I'd like to understand any trends around how self-harm or suicidal behavior posted on social media changed from 2007 until now.
Hello! Thanks for your question about how posting on social media about self-harming or suicidal behavior has trended over time. The short version is that after searching extensively through dozens of peer-reviewed articles, we’ve determined that the information you requested is not publicly available because the major social media platforms forbid "glorifying" self-harm, forcing their users to employ constantly-shifting and ambiguous terms to avoid censorship. As a result, it is impossible for any researcher to pull the data needed to discover these trends. However, we were able to learn that all researchers agree that teens will overwhelmingly seek out peers via social media rather than seek adult help for their desire to self-harm.
Below you will find a deep dive of my research and methodology.
We began by reviewing our colleague's first report to find out where the misunderstanding had occurred and see if any of the sources used were pertinent to your true question. Using the reference to Megan Moreno's work in this area as a starting point, we conducted a study of scholarly articles to find out exactly what information was available on this subject. The majority focused on the general social effects of social media or cyber-bullying rather than the subject at hand, but we still found numerous pertinent articles. Many of these were available only as abstracts or previews, but we still cite them where they yield useful data.
Other than to point us towards more scholarly material, we did not search for popular media articles on the subject. We also did not look into any possible link between search term trends and suicide, as this was out of scope for your question and the intent behind it.
A LACK OF STATISTICAL DATA
We found literally dozens of articles written on this subject, spanning over a decade. However, there were no studies which even attempted to provide a statistical analysis of the number of social media posts related to self-harm or suicide over time. One recent study (from 2016) of the posts of Twitter users who attempted to take their own lives notes, "This kind of data has not previously been available in sufficient quantities or at this granularity" and so the study attempts only "broad intuition and interpretation of trends, rather than testing specific hypotheses."
One major reason why this data is simply unavailable is because of the attempts of the social media platforms themselves to curb self-harm, possibly influenced by earlier studies which surfaced primarily the negative possibilities of those contemplating or engaged in self-harm having unmoderated contact with each other in an online forum. As a result, though there are more generalized studies on how conversations in social media about mental health have changed over time, those attempting to study self-harm must navigate an ever-changing vocabulary of ambiguous hashtags.
This has not stopped concerned researchers from conducting studies, of course, but such studies are snapshots of time, either examining a few thousand posts or surveying willing forum participants. The only statistical analysis over time that we could find is a graph showing that those who have already attempted suicide use fewer emojis in their tweets both before and after their attempts than the general population. This means that it is impossible to prove that the use of social media as a vehicle for expressing thoughts of self-harm or suicide is actually on the rise, except insofar as the use of social media more generally is on the rise.
However, for the purpose of promoting SocialSentinal, we do not believe that this matters. The key point that you need to prove in order to promote your solution to school districts is that far more students will discuss thoughts or actions involving self-harm with peers via social media than will seek out an adult's help--and this is something that every study agrees on!
Megan Moreno explains, "If you're cutting, it can be stigmatizing. But teens in my clinic go on Instagram and use hashtags because it provides them a key to others just like them." At the time of her study, the hashtag #MySecretFamily, which "described the broader community of NSSI [non-suicidal self harm] and mental illness," grew from 900,000 search results to 1.5 million.
Another aspect that needs to be considered is the near-impossibility of a parent or teacher discovering a teen's self-harm material on their own. While Pew shows that the most popular platform from ages 18-29 is Facebook (88%), Instagram is favored by more in this age bracket than any other age (59%). More importantly, teens use different platforms for different purposes: Snapchat is useful for saying something that you don't want to risk regretting later (and so impossible to use for your purpose), while Instagram is used for communicating with friends and family. Teens frequently have multiple social media accounts, and 83% of those considering self-harm post anonymously.  In fact, faces are only shown on 18.8% of self-harm-related pictures, and only 0.7% identify their environment.
The need to automatically detect posts about self-harm is clear, and at least one study attempts to establish a framework for doing so. This paper is behind a paywall, so we could not read more than the abstract, but is inexpensive and may be useful to you.
POSSIBLE CONTINUED STUDY
Due to the manifest lack of statistical data, and the near-impossibility of collating it, we do not believe that further research down that track will be fruitful. A deeper dive into the mindsets of teens discussing self-harm online or pulling together a body of case-studies may be more useful for your purposes.
To wrap it up, after searching extensively through peer-reviewed scholarly literature on this subject, a direct answer to your question is not publicly available because the major social media platforms forbid "glorifying" self-harm, forcing those using them for that purpose to employ constantly-shifting and ambiguous terms to avoid censorship. As a result, it is impossible to pull the data needed to discover these trends. However, we were able to learn that all researchers agree that teens will indeed overwhelmingly seek out peers via social media rather than seek adult help for their desire to self-harm. If you’d like to continue research on any of the other topics I’ve outlined above, just let us know!
Thanks for using Wonder!