The taskforce, which has been evaluating the reliability of online panels for the past 18 months, reported back this week with a warning that there is “no generally accepted theoretical basis from which to claim that survey results using samples from non-probability online panels are projectable to the general population”.

“Claims of ‘representativeness’ should be avoided when using these sample sources,” says the taskforce in its 81-page report.

However, it notes, “One special case is electoral polling where studies using non-probability panels sometimes have yielded results that are as accurate or more accurate than some surveys using probability samples.”

The taskforce – chaired by Market Strategies’ chief operating officer Reg Baker (pictured) – also acknowledges that there are times when a non-probability online panel “is an appropriate choice”.

“Not all research is intended to produce precise estimates of population values and so there may be survey purposes and topics where the generally lower cost and unique properties of web data collection is an acceptable alternative to traditional probability-based methods,” says the report.

Click here to download.