• by Joe Mandese , August 25, 2009

TV ratings giant Nielsen has decided not to adjust its national TV ratings sample to ensure that it properly represents households equipped with personal computers or Internet access. The decision follows an annual review of the process Nielsen uses to determine what types of households and individuals to mathematically "weight" to ensure that they represent the population at large.

Nielsen historically has used the weighting process to adjust for an under-representation of certain kinds of households, especially harder-to-recruit ones such as minorities, to ensure that their viewing habits are properly represented in its national TV ratings.

But after conducting some testing, Nielsen executives concluded that adding weights for the presence of a personal computer or Internet access in under-represented households would provide "no significant change or enhancement" to its national TV ratings sample.

The decision is interesting, because Nielsen is pushing ahead with so-called "three-screen" measurement - research on how people consume video programming across TV, PC and mobile hand-held screens - as an as-yet-undisclosed industry consortium reportedly is mounting a rival measurement service that would specifically report audience estimates across those platforms.

It also comes as Nielsen has begun integrating a so-called "convergence panel" - a sub-panel of households reporting their Internet usage - into its national TV ratings panel.

Nielsen also looked at adding a new weight for the presence of high-definition TV sets, and for changing the weight it uses for digital video recorders to factor every DVR present in panelists' households, but decided not to implement those as well.

It said it would continue with its other existing weighting variables, and that it would begin adding them for race and ethnicity in some of its local people meter markets.

The changes affect Nielsen's ratings effective Aug. 31st.