Archive for May 12th, 2023

Is Less the Merrier for Ad Exposure and Audience Attention? A Media Measurement Case Study

Abstract:

Advertisement-supported media models are built upon an implicit agreement that media publishers make with their audience: the price of free or subsidized content is the audience’s attention to advertisements. A key component in the evaluation of an advertiser’s return on investment is whether an advertisement held the audience’s attention well enough to generate a conversion (e.g., a sale, website/store visit, newsletter sign-up). In our modern era of near-constant access to information, arguably one of the greatest threats to audience attention is information overload. This threat is likely to increase in severity as ad content increases in quantity. Recent research by Abcarian and Rao (in progress) has demonstrated an increase in the number of ad spots generated by media companies over time. How do audiences react to increased ad exposure? Is there an optimal ad duration that elicits higher levels of audience attention? This study seeks to answer these questions. This study adds to the growing body of research on audience attention in the context of ad exposure measurement. Preliminary results indicate that shorter-duration ads (15 seconds) are more effective in garnering audience attention than ads of longer duration (30+ seconds). Our paper includes a more extensive discussion of our findings along with recommendations for future research.

Recommended Citation:

Nandi, S., Craft, M., & Rao, K. (2023). Is Less the Merrier for Ad Exposure and Audience Attention? A Media Measurement Case Study. American Association for Public Opinion Research, Philadelphia, PA.

Attached Documents/Links:

  • AAPOR 2023 Program
  • For a copy of this presentation, please send me a comment with your email address in the box below or an email to kumarrons at gmail dot com.

Latent Class Analysis: A Probabilistic Approach to Uncovering Latent Classes of TV Advertisements

Abstract:

In media measurement, examinations of relationships between audience behaviors (e.g., viewership of TV advertisements) and demographics (e.g., age and gender) are common practice. These examinations can follow a variable-centered approach that differentiates variables into dependent and independent variables (e.g., linear regression) or a person-centered approach which groups observations together according to their configural profile of characteristics (e.g., K-means; Morris & Kavussanu, 2008). This paper extends the boundaries of media research by using a latent clustering method — Latent Class Analysis (LCA) for analyzing advertisement viewership data. LCA is a probabilistic mixture modeling technique used to identify a set of discrete latent classes of observations, in this case advertisements, based on the values of a set of categorical indicators. The technique focuses on generating classes of advertisements that share similar patterns of responses and comparing with other classes to subsequently determine how they are differentially related to predictors and outcomes (Wang & Hanges, 2011).

The dataset used in our analysis is characterized by extreme volume (7M rows) and a variety of variable types (variables are a combination of continuous, categorical, and count). Preliminary findings indicate the classification of advertisements into two latent classes as defined by a set of categorical indicators of interest. The most salient features of Class 1 in comparison to Class 2 are a greater likelihood to air within News/Information and Sports TV genres as well as a greater likelihood to air on a weekday as opposed to a weekend or weeknight. We conduct additional analyses to understand how these latent classifications are related to a measure of what the media world refers to as co-viewing. This paper concludes with a discussion of the results and their implications along with recommendations for future research. R code and a sample dataset is provided for interested readers and practitioners.

Recommended Citation:

Craft, M., & Rao, K. (2023). Latent Class Analysis: A Probabilistic Approach to Uncovering Latent Classes of TV Advertisements. American Association for Public Opinion Research, Philadelphia, PA.

Attached Documents/Links:

  • AAPOR 2023 Program
  • For a copy of this presentation, please send me a comment with your email address in the box below or an email to kumarrons at gmail dot com.

Does Ancillary Data about Ancillary Data Help Treat Missing Responses in Ancillary Data?

Abstract:

Consumer-file marketing companies collect and sell customer’s demographic and behavioral information wherever they can: through loyalty cards, website visits, social media posts, etc. These ancillary data are tremendously valuable and have a variety of applications in the field of television and digital media measurement that include targeting households or individuals for marketing campaigns, adjusting for nonresponse bias, and projecting estimates from a sample to an underlying universe. These potential advantages, however, depend on the completeness of the information contained in the data. Today, most applications involving the use of ancillary data exclude incomplete records from the substantive analysis. Exclusion as a missing data handling strategy has been known to result in inaccurate conclusions depending on the nature of the mechanism assumed to have generated the missing data. In this paper, we demonstrate a model-based alternative to exclusion, called multiple imputation (MI), which is used to fill in the missing values. A correctly specified imputation model preserves important characteristics of the data (e.g., variances, correlations, interactions) and can mitigate any bias resulting from exclusion-based missing data handling strategies. Our paper demonstrates MI in the context of a substantive analysis of the relationship between a completely observed measure of the number of advertisement impressions served across television networks and an incomplete measure of the number of children per household. We include likely correlates of the suspected cause of the missing data in the imputation model to evaluate whether their inclusion improves the quality of the imputed values. These correlates are measures pulled from a second, more completely observed, ancillary dataset containing television and digital media audience measurements. This paper concludes with a discussion of our findings, which are likely to have important implications for the use of ancillary data in television and digital media audience measurement.

Recommended Citation:

Craft, M., Raghunath, A., & Rao, K. (2023). Does Ancillary Data about Ancillary Data Help Treat Missing Responses in Ancillary Data? American Association for Public Opinion Research, Philadelphia, PA.

Attached Documents/Links:

  • AAPOR 2023 Program
  • For a copy of this presentation, please send me a comment with your email address in the box below or an email to kumarrons at gmail dot com.