© VisionMobile 2014 | www.DeveloperEconomics.com/go
10
VisionMobile over a period of five weeks between October and late
November 2013. One to one interviews were conducted in November and
December 2013.
The online survey received over 7,000 responses, making this the largest
mobile developer survey to date. Respondents to the online survey came
from over 127 countries, including major app development hotspots such as
the US, China, India, Israel, UK and Russia and stretching all the way to
Kenya, Brasil and Jordan. The geographic reach of this survey is truly
reflective of the global scale of the mobile app economy.
The online survey was translated in 6 languages (Chinese, German, Japanese,
Korean, Russian, Spanish) and promoted by 60 regional and media partners
within the app development industry. As a result, the survey reached an
unprecedented number of respondents, globally balanced across Europe
(36.5%), Asia (32.1%) and North America (19.7%). The online survey also
attracted a significant developer sample from Africa (6.6%) and South
America (3.4%).
To eliminate the effect of regional sampling biases, we weighted the regional
distribution by a factor that was determined by the regional distribution
identified in our App Economy Forecasts (2013 - 2016) report published in
July 2013, as shown in the graph above.
The survey gathered responses from developers across 15 platforms
including Android, Bada, BlackBerry 5/6/7, BlackBerry 10, Chrome,
Facebook, Firefox OS, iOS, Java ME, HTML5, OSX (desktop), Windows
(desktop), Windows Phone, Windows 8 and Tizen. As our research focuses
on mobile developers, we have excluded from the analysis all respondents
that are not developing for mobile platforms.
To minimise the sampling bias for platform distribution across our outreach
channels, we weighted the responses to derive a representative platform
distribution. We compared the distribution across a number of different
developer outreach channels and identified statistically significant channels
that exhibited the lowest variability from the platform medians across our
whole sample base. From these channels we excluded the channels of our
research partners to eliminate sampling bias due to respondents recruited
via these channels. We derived a representative platform distribution based
on independent, statistically significant channels to derive a weighted
platform distribution.
By combining the regional and platform weighting we were able to minimise
sampling biases due to these factors. All results in the report are weighted by
main platform and region.