Sound Statistical Principles

COMMB uses the highest statistical research standards in its measurement methodologies because even the slightest deviation from statistical norms can create unreliable results. These statistical standards include:


Expansion factors

Outdoor

Statistically-reliable factors are utilized to adjust raw traffic and pedestrian counts to traffic volumes that represent an average day of the year. These factors take into account traffic fluctuation on various days of the week, at different times of the day, in different months or seasons of the year, and in different land-use areas such as residential, industrial or commercial. Extensive traffic studies conducted by municipalities, road authorities and traffic engineering firms determine the various factors and how they are applied.

Indoor

In venues such as restaurants, bars, malls, fitness clubs and university-college campuses, expansion factors are determined using regression analysis of full-week traffic counts. These reflect how pedestrian traffic levels change depending on the time-of-day, day-of-week, month/season, type of urban area and hours of operation. They are utilized to convert in-field traffic counts to average weekly circulation.


Methods of averaging

Methods of averaging, used to produce average daily or average weekly circulations, are dependent on the nature and complexity of the various OOH networks. The number of advertising faces located in indoor place-based venues, for example, varies considerably from establishment to establishment, dictating the need to use weighted averaging when calculating average weekly circulation per face in a given market. Weighted averaging takes into account the proportion of faces in each establishment rather than treating each establishment equally.

The COMMB Research Committee uses its expertise to determine the appropriate averaging method for each methodology.


Sampling

COMMB conducts many different research studies to deliver reliable traffic counts, factors used in circulation formulas, and audit quotas. Sampling for all studies is conducted to obtain the highest degree of reliability, maintaining a 95% confidence level and a margin of error of under 10%.