Typically, users compare data from several NADbank Studies when evaluating newspaper readership. There are no hard and fast rules. Users can look at changes from year-over-year (NADbank 2008 versus NADbank 2007) or over a specific period, using several consecutive studies (NADbank 2005, 2006, 2007 and 2008). Generally, newspaper readership is expressed as incidence (percentage of adults 18+ who read newspapers) or number of readers.

The decision to use incidence or number of readers depends on the user’s objective. The key question is whether the chosen metric provides adequate information.

It is important to consider the following:

1. When evaluating changes in newspaper readership, it is prudent to consider all measures, incidence, number of readers, and changes in population over time.
2. The NADbank database is made up of a combination of full sample and blended or rolling sample and three-year data.
3. When sourcing NADbank data, clearly reference the source as NADbank and specify the year of the study. This avoids confusion and provides context for the reported data.
   1. Full sample data provides full year-to year comparison,
   2. Data for rolling sample markets represents an “average” over the two-year period.
   3. Three-year data is only comparable against, the previous three years. For example, data collected in 2008 is comparable to data collected in 2005.
4. Incorporate Margin of Error when evaluating changes in newspaper readership
5. Caution should be used when analyzing results where sample size is less than 40, for both incidence and number of readers.