• by Joe Mandese  @mp_joemandese, July 8, 2005

In what is likely the most radical development in TV audience measurement since the introduction of people meters, Nielsen and a small industry steering committee have decided to change the way Nielsen prompts people to register their viewing, abandoning the concept of channel tuning and moving to a time-based prompt that would interrupt TV viewers every 42-minutes. The move has the potential to impact ratings for TV shows and networks, as well as how Nielsen households watch the medium of television itself.

Nielsen has been contemplating the change for several years - ever since it began developing the concept for the so-called A/P meters it is beginning to rollout to measure viewing done via digital TV platforms. The A/P meters, which stand for active/passive, utilize a digital code embedded in TV shows. But unlike its older generation of people meters, which prompted people to register their viewing whenever they changed a channel or turned on their set, Nielsen said digital technologies have rendered the concept of a channel "obsolete."

The problem for Nielsen and the steering committee is that there has been no solid research on what the ideal duration of time should be to interrupt TV viewers to measure their ratings.

Nielsen conducted qualitative research before making this decision, and canvassed data in other parts of the world where time-based prompts are used, but said they did not yield any results on the methodological impact of time-based prompts.

However, the committee, and Nielsen appear to have resisted TV industry pressure to utilize an even shorter duration prompt - one that would interrupt viewers every 21 minutes - which many believe would have captured more viewing and registered higher ratings, but which some also feared could negative impact actual viewing behavior.

Nielsen said the first meters to use the new time prompts will be installed in its national and local people meter samples later this month, and that it would continue to conduct research on time effects of time prompts.

"The next steps are to increase the size of the test samples to facilitate the analysis of key population segments, and to look into differential prompting intervals," Nielsen said.