Statistics in the Media II

Surveys in Digital Media

The article Origin clash smashes ratings records on the Sydney Morning Herald website claims that State of Origin III in 2012 set a rugby league television ratings record as the “highest-rating rugby league contest in the history of OzTam.” But how does OzTam collect data on the number of people watching a football game on TV or watching any program for that matter?

OzTam, like other television ratings companies, did not count all the people whose television sets were tuned in to the State of Origin game. Instead, they made audience estimates using samples. How did they form this sample?

To form this sample, they initially conducted a large-scale survey to find out the characteristics of the population that was to be represented. Some respondents to this survey were then recruited to form a sample that reflected the demographics of the population and a metering system was installed on every TV set in their household to record their viewing activity, which includes information on who is watching, when they are watching, what they are watching, for how long they are watching and also information on the television audio signal. This information is then sent through the telephone line every night to be processed. OzTam’s production software then collates, processes and analyses the data from all households.

From this process, OzTam can track the number of people watching a particular program and can then project the result to the entire population. For example, if they wanted to know how many people watched the State of Origin game, they would simply count the number of people in the sample audience that watched the game, and then generalise from this sample audience to estimate how many people in the whole population watched the match. Not only that, they could also estimate the age, gender, city and other characteristics of the viewers.

Exercise

The article Making sense of the bewildering mess of US polling on ABC’s The Drum website talks about the disparate findings of polling organizations of US voters’ preferred choice for president. Read the article and answer the following questions.

1. Investigate the methodology used by Gallup to come up with their rating.
2. Investigate the methodology behind the ABC News/Washington Post poll.
3. What are the major differences between the sampling techniques of the two polling companies?
4. After learning about the two companies’ ways of collecting data, which one’s rating do you trust more?
5. Investigate the methodology used by polling aggregation websites such as fivethirtyeight.com.

Exercise

The article Poor disadvantaged by ‘broken’ system on The Australian website shows the disparity in government funding for schools of different levels of socio-economic advantage/disadvantage. The article bases this kind of socio-economic advantage/disadvantage on the Index of Community Socio-Educational Advantage (ICSEA) that is published on the My School website.

1. Find out your school’s ICSEA score from the My School website.
2. Investigate how ICSEA values are calculated.
3. Can you think of any drawbacks to this methodology?

Exercise

The article Why Jack Welch Has A Point About Unemployment Numbers on the Forbes website discusses the criticisms leveled at the way the unemployment rate is calculated in the US.

1. Investigate how the official unemployment rate in Australia is calculated by the Australian Bureau of Statistics (ABS).
2. What are the differences in this methodology and the methodology used by the US Bureau of Labor Statistics?
3. Are any of the criticisms discussed in the article applicable to the ABS’s methodology?
4. Are there any changes you think should be made to improve the way the unemployment rate is calculated?