Data collection
Each website was visited twice across a 1-month period
with site features and content recorded according to a
pro-forma. The pro-forma was designed based on previous
tobacco research [27] and following consultation
of the extant literature, then adapted to suit the UK
alcohol market and brand websites. This recorded
information on age restricted entry or messages, headings,
site sections and content, games or other interactive
activities, brand databases, special features,
competitions and links to other websites. Inter-coder
reliability checks between two researchers were conducted
to ensure consistency in coding, using joint
probability of agreement, with a minimum level of
agreement set at 85% and any inconsistencies discussed
and resolved upon discussion [28].