials' social media habits will change when they reach a certain age, or are enduring characteristics related to the cultural moment they grew up in.

"Our interest is in seeing how human behavior develops over time in relation to different social media platforms," Han said.