Politics Survey!!! POL 1013.010 Q: Introduction to American Government Fall 2014 Dr. Gervais Due Date: November 7th, 2014 Survey Report Assignment Public opinion polls play an important role in contemporary American politics. For this assignment, you will be learning about the margin of error in public opinion polls and some basic knowledge in how to conduct professional surveys. You will be conducting your own survey and authoring a report on the survey results. The assignment consists of five parts, and a grading rubric is included at the end of the document for your reference: Part I: Collect responses to a survey question Part II: Organize the collected data in a spreadsheet Part III: Graph your survey results Part IV: Calculate the margin of error and 95% confidence interval Part V: Summarize the results of your survey, margin of error calculations, and confidence interval calculations, interpret the statistical significance of your results based on the margin of error and confidence intervals calculations, and critique the survey methodology Once you have completed the assignment, submit it on BlackBoard Learn in the appropriate place under the Assignments link (it will be submitted via SafeAssign) as a single document. Part V should come first, followed by Part III and Part II. This assignment is due by 11:59 PM on November 7th. Assignments turned after this will be docked two full letter grades. Assignments turned in after November 8th will not be accepted under any circumstances. Introduction to the Margin of Error in Public Opinion Polls Here is a brief overview of the margin of error. Lets suppose a poll was conducted in which 1,000 UTSA students were asked if they approved of the job President Barack Obama was doing. About 51 percent of the student respondents said they approved of Obamas job performance. Could we conclude that a majority of the UTSA student approved of Obama? No, not yet. First we need to consider the margin of error, which we can understand to be natural errors in statistical measurement that involves sampling from a larger population. The margin of error for this sample, based on 95% confidence interval, is 3.05. What does this means? Essentially, it means that if we were to conduct this survey 100 times, 95 of those times the percentage approving of Obama would fall within the range of 47.95 percent and 54.05 percent. We get 47.95 by subtracting 3.05 from 51 percent, and 54.05 by adding 3.05 to 51. Because the low end of our confidence interval, 47.95, is below 50 percent, we cannot conclude that a majority of the student body approves of Obamas job performance. That is, the difference between the percentage that approves of Obama and does not approve of Obama is not statistically significant. So, how did we get the margin of error? We can take our sample size (1,000) and the size of the population we are sampling from (the UTSA student body population is currently about 29,000), and enter into an online calculator, such as this one here: http://americanresearchgroup.com/moe.html. Try it and see if you get 3.05. Now, if we were to decrease the sample size from 1,000 to 500, what happens? The margin of error increases to 4.34. Our poll becomes less accurate. This tells us that if we want a poll to be accurate, we need a big sample. The general rule of thumb is 1,000 respondents. Additionally, we want the sample we select to be representative of the population we are trying to learn about. When it is not, our survey may not be accuratewe call this selection bias. For example, if our survey consisted of only female, there may be a gender bias. We know not all UTSA students are female, and males may approve of Obama more or less than females. Other concerns include the wording of the survey questiondoes it push individuals to answer in a certain way? Does it confuse respondents? Bad wording can bias a poll. So can how we ask the questionif we do it face-to-face, do respondents feel pressured to give an answer they believe will please the interviewer (you!)? We call this social desirability bias. Assignment, Part I. The poll discussed above was imaginary. You are going to collect actual responses from the UTSA student body, which will involve using some basic survey methodology. You will be required to ask at least 16 UTSA students about whether they approve of President Obamas job performance, using the following prompt: I am conducting a survey for my Introduction to American Government class. I would like to ask you about your opinion of President Obamas job performance. Your answer will be kept confidential. First, can you tell me if you have already been asked this question by another student and provided him or her with an answer? If a student has already answered the questions with another student, do not record his or her answer. Instead, reply: This concludes the survey. Thank you for your time, and move on to another student. If the student has not been asked this question previously, then ask him or her the following: Do you currently approve of the way Barack Obama is handling his job as president? Please answer yes or no. After he or she answers the question, ask them for his or her name. End the survey with the following: This concludes the survey. Thank you for your time. Make sure you record both the names and answers or your respondents! Note: graders will be able to determine if student names and answers are invented; fabricating responses will be considered academic dishonesty. Assignment, Part II. After collecting answers in the field, it is important to organize the data. You will put the names of your respondents and their answers in a spreadsheet, as well as the date you surveyed them on. Their names should go in the first column. In the second column, give them a score of 0 if they did not approve of Obamas handling of his job, and a score of 1 if they did approve of Obamas job as president. Tabulate how many individuals approve of Obamas job performance, and provide the percentage of those who approve. A blank spreadsheet that you can use is provided in a separate document on Blackboard. Your table should be labeled and included after your paper (Assignment Part V). See below for an example: Table 1: Survey Respondents Names and Obama Job Approval Respondents Name Date Surveyed Obama Job Approval (0=No, 1=Yes) Johnny Carson October 1, 2014 0 Jack Parr October 1, 2014 1 Jimmy Kimmel October 2, 2014 1 Jay Leno October 3, 2014 0 David Letterman October 6, 3014 1 Jimmy Fallon October 8, 2014 0 Conan OBrien October 9, 2014 1 Approval Percentage 57.1% Margin of Error 37.0 95% Confidence Interval 20.1%-94.1% Assignment, Part III. Now we want to visualize the data and put the data into a bar chart or pie chart. You can generate your chart using Microsoft Excel. The figure should have a title and data labels. Below is an example of a pie chart: Note: If you are not familiar with spreadsheet programs, here are several short tutorials on preparation of charts or graphs in Excel: http://spreadsheets.about.com/od/excelcharts/. If you do not have access to a charting or graphing program, there are several online, free chart and graph programs available if you do a search for How to make charts or graphs. Assignment, Part IV. Next, we want to figure out the margin of error. You can do this by using the online margin of error calculator located here: http://americanresearchgroup.com/moe.html. The population we are sampling from is the UTSA student body, which currently numbers around 29,000. Enter this in Population Size. The sample size you enter will be the number of respondents you collected answers from (remember, the minimum is 16). Once you have entered this information, click on Calculate Error. This is your margin of error, and should be entered below Approval Percentage in your spreadsheet. Next, using the approval percentage and the margin of error, calculate the 95% confidence interval and included in in the Table 1 spreadsheet. See Table 1 above in the Part II instructions for an example. 57% Approve 43% Do not Approve Approval Ratings of President Obama's Job Performance Assignment, Part V. Finally, report the results of your survey. Do a majority approve of Obamas job performance? You should report the number of respondents you polled, the percentage of respondents who approved of Barack Obamas job as president, and the dates you conducted the survey (e.g., from November 1st to November 5th). Next, summarize the results of your margin of error calculation and 95% confidence interval. Within what range can we expect the percentage approving of Obama to fall 95 out of 100 times? After this, interpret the statistical significance of your results based on the margin of error and the 95% percent confidence interval you calculated. Is the difference between the percentage approving and not approving of Obamas job performance statistically significant? If it is significant, what can you conclude about UTSA students perceptions of Obamas job as president? If the difference is not significant, what could you have done differently to have generated a statistically significant result? (Hint: think about the sample size). Finally, what are the weaknesses of the poll you conducted? Consider the question wording, the sample that was surveyed, and the manner in which the survey was conducted (a description of these errors can be found in the Introduction to Public Opinion Polling section above; also reference pages 228-241 in your textbook). If you were to conduct this poll again, what changes would you make in order to produce more accurate results? The Writing Center is a great resource that all students are encouraged to utilize (https://utsa.edu/twc/). Reports are required to meet the following format requirements: Font size should be 12 point font Font style should be Times New Roman The report must be double spaced The report should have 1 margins all around You must use complete sentences and present coherent thoughts. Grading Rubric Collect responses to a survey question (12 Points) 12 Points: Collected responses from at least 16 UTSA students 8 Points: Collected responses from 10 to 15 UTSA students 4 Points: Collected responses from 3 to 10 UTSA students 0 Points: Collected responses from less than 3 UTSA students Organize the collected data in a spreadsheet (12 Points) 12 Points: All of the components of Part II are complete. (1) Names of all respondents are recorded, (2) dates respondents were surveyed are recorded, (3) survey responses are entered as 0 and 1, and (4) the approval percentage is calculated 8 Points: One of the components of Part II is incomplete or missing 4 Points: Two of the components of Part II are incomplete or missing 0 Points: Three or more components of Part II are incomplete or missing Graph survey results (14 Points) 14 Points: Three components completed successfully. (1) Approval percentage is accurately and properly presented in a bar graph or pie chart, (2) the chart/graph has a title, and (3) data labels are included 9 Points: Two components completed successfully: Bar/pie chart is included, but approval percentage is not accurately/properly displayed, chart/graph title is missing, or data labels are missing 4 Points: Bar/pie chart is included, but two or more of the elements listed above are not included 0 Points: Bar/pie chart not included Calculate the margin of error and 95% confidence interval (12 Points) 12 Points: Margin of error and 95% confidence interval are included in Table 1 and are accurately calculated 5 Points: Margin of error and 95% confidence interval are included in Table 1, but were not accurately calculated 0 Points: Margin of error and/or 95% confidence interval are not included in Table 1 Summarize the results of the survey, margin of error calculations, and confidence interval calculations (16 Points) 16 Points: Successfully reports the number of respondents surveyed, the percentage of respondents who approve of Obamas job performance, the dates of the survey the margin of error and 95% confidence interval. 10 Points: Reports the margin of error and 95% confidence interval, but does not report all of the survey result components (sample size, support percentage, polling date) 4 Points: Reports survey results but does not report the margin of error and 95% confidence interval 0 Points: Does not report survey results or the margin of error and 95% confidence interval Interpret the statistical significance of the results (14 Points) 14 Points: Correctly interprets the statistical significance of the results and makes correct conclusion 4 Points: Incorrectly interprets the results 0 Points: Interpretation of the results is missing Critique and improve the survey methodology (10 Points) 10 Points: Accurately identifies a weakness in survey methodology and offers valid recommendations for improving survey 5 Points: Either accurately identifies a weakness in survey methodology or offers valid recommendations for improving survey, but not both 0 Points: Does not accurately identify a weakness in survey methodology or offer a valid recommendation for improving survey Reports are written with proper grammar, style, organization, and formatting (10 Points) 10 Points: Meets all requirements 6 Points: 1-2 errors 2 Points: 3-4 errors 0 Points: 5 or more errors