Look for more knocking on doors and larger samples as Arbitron advances it continuous improvement program for Portable People Meter (PPM) markets this year. The PPM sample improvement initiatives were announced last year and included in Arbitron’s settlement with the PPM Coalition – with scheduling for the remaining elements spelled out in the company’s monthly conference call with PPM clients and reporters for the February results.
The sample target increase begun in 2010, with an 8% increase implemented, meant that the sample targets for PPM markets and Designated Delivery Index (DDI) calculations were reset effective with the January 2011 survey. But the remaining two points of the 10% total increase is yet to come, with implementation scheduled by mid-year 2011.
A major change is in-person recruitment, which combined with moving to an address-based sample is expected to improve representation of minority households and younger demographics. When no phone number can be linked up to a computer-selected address and the household does not respond to a mailed screener survey, Arbitron representatives go out to physically knock on the door.
“The implementation did begin in 2010 in the high-density Black and Hispanic markets,” noted Denise Safko, PPM Project Manager. She said the field staff got new materials to help in recruiting panelists. “As a result of these efforts we’re seeing very good cooperation. We’re seeing very positive trends relative to our success at reaching and finding panelists,” she said.
The in-person recruiting effort started in high-density minority areas, but is scheduled to be deployed in all remaining PPM geographies in an August-December 2011 timeframe.
Address-based sample recruitment is currently used 100% in the Houston market and for reaching cell phone only and unlisted landline households in all other PPM markets. A switch to 100% address-based recruitment in all PPM markets is due to take place this year. That changeover is to begin in August and by December Arbitron says all new panelists will come from the address-based sample.
