There are a number of different biases that can impact panel and census research and big data sets which utilize artificial intelligence (AI) or machine learning (ML). The attached Knowledge at Hand identifies over two dozen unique biases. It discusses how they might occur and what mitigation options researchers should consider to account for them.
Member Only Access
on the upper right of the web site or create an account to view this page. on the upper right of the web site or create an account to view this page.
Data sets for STBs and Smart TVs have, up until now, been separate from one another, which has hampered the ability to accurately assess viewing habits. This data would be very useful for planning, buying and optimizing ad campaigns. Since the two are complementary, combining them can help us form scaled, granular, TV tuning data sets that are more nationally representative. Luckily, a new report from the Coalition for Innovative Media Measurement (CIMM)—a subsidiary of the ARF, outlines best practices for combining set-top box data and smart TV ACR data.
Member Only Access on the upper right of the web site or create an account to view this page. on the upper right of the web site or create an account to view this page. on the upper right of the web site or create an account to view this page. on the upper right of the web site or create an account to view this page.
Establishing causality in the TV advertising-to-sales relationship, ideally, is studied through randomized trials, which match respondents to exposure and non-exposure groups. Given the difficulties of such experiments, analysis of observational data, alternatively, can reveal links between ad campaigns and sales. But new research shows how failure to control for variants can produce misleading signals to advertisers.
Member Only Access on the upper right of the web site or create an account to view this page.