Mixed use of big data and survey data for media audience measurement in France : an overview

Hybrid methods, big data, sample surveys

Meet the scientific direction of Médiamétrie at the 8th Conference of the European Survey Research Association in Zagreb (15th-19th July 2019)

Founded in 1985 in response to new demands in the French audiovisual sector, Médiamétrie is now the benchmark when it comes to audience measurement of audiovisual and digital media. Médiamétrie’s main mission is to measure and analyse the behavior of the audience and the trends in the market. In a context of ever-wider media offerings and increasingly complex media consumption modes, Médiamétrie is always on the lookout for new trends to design and develop new methodologies and technologies for audience measurement.

Audience measurement was originally based on panels of individuals. The emergence of return path data has provided the media ecosystem with new sources of data that are comprehensive and available in near real time. Rather than oppose big data and survey data, Médiamétrie has chosen to develop hybrid approaches that mix the different data sources to create a new, richer or more detailed one.

Two types of approach have been developed.

The first one is based on a basic principle in survey sampling theory: the use of any auxiliary information related to variables of interest improves the statistical precision of the results. Big data is considered here as auxiliary information in the same way as Insee census data.

The second approach responds to a different need. Actually, the survey sample may be too small to estimate audience results of small TV channels or websites. Big data then brings the necessary deepness and becomes the reference data and survey data is used as training data set in the construction of a qualification model.

The purpose of the presentation is to discuss the strenghts and weaknesses of each source, especially with regard to the new European General Data Protection Regulation, and to provide an overview of the hybrid methods developed by Médiamétrie since 2012.

Calcul d’intervalle de confiance à 95%

Taille de l'échantillon ou d'une cible dans l'échantillon

n =

Proportion observée dans l'échantillon ou sur une cible dans l'échantillon

p =

%

Attention : ne s'applique qu'à une proportion. Le Taux Moyen est une moyenne de proportions et la Part d'audience un rapport de proportions.
Cet outil est donné à titre indicatif. Il ne saurait pouvoir s'appliquer sans autres précautions à des fins professionnelles.

Test de significativité des écarts entre deux proportions

Permet d'évaluer si la différence entre 2 proportions est significative au seuil de 95%

Proportion

Taille de l'échantillon

Échantillon 1

%

Échantillon 2

%

Attention : ne s'applique qu'à une proportion. Le Taux Moyen est une moyenne de proportions et la Part d'audience un rapport de proportions.
Cet outil est donné à titre indicatif. Il ne saurait pouvoir s'appliquer sans autres précautions à des fins professionnelles.