TV Commercial Variety vs. Cost of Advertising Part 2

With an understanding of the methodology I would use on the samples in place, I then had to decide from where to take the samples such that it would be unlikely for a confounding variable to affect the percentage of commercials containing celebrity appearances in a way that would prevent the drawing of conclusions from the data.  The most important inclusion would be samples from programs that draw very high ratings and programs that draw low to average ratings.  The most obvious representative of the high ratings group is the most viewed television program in America every year: the Super Bowl.  I also included several games from the NBA Finals, which have significantly lower ratings than the Super Bowl but still have much higher viewership than average television programs.  Those selections were the most readily available for my research because the NBA Finals occur during the summer and the Super Bowl commercials have become a sensation in their own right such that they get posted on the Internet for posterity on web pages titled “All Super Bowl LI Commercials”.

Recognizing that both of my selections to represent programs with high ratings were sporting events, I ensured that the subject matter of the television programs would not become a confounding variable in my data by sampling from the daytime programming of a sports channel, ESPN.  I also took from several other genres: late night dramas (CBS), game shows (FOX), news shows (MSNBC), science-related programs and documentaries (Weather Channel), and lifestyle shows (HGTV).

For further research, I would select low demand live sports like midday baseball games for further comparisons to the high demand live sports data.  A separate study could be done on the possible disparities between national programming and regional programming.  I am very interested as well in collecting data from all types of sports coverage and searching for patterns.