One of the strengths of MSG’s new Advanced Cellular Frame is its ability to address inward and outward migration of cellular telephone users.  This feature allows for the employment of Disproportionate Stratified Sampling (DSS) designs – or in other vernacular – split-frame designs. We can illustrate this using a typical scenario.  

Say you are targeting a metro area like Multnomah County, OR – home to the city of Portland.  There are, in essence, three ways you can define a cellular telephone sample frame targeting Multnomah County. 

One method is using a traditional EPSEM (equal probability of selection method) RDD design.  This is the method that has been around for well over 10 years.   We first identify the rate centers that best fit Multnomah County.  Every possible telephone number in the cellular thousand blocks (first 7 digits of a telephone number) that originate from the selected rate centers make up the sample frame.  This RDD frame design contains a mixture of working and non-working/unassigned telephone numbers. 

Multnomah County, OR
RDD Frame (Traditional EPSEM)	2,057,000	100.0%
Listed In-Area	348,497	16.9%
Listed Out of Area	804,686	39.1%
Unlisted/Unassigned	903,817	43.9%

This method, however, has some serious drawbacks, but more importantly, it does not address migration – people who have moved in and out of Multnomah County and kept their cellular telephone number.   Look at the highlighted line in the table above.  39% of the RDD EPSEM frame is known to be outside of Multnomah County.  Additionally, the RDD EPSEM frame excludes anyone who has since moved into the county from other parts of the country.  The result of this behavior is high under coverage and high out-of-area.  It’s simply an inefficient and costly design. 

The Advanced Cellular Frame dramatically improves upon the shortcomings of the traditional RDD design.  We can still employ an EPSEM design by including both known and unknown cellular telephones.  But, this design addresses the migration issue.   Look at the highlighted row in the table below.  This design now includes cellular numbers for individuals who have since moved into Multnomah County from other parts of the state or country (inward migration).  It also excludes cellular numbers known to be outside Multnomah County (outward migration).  This all but erases the inefficiencies of dialing known out of area out of area numbers while at the same time improving in-area coverage.  This design is still an EPSEM methodology because it includes the unlisted/unassigned portion in order to provide full coverage of Multnomah County.  When sampling, every telephone number has an equal probability of selection.  

Multnomah County, OR
ACF Frame (EPSEM)	1,449,089	100.0%
Listed	545,272	37.6%
In Area	348,497	24.0%
Inward migration	196,775	13.6%
Outward migration	0	0.0%
Unlisted/Unassigned	903,817	62.4%

Building on the improved EPSEM design noted above is the fact that we can now employ Disproportionate Stratified Sample designs (DSS) using the Advanced Cellular Frame.  We can separate the Listed and Unlisted/Unassigned components of a sample frame into individual strata.  This enables you to sample the more efficient listed stratum at a higher rate and under sample the less efficient unlisted/unassigned stratum at a lower rate.  Every number in the cellular sample frame still has a probability of selection but DSS will yield a more productive sample over a simple EPSEM design. 

Click Here to learn more about the Advanced Cellular Frame today!