Please scroll down if you just want to see the tweets. The context and method are described in detail here in case this information is useful for #SoMe4Surgery research.

This analysis is an attempt to bring together some rules that could be used to start analysis of #SoMe4Surgery activity on Twitter.

I have listed all tweets from 28 July to 12 September with at least 10 retweets (up to the point of the final extract on 12 September). This has produced a reasonably comprehensive overview of #SoMe4Surgery tweeting, but will sometimes point to other areas for exploration - eg sometimes a list of numbered questions will only include some of the questions (which is interesting in itself, as it allows reflection as to why some questions were more popular than others - timing? clarity of explanation? topic? etc). Judicious use of the Twitter advanced search function will help find tweets that are not included in this list - or you can find them in the NodeXL raw data.

I started extracting data via NodeXL using a search strategy that would include Julio Mayol's original tweet (first tweet below). More recently I have moved to searching just for #SoMe4Surgery as more complex search terms appear to sometimes muddle up access to the Twitter API.

I have ordered these tweets chronologically. The method is written up here. There is a period on 2 September where data were not available using NodeXL so I extracted data as a backup using TAGS (on 5 September). These data could be added to the summary, but in fact the most popular tweets from that period should be picked up by subsequent NodeXL extracts, as NodeXL records full data about tweets that are retweeted during the stated period of extract. Nonetheless, for completeness it might be worth double checking against tweets from 2 September.

As I have illustrated in a separate analysis, most replies to #SoMe4Surgery tweets do not use the hashtag. Tweets without the #SoMe4Surgery hashtag are not included in this current summary. They could, however, be pulled out manually for each of the individual tweets in this list, but that is a topic for another day and would involve a lot of time and effort (and may not be entirely accurate - eg if responses to individual questions did not use the hashtag and were not posted as replies to the original tweet).

The NodeXL extract on which this current analysis is based has 7,329 tweets by 456 tweeters with 14,341 retweets. Overall, 64 (14%) tweeters received 80% of all retweets while 76 (17%) received no retweets. This summary includes 253 tweets by 66 tweeters. The top 20 tweeters (by retweets) are listed below before listing the individual tweets.

Further information can be obtained via Followthehashtag (mainly geography and gender), with an extract completed for the same time period as this analysis in a tweet from 12 September.

There are a number of questions that emerge from the summary below:

1) Methodological challenges of gathering all the information that unfolded from the most popular tweets

2) The quantitative data that can be gathered from questions with straightforward options

3) The qualitative data that can be unpicked using thematic analysis

4) And others that colleagues can share in response to this summary...

Items