Skip to main content icon/video/no-internet

Cell Phone Sampling

The rise of personal cell phone ownership in many industrialized countries and, more important, the increase in the number of people who can be contacted only via cell phone poses some challenges to traditional telephone surveys. Some of the sampling techniques used for selecting traditional landline (wired) telephone samples still apply when selecting cell phone samples. There are, however, specific characteristics of the cell phone that impact frame construction and sample selection that should be incorporated into designs to maximize yield from cell phone samples. The sampling issues will vary by country as a function of differing cell phone penetration rates, numbering taxonomies, and local market conditions, including technology and plan attributes. Designs for cell phone sampling and weighting, along with a general consensus for their use in practice, are currently continuing to emerge within the survey research community. Based on a query of cell phone systems worldwide, it does appear that the cell phone situation in the United States has a tendency for more complexities. The solutions for other countries may be much simpler versions of these designs.

The New Phone Subscriber Population

The cell phone subscriber population is expanding worldwide and is rapidly changing telephone systems and how people communicate within them. In some countries, the ratio of cell phone subscribers to total residents is quickly reaching a 1:1 ratio. Only 15 years ago, these ratios were in the range of 1:20 to 1:10. Table 1 summarizes the penetration rate of cell phones in selected countries (unadjusted for multiple cell phone ownership) collected by the International Telecommunication Union in 2005. Comparisons between countries should be made carefully due to variations in age distributions within different countries, since age is associated with cell phone ownership. The table gives an idea of potential undercoverage biases that may result in samples of landline phones that exclude cell phones.

Table 1 Cell phone penetration rates by selected countries, 2006
Australia97
Austria103
Belgium93
Canada53
Denmark107
Finland108
France85
Germany102
Greece100
Hong Kong131
Italy124
Japan79
Netherlands97
Portugal116
Russia84
Spain106
Sweden105
Turkey71
Taiwan97
U.K.116
U.S.77
Source: International Telecommunication Union (2006).

The percentage of cell phone numbers (CPNs) to total inhabitants generally overestimates the number of unique users as reflected by the reality that multiple numbers may be used by a single subscriber. Thus a sampling frame of CPNs may have a problem of multiple listings for some individuals, thereby increasing the probability of selection for those subscribers with multiple CPNs. Another phenomenon that has direct impact on telephone surveys in general is masked in Table 1: In many countries the number of people dismissing a landline or not having one in the first place is also rising. Currently, it is not unrealistic to predict that, in the near future, in some countries everyone could potentially be reached more easily via a cell phone than by a landline phone.

Diversification of Telephone Sampling Frames

As a result of the new presence of cell phone subscribers, the telephone subscriber universe as we know it is changing and can be best described in four parts: (1) cell phone only (CPO), (2) landline only (LLO), (3) cell and landline (C&L), and (4) no phone service of any kind (NPS), as depicted in Figure 1.

...

  • Loading...
locked icon

Sign in to access this content

Get a 30 day FREE TRIAL

  • Watch videos from a variety of sources bringing classroom topics to life
  • Read modern, diverse business cases
  • Explore hundreds of books and reference titles

Sage Recommends

We found other relevant content for you on other Sage platforms.

Loading