Lecture 3. Lecture Outline. 1. Turn in Homework 2. Sampling Quiz 3. Essay Writing Lecture. Assignments

Transcription

1 Lecture 3 Lecture Outline 1. Turn in Homework 2. Sampling Quiz 3. Essay Writing Lecture Assignments 1. Thesis Activity 2. Review Essay 3. Improving the Paper Preparation for Next Class 1. Read Chapter 1 in Textbook 2. You must have an M Drive by next class! Quiz for Next Class- Chapter 1 (Levels of Measurement)

2 Essay Structure Example Used in Class Discussion Assignment: Compare and contrast administration of a survey about interracial relations using an online (AOL) and interview (GSS) method. How will the results of the studies be different? Points to Address: Who participates in each study? Are the participants likely to be representative of the United States population? Are participants of both studies equally likely to give honest answers? Possible Essay: Collecting accurate statistics that represent the entire U.S. population is a challenging task. In this example, both the General Social Survey (GSS) and America Online (AOL) are attempting to study current American attitudes toward interracial relationships. Although the GSS study is not perfect, the results will be superior to the AOL results because the survey method will result in more truthful answers and data from a more representative sample of the U.S. population. The participants of the GSS are more likely to give honest answers for three reasons: 1) the method of asking the questions allows researchers to validate the answers, 2) the survey is not presented in an influential context and 3) the survey is anonymous. The GSS format instructs interviewers to ask questions one at a time. Therefore, the results of the general questions are validated or refuted based on the results of the networking questions. The AOL survey is presented in whole, allowing participants to modify answers in order to create a false congruency. Secondly, the AOL survey is presented with an article about racial tolerance which is likely to influence survey results by suggesting a socially acceptable answer. Lastly, in order to access AOL, survey respondents must provide a log-in and password which makes survey results not anonymous. In contract, interviewers conducting the GSS inform participants that results are anonymous. Further, the GSS survey participants are more representative of the U.S. population. The GSS uses a complex system for identifying and then interviewing a large random sample of the population. The repeat visits made by interviewers to each household increase the probability of reaching all segments of the population. However, the GSS is unlikely to survey members of the population without a home such as the homeless or incarcerated. In contrast, AOL exclusively presents the survey to members who choose whether or not to participate. Elderly and poor populations are underrepresented in the AOL survey because they are less likely to have internet access or a paid membership to AOL. In summary, though no study is perfect, the GSS results in the most accurate and truthful measure of attitudes towards interracial relationships. Through the use of random sampling, anonymous surveying, and good interview design, the GSS results in data that is superior to the data generated by the AOL survey.

3 Thesis Activity Who should be responsible for paying for college? Write a thesis and topic sentences for an essay that answers this question. Numbers We Should Know Paper Reviews Write a one-page review of the paper you have received. Please bring two copies of each review (one for the author, the other for me). In this review highlight what you most value about the paper and list suggestions for improving the paper. Make minor corrections on the paper itself.

4 Example Essay for Improving The Paper Activity Assignment question: Should sampling be allowed in the Census? The United States Census is perhaps the most extensive data collection effort in the United States. The results of the Census are important in determining the amount of Federal funding and the number of United States Representatives an area receives. The Census is used to count and collect data about every individual in the United States; however, it is not always successful. Sampling is viewed as a potential solution. The Census requires enormous resources to conduct, yet, locating some individuals is more difficult than other locating others; sampling can help compensate for missed individuals. Individuals such as the homeless are difficult to identify and are, as a result, undercounted. We know they are undercounted because, in some areas, the Census bureau conducts an intensive door-to-door search to confirm the official count. The difference between the more intensive and traditional count provide an estimate of the number and type of people missed. Sampling can be use to establish the likely characteristics of missed people. Sampling may decrease the integrity of the Census. Currently, each person in the Census is uniquely identified. If sampling is introduced, the final count is dependent on the procedures adopted by a statistician in the Census Bureau who may not be held publically accountable for their actions or may be politically motivated. However, arguably, the resources allocated to conduct the Census thoroughly and the decision not to sample is also political. In summary, sampling in the Census is controversial. It may be a solution. However, it also has a potential for misuse.

5 Improving the Paper Activity Rewrite the thesis to make it more thoroughly recount the main idea and argument of the paper. Rewrite one of the sentences in paragraph 2 to support the main idea of the paragraph numerically. Rewrite the topic sentence for paragraph 3 to better reflect the content of the entire paragraph. Rewrite and improve the content in the conclusion.