Early Bitcoin Surveys

I sporadically get asked for access to the primary data from two surveys I did with Bitcoin community members in 2013 and 2014, so for the sake of posterity (and future requests), here it is.

2013 Survey: full data, cleaned data

2014 Survey: full data

For context – during this period I was writing a dissertation on the Bitcoin community, largely from an anthropological perspective. To support my research with some numerical data, I conducted two online surveys, which were both open to Bitcoin users globally.

2013 Survey

The first survey was conducted in February-March 2013, and received 1,087 responses.

I created this survey on 12 February 2013 using Google Docs, and posted the link on Bitcointalk.org, Reddit, Twitter, and Google+. From there, the survey was reposted on various Bitcoin related sites. I posted preliminary findings to my blog after 102 responses and 533 responses. In both cases, the results sparked more interest and helped garner more responses. I closed the survey on 4 April after receiving 1,087 responses, so that I could clean up the data definitively.

The “cleaned data” have been through the following process:

Responses which gave a range (e.g. “300-500”) were converted to a mean figure (“400”). Responses which gave a limit figure (“less than 100”, “500+”) were converted to the limit figure (100, 500). I deleted anyone who had left four “major blanks,” ie they had not responded to four simple, short, numerical/multiple choice questions which applied to everybody. I deleted anyone who left the age or gender field blank. I manually removed a very small number of highly erroneous answers, such as a respondent who claimed to own one million bitcoins. After all that cleaning, there were 1,011 responses left, so for the sake of clean percentages I used the first 1,000.

2014 Survey

The second survey contained largely the same questions, received far more promotional effort from myself and acquaintances, was posted on all the same channels as my 2013 survey, and was kept open from February to June 2014, much longer than the first survey. Despite this, it only managed to get just over 400 valid responses. This second survey was also greeted with more suspicion and mockery, and more users solicited payment for their responses.

The difference between the two community responses hints at the transformation experienced by the community during the intervening year. During my 2013 survey each bitcoin was worth about $50, and during my 2014 survey each bitcoin was worth about $500. The tenfold increase in digital wealth appeared to change the way people viewed their relationship with the project. In addition, the twelve months between the two surveys saw a flurry of mainstream media attention and a number of other survey efforts, so there may also have been an element of research fatigue.

Due to improvements in the Google Docs interface there were fewer format fixes needed for the 2014 survey data.

Both surveys were dominated by respondents from North America and Europe, thus likely under-representing the importance that community members in East Asia (especially China) play in Bitcoin.