One in five Australians are not contactable via a landline telephone – What are the implications for your telephone survey estimates?
As in other parts of the world, one of the issues facing Australian survey researchers is the increasing proportion of households without a landline telephone connection.
According to a December 2011 report from the Australian Communications Media Authority, the proportion of adults without a landline telephone connection has more than trebled in the last 5 years –from 6% to 19%. The vast majority of these persons (96%) form the ‘mobile phone only’ population.
This growing trend in the proportion of ‘mobile phone only’ persons is being seen around the world. In the US for example, a National Center for Health Statistics survey found that in the first six months of 2011, three in 10 persons were only contactable via their cell phone.
Extensive international and emerging Australian research indicates that the non-coverage of mobile phone only persons via traditional landline sample frames is a source of non-ignorable bias, to which Australian survey researchers have been slow to respond.
The methodological response to this phenomenon is to undertake ‘dual-frame’ surveys. Dual-frame surveys include randomly generated landline telephone numbers and randomly generated mobile phone numbers in the sample frame to ensure a near complete coverage of the population (i.e. only missing the estimated <2% of people not contactable via either a landline or mobile phone).
The Social Research Centre (a privately owned survey research and social research organisation based in Melbourne) is the leading exponent of dual-frame surveys in Australia. The Social Research Centre has hosted two highly successful workshops devoted to this issue in March 2011 and July 2012 and undertaken several dual-frame surveys, including the first Social Research Centre Dual-Frame Omnibus Survey in December, 2011.
Among other things, the findings from the first Omnibus Survey showed that mobile phone only respondents are typically younger than landline users, more likely to be male, have lower socio-economic status, higher levels of participation in education, are more geographically mobile, and more likely to be Aboriginal or born overseas.
A higher proportion of mobile phone only respondents also reported mental health problems, binge and heavy drinking, smoking, legal and illicit drug use, problem gambling and financial hardship, and had fewer social supports.
Many of these initial findings reflect those in the US research literature and have been replicated in ensuing dual-frame studies undertaken by the Social Research Centre.
The findings reinforce the view that there are significant and non-ignorable biases in telephone surveys that rely solely on landline sample frames.
While major research organisations such as the Orygen Youth Health Research Centre, the Cancer Council Victoria and the Australian National Preventative Health Agency have all commissioned dual-frame surveys in recent times, this is still an emerging methodology.
To further expand knowledge in this area, the Social Research Centre is planning to conduct a second Dual-frame omnibus survey in early 2013. This second survey will enable the further exploration of this methodology and provide an opportunity for researchers to include key measures of interest to this omnibus survey so as to better understand the impact of the ‘mobile phone only’ population on their established survey estimates.
The survey will comprise 1,000 interviews with persons via a traditional landline RDD sample and 1,000 interviews with persons via a mobile phone RDD sample frame. A sample of these dimensions will yield approximately 250-300 interviews with the ‘mobile phone only’ persons.
While understanding that for some subscribers their questions and results may need to be treated as ‘commercial-in-confidence’ wherever possible we would like to share all of the data collected with all subscribers. Our preference is to allow subscribers to seek to be granted a royalty free, non-transferable, non-exclusive licence to use, reproduce, copy, circulate or publish findings from the survey.
The Social Research Centre will be responsible for coordinating the survey, compiling the survey questionnaire, data collection and data processing and producing and disseminating the final data set and support materials. A detailed methodological report will also be provided to each subscriber. The University of Queensland, Institute for Social Science Research (ISSR) will be responsible for obtaining ethical clearance for the survey via the UQ Human Research Ethics Committee.
Our intention is to share the learnings from this survey as widely as possible and to use the results for both research and advocacy purposes.
The first step is to ascertain whether there is enough support to make this second dual-frame omnibus survey viable. If you are interested in possibly including some questions please contact Darren Pennay by close of business 16 November, 2012.
The cost per question, assuming a national sample of n=2,000, will be $3,000 (excl GST) for a closed question and $3,440 (excl GST) for an open ended question.
To find out more please contact email@example.com or go to the news page on the Social Research Centre’s web-site (www.srcentre.com.au/news-events/news).
Darren Pennay is Managing Director and Head of Research Strategy at the Social Research Centre and Adjunct Professor at the Institute of Social Science Research at the University of Queensland.
Image: fd / flickr