New readership survey to be out on Tuesday

The Indian Readership Survey is used to buy & sell ads in the Rs 22k-cr print media industry

The new, overhauled Indian Readership Survey (IRS) will be out on Tuesday. That is when various stakeholders, such as publishers and media agencies, get a first look at the new IRS. The last time a readership survey came out was in December 2012.

IRS is a critical currency used to buy and sell advertising in the Rs 22,400-crore Indian print media industry. It measures the readership, viewership, etc, of newspapers, magazines, TV, radio, internet and cinema in India. It also offers a demographic map of Indian consumers.

There were, however, complaints about IRS - its method and ability to capture the heterogeneity of the market. For instance, magazines always suffered because the bias was towards newspapers, with their large numbers. To tackle many of these issues, in 2009, the Media Research Users Council (MRUC) joined hands with the Audit Bureau of Circulations (ABC). By 2011, it had set up Readership Studies Council of India or RSCI. In 2012, it gave Nielsen India the mandate to research IRS, instead of Hansa Research, which had been tackling it for ages.

While the big changes in what the data show are yet to be revealed, this time the survey, based on a sample of 235,000 people, aims to eliminate some of the major sampling and non-sampling errors in IRS. For instance, under the old system, it took 90 minutes to administer one questionnaire rendering the quality of response suspect. The new system does it in 40 minutes.

"The big enemy of any large survey is respondent fatigue at the end of 30 minutes," says Paritosh Joshi, chairman of the technical Committee for IRS and an MRUC board member. This has been tackled by using computer-assisted personal interviewing (CAPI) and data fusion.

Double-screen CAPI helps reduce interview time. The idea behind data fusion is to do a suite of studies, each with one set of questions, and then fuse the data. The data collection could happen from various samples. This gives a lot of data without increasing sample size and overloading the respondent.

To tackle data corruption, there is a new emphasis on security and integrity with geo-tagging, voice recording and live interviewer route-tracking to deter fraudsters. There are other major changes such as making average issue readership the only measure, instead of having options such as total readership.

Does the data have any surprises? Joshi doesn't say so but there are bound to be major upsets. Watch out for the fireworks on Tuesday.