Pew finds regular non-coverage bias in landline-only phone samples

Pew sounded the alarm in an update to its ongoing study comparing the results of surveys fielded to landline-only samples and combined landline and cellphone samples.

The latest report, published this week, analyses a host of surveys conducted in the second half of 2009 and in the first half of this year.

Of 72 questions examined, Pew says 43 show differences between the two samples of up to two percentage points. A further 29 differences are of three percentage points or more – “all of which are statistically significant”, says Pew.

“For some estimates, even a small amount of bias may have important substantive consequences for the political or social implications of the research,” say report authors Leah Christian, Scott Keeter, Kristen Purcell and Aaron Smith. “Since the decline of landline coverage has not been uniform across demographic groups, non-coverage bias among certain subgroups may be even larger than for the full sample.”

Pew gives the example of a survey of wireless internet use. In landline surveys, wireless usage is underestimated by two percentage points, but the bias is eight points among African Americans.

Other differences include a tendency towards more liberal views on social values and certain lifestyle behaviours in the dual landline and cellphone sample, while estimates of gun ownership are four points higher in the landline sample.

Pew’s research is available for download here. The latest report from the National Centre for Health Statistics, which revealed that the numbers of cell-only homes had risen from 22.7% in the first half of 2009 to 25% in the second half, was published last week.