Regarding yesterday’s Executive Comment from Charlie Ferguson of Northern Broadcast Inc., we received this response from Thom Mocarsky, Arbitron Sr. VP of Press and Investor Relations:
On randomly selecting radio station personnel: The sampling rate is so high in Traverse City that it¹s inevitable that households with radio station employees get selected.
With each survey, approximately on out of every 45 households gets selected to be in our starting sample. And, once the survey is done, one out of every 129 people in the market ends up in our survey.
Despite the odds of getting selected, people in the radio industry are remarkably honest. It is an uncommon event for us to re-issue a report because of media affiliation. (Only once this year, and we issued almost 800 separate quarterly reports.)
On keeping a list of radio station personnel: This would be a logistical nightmare and it raises significant privacy issues. (Remember, we track diary/PPM entries from nearly 14,000 separate radio stations and sample more than a million and half people each year.) We already spend a lot of time getting an up to date Station Information Package from stations. As I’ve said before, media affiliated respondents are not a common occurrence.
On software processing: While it is possible to "average the averages" in the Smart Plus software, there is a warning right on the front page after an agency signs in to get the data which cautions against averaging small market books.
Additionally, there is additional reference information regarding this in the SmartPlus help system. Also, when we began the two-book averages in select markets, we spent time educating the users of the software about the proper use of the data. This
is certainly something we can revisit.