Final Paper

1773 Words4 Pages

Coverage/Noncoverage and likely Coverage Error
The target population is people who are 18 and older in the Flagstaff area who are English speaking. In order to sample those specific people the sampling frame used was a method called Random Digit Dialing (RDD). Specifically the RDD method was used with landline telephones and all of the numbers had a 928 area code. When the professor researched in the phone directories he found that there were specific prefixes that pertained to the Flagstaff area. He also found that each prefix had suffixes that ran from 0-9000.
The sampling frame did not represent the target population very well because there are a lot of English speaking Flagstaff residents without a landline. We did have some coverage problems because we did not have any cell phone prefixes, and if we did that would have covered the target population considerably better. The fact that there was coverage error is a large issue because I speculate that people who have landlines and people who don’t are significantly different. These people are different when it comes to things like education, age and income.
Generally people that have landline telephones have average to higher education completed in their life. While people that are currently still in college do not have a landline because they are always moving. The age gap between people with landlines and cell phones is another considerable factor, because typically people with landlines are much older because cell phones are still a new technology to them. Income is another large factor, because lower income people usually do not have a landline telephone. To add on to that, minorities are what make up a large part of the lower class. Their views on political subjects f...

... middle of paper ...

...e said road construction, those can be in separate categories, but they can both be in a general category of traffic also. Some people could have been placed in the wrong race because they answered the race question saying they were American. Careless errors could have also occurred, such as pushing a wrong number on the key board while imputing data.
Inferential Error
Overall, there was definite Internal Validity (did X cause Y) in the data that was collected in the survey. One could argue that the order of the questionnaire gave the whole survey strong Internal Validity. There were two experiments built within the questionnaire in the same sex couple question and in the marijuana question. These two questions specifically both had strong Internal Validity. Both of the questions were able to show how one change in the question can change someone’s answer.

More about Final Paper

Open Document