It used to be that 97%+ of US households had a phone (meaning a landline), and a phone survey conducted using RDD (Random Digit Dialing) was considered a "random sample" because it gave you access to all those households that had a phone – meaning nearly every house in America. 

But times have changed.  Recent estimates indicate that 16% of US households now have a cell phone only – no landline.  While market research is exempt from Do Not Call legislation, the fact is that with 76% of Americans signed up for it, it has created a somewhat more “hostile” environment for survey calls. 

So we have what researchers see as a decidedly disharmonic convergence – legislative pressure + declining landline penetration – which has led to a now shaky definition of "random sample."  In light of all this, then, what is a random sample?  My answer, really, is that in the strictest statistical definition, it doesn’t exist.  At least not in a practical way. 

So what’s a researcher to do?  Get out the blender.  Mixed methodologies that combine data streams from different sources (phone, online, in person, etc.) is a solution that makes sense to researchers and clients alike.  In this spirit, we’ve teamed with Zoomerang, the the nation’s #1 online survey provider, to provide the on-demand research community with an easy way to reach a broader audience. 

Click here to learn more about how to get phone interviews from your online survey with BETi™, brower-enabled telephone interviewing.