Television Ratings

Jay James Ray Shannon

Class: Statistics.

Materials: Poll Sheets, data sheets, and calculators.

Goals: To examine the reliability of survey techniques using small samples to predict population trends. To enable students to become familiar with oral and written presentations in statistics.

Time required: Three class days. Initially use about 10-15 minutes for presentation of the problem to students (preferably on Friday to allow more time to gather data outside of class). Allow one day in class for write-ups and organization of materials and one day for presentations.

Background: Knowledge of frequency tables and related statistics.

Setting: Ratings of various sorts constantly make the news. Ranging from popularity polls for politicians to dietary habits of teen-agers, these polls not only make the news, they sometimes have a dramatic economic impact. Many polls claim to be sampling from the entire population of the United States, even though the number surveyed is often quite small. One rating scheme that has been criticized for its small sample size is the Nielsen Rating for television viewing.

Problem: You are to poll friends, relatives, neighbors, and/or members of the community about their 5 favorite television programs. Then use the data to predict your community's TV program preferences and compare the resulting data to current Nielsen Ratings which are used to determine the charges for TV advertising. In your opinion, is this a fair system? Support your opinion. Each group of students will present an oral and a written opinion with support.

Evaluation: Points may be awarded to each group as follows: 10 points for the written report with emphasis on their assumptions and 5 points for the oral presentation. Bonus points may be awarded at teacher discretion for outstanding work and effort.

Extensions: The students might investigate TV program preferences of grade levels, age groups, boys vs. girls, adults (20 and up) vs. teenagers vs. preteens, and so on. You may wish to collect the data from each group and compare the conclusions with those of each group.

Teacher Notes: Encourage students to be thoughtful as they collect data. For example, if data were collected at the local barbershop, the conclusions might be biased toward men; if data were collected at a local grocery, the data might be biased toward a certain socio-economic group. Students need to be cognizant of the many factors that influence their decisions. The written presentation should include: 1) a statement of the problem, 2) how the problem was approached, 3) all calculations performed, and 4) a clear and concise statement of conclusions. The oral presentation should present the written report to the class; It should not merely be read aloud!

Television Ratings

Sample Solution

Project Description: We were assigned to study the validity of the Nielsen Ratings System in our community. We used the poll sheet provided to survey a sample of 19 people. We found that for first place Seinfeld received 5 votes, Home Improvement and Star Trek - The Next Generation tied with 3 votes, Murder, She Wrote received 2 votes and the other nine shows received 1 vote. So we tried various ways to make adjustments as shown in our calculations.

Assumptions:

1. First place votes will be the only data used to determine the most favorite shows. We assigned 5 for each first place vote, 4 for each second place vote, 3 for each third place vote, 2 for each fourth place vote and 1 for each fifth place vote to make our frequency table.

2. The people we sampled are representative of our community as a whole.

3. We had chosen enough popular TV shows that our data would not be skewed by anyone who forgot a show he/she would have chosen had it been listed.

4. Each survey would reflect a response as directed.

5. Each vote represents a family of four that watches that particular show.

6. The 1990 census of our community which is listed in the Rand McNally Atlas as 19,760 has not changed.

7. Each person surveyed watched the show he/she voted for each week.

Calculations:

Conclusions: There are several ways that statistics can be used to show the top - 5 television shows. In just our small sample we found that using different approaches gave us the following:

A. The top five shows by first place rating only are:

1. Seinfeld - 5

2. Home Improvement - 3

2. Star Trek - The Next Generation - 3

4. Murder, She Wrote - 2

5. nine shows tied with - 1

B. The top five shows by assigning values to votes and collecting points are:

1. Seinfeld - 48

2. Home Improvement - 35

3. Murder, She Wrote - 27

3. Frasier - 27

5. Coach - 25

C. The top five shows by the number of people watching are:

1. Seinfeld - 13,436

2. Home Improvement - 9,287

3. Coach - 8,299

3. Jeopardy - 8,299

5. four shows tied - 7,311

Our comparison with the Nielsen=s is somewhat favorable as the July 11 - 17, 1994 top five ratings were in order:

1. Home Improvement

2. Grace Under Fire

3. Major League Baseball All-Star Game

4. Seinfeld

5. Frasier

However, it appears that if advertising rates are based locally on the Nielsen=s, there is some obvious over-charging for certain shows and some obvious underCcharging as well. It is the recommendation of this group that more extensive sampling be done in the future.

Television Show Popularity Poll

Sample Polling Sheet

Please number your 5 favorite television shows in order from 1 (most favorite) to 5. You may add to the list if your favorite programs are not listed. Thank you for your time.

Funded in part by the National Science Foundation and Indiana University 1995