Listening audioscript

 
Number 2, Page 275
[ mp3 115] 
Now listen to a lecture on this topic in a political science class. 
(Professor)
There are many ways that sample bias can affect the results of an 
election poll. The use of telephone surveys has always presented 
pollsters with several kinds of bias that have changed along with 
changes in technology. When telephones were less common among 
rural voters and poor voters, surveys conducted by telephone tended to 
exclude these people. This created an exclusion bias because people 
who did not have phones were underrepresented in, or even excluded 
from the polls.
 
 
As technology has made it possible for more people to replace their land 
lines at home with mobile cell phones, another bias has been introduced. 
Surveys conducted by land line over-represent older and more rural 
people. That’s because they are more likely to rely on a land line than to 
have gotten rid of it in favor of a cell phone as younger, more urban 
individuals have done.
Another type of bias that is always present in telephone surveys is the 
self-selection bias. That is, since you are not obligated to answer survey 
questions over the phone, those who don’t care about or know little about 
the issues are more likely not to agree to answer the survey questions. 
On the other hand, those who are willing to answer survey questions are 
people who may already have strong opinions and feel knowledgeable 
about the issues or political candidates. In all of these cases, the 
samples do not accurately represent the total population, and so the 
results might reflect more heavily the views of urban people, the elderly, 
or people who are actively involved in politics. To accurately represent 
the entire population, statistical methods must be applied to correct for 
these biases. 

---

LPREP IBT 3 E AudioScript 
79