Sit Back, Relax and Enjoy the Commercials

1
Sit Back, Relax and Enjoy the Commercials

• Jenny Clift
• Gena Omelyaneko
• Tori Langan

2
Background

• The first TV commercial was broadcasted on July
  1, 1941
• It was on the New York station WNBT
• The ad was ten seconds long and advertised a
  watchmaker
• The cost of a commercial depends on the number of
  viewers
• The cost of a 30 second ad during the 2010 Super
  bowl, one of the most watched programs, cost 2.6
  millions dollars

3
Procedure

• Watched TV commercials on different channels
  during different times and recorded their types
  and how many for each
• Watched Food network and ABC to compare number of
  food commercials for each
• We each watched one network (ABC, NBC, or CBS)
  and cited the types of commercials and their
  numbers
• Collected all of the recorded female commercials
  for both daytime and primetime (on NBC) and
  compared

4
Assumptions for 1st Test

• State

• Check

• 2 independent SRS
• n1p1 10
• n1 (1-p1) 10
• n2p2 10
• n2(1-p2) 10
• 3. pop. (1) 10n1
• pop. (2) 10n2

• 2 independent SRS
• 54(.037) 10
• 54(1-.037) 10
• 111(.27) 10
• 111(1-.27) 10
• 3. pop. (1) 10(54)
• pop. (2) 10(111)

Not all assumptions pass, but we will continue
with the test.
5
Food Channel vs. ABC

• 2 proportion z-test
• Ho P1 P2
• Ha P1 gt P2
• z3.6529
• p(z gt 3.6529) 1.2967 x 10-4
• We reject Ho because our p-value lt a0.05. We
  have sufficient evidence that the proportion of
  food commercials on Food network is greater than
  the proportion of food commercials on ABC.

P1 Food Channel P2 ABC
6
Visual Statistics, Analysis, and Conclusion (1st
Test)

• There were overall a greater number of
  commercials recorded on the Food network
• The Food network did, in fact, have a larger
  number of food commercials than ABC did
• We were surprised by how little food commercials
  there were when compared to the total number

7
Assumptions for 2nd Test

• State

• Check

• 2 independent SRS
• n1p1 10
• n1 (1-p1) 10
• n2p2 10
• n2(1-p2) 10
• 3. pop. (1) 10n1
• pop. (2) 10n2

• 2 independent SRS
• 73(.068) 10
• 73(1-.068) 10
• 83(.096) 10
• 83(1-.096 10
• 3. pop. (1) 10(73)
• pop. (2) 10(83)

Not all assumptions pass, but we will continue
with the test.
8
Female-based commercials on NBCDaytime vs.
Primetime

• 2 proportion z-test
• Ho P1 P2
• Ha P1 lt P2
• z-.6289
• p(z lt -.6289) .2647
• We fail to reject Ho because our p-value gt
  a0.05. We have sufficient evidence to say that
  there is an equal amount of female commercials
  during the daytime block and during the primetime
  block.

P1 of female commercials during daytime P2
of female commercials during primetime
9
Visual Statistics (2nd Test)
10
Analysis and Conclusion (2nd Test)

• The bar graphs for the primetime and daytime were
  almost identical in ratings
• Although the primetime block had more commercials
  shown, the proportions between the nos and the
  yess for both times was very similar

11
Assumption for 3rd Test
State
Check

1. SRS
2. All expected counts gt (or equal to) 5

1. Not SRS
2. Not all counts gt (or equal to) 5

Not all assumptions pass, but we will continue
with the test.
12
NBC, ABC, CBS

• Chi-square Test
• Ho The channel and the type of commercial shown
  are independent
• Ha The channel and the type of commercial shown
  are dependent
• (6-9)² (9-4.34)² 30.259
• 9 4.34
• P(? ² gt 30.259 / df 12) .0026
• We reject Ho because our p-value lt a0.05. We
  have sufficient evidence to say that the channel
  and the type of commercial shown are dependent,
  therefore some channels specifically show more of
  one type of commercial than others.

13
Visual Statistics (3rd Test)
14
Analysis and Conclusion (3rd Test)

• According to the overall bar graph for the three
  networks, NBC had at least ten more insurance
  commercials than CBS and ABC
• Another notable difference CBS had virtually no
  female-based commercials shown during the
  allotted time
• For the most part, the other and the other shows
  categories were fairly similar in all three

15
Bias and Error

• We only watched certain shows, and they were not
  picked at random
• The channels we watched were not picked at
  random(ABC, NBC and CBS) but they are the major
  networks
• Differences in half-hour shows and hour shows
  could have an effect on the results
• Another affect could be weekend vs. weekday
• A show we watched for two hours had an effect on
  the results because it was a golfing tournament
  and ¾ of the commercials were focused on men

16
Application to Population and Personal Opinions

• Our conclusion about this project was not as
  straightforward and accurate as it could have
  been.
• Watching at the same exact times as the other
  members of our group could have made the results
  more accurate.
• It also tended to get very boring watching
  daytime shows on the networks during the weekend.
  
• We assumed that during the daytime there would be
  more female-based commercials, and it ended up
  being an equal amount between the blocks.
• The populations results may or may not be
  different because of different networks across
  the state and country.

17
Conclusion

• All in all
• The Food Network does indeed have an increased
  number of food commercials over the average
  network (NBC)
• Female-based commercials rate equally in the
  daytime and primetime blocks
• And even though ABC, CBS, and NBC are all the
  top-dog networks and are similar in our minds,
  our studys results prove that the channels show
  more of one type of commercial over other types.