Using a single item scale to measure online and offline
behaviour was most appropriate for a number of reasons.
Firstly, gendered behaviour had to be measured twice (online
and offline), thus using an extensive multi-item scale would
have been difficult because of its repetitious nature for
respondents. Secondly, data was gathered in street intercept
face-to-face interviews. Therefore it was paramount to keep
interview length to a minimum to ensure high completion
rates and to maintain respondents’ interest in the interview.
Thirdly, whether gendered behaviour is best represented in
a single or multidimensional space has not yet been conclusively
shown (Bettany et al., 2010). Smiler and Epstein
(2010: 133) in their conceptual and methodological review
of measures of gender in that they argue that there are a
‘plethora of measures’ and ‘substantial disagreements
about measurement’. They cite Beere (1990) who reviewed
over 1400 different measures of gender. MacLaran et al.
(2004: 165) also concurs arguing that ‘gender has no single
and universally agreed meaning. Gender theories are
informed by different, and often irreconcilable, assumptions.’
Lastly, single item scales have been shown to measure
constructs as well, and sometimes even better than, more
elaborate multi-item scales (Wirtz and Lee, 2003). Therefore,
a single item scale was used to measure gendered behaviour
online and offline.