Mass Communication Research
Final Group Project —
Spring 2002
INDEX SYLLABUS SCHEDULE e-MEDIA COMM-STOP

The final project gives your group the opportunity to show its mastery of the skills taught in the C300 course. To complete the project, the group should get together and decide what variables in the Spring 2002 Survey Codebook [cbs02.html] provide interesting angles for analysis. This analysis will provide the basis for your group's technical report and news release, as well as individual articles. Bonus points and possible publication await the best projects.

NOTE: You have already earned up to 25 individual points for correctly gathering the data, and up to 25 individual points for correctly entering the data. The remaining 150 points of the group project will be split between writing a technical report (100 points) and a news release (50 points). To make it easy for you to have the correct format, click here for the required template.

The group news release will focus on one of three primary topics. In other words, your group MUST examine the primary issue assigned, but you should also examine other issues which might be newsworthy. The group technical report will provide details of the issues under study, while the news release will highlight the newsworthy findings of the group's analysis.
Group Topics
Class Availability (Qs 6-11) M2M6S1S4S6S10
Budget Problems (Qs 12-14) M1M5M8S3S8S11
Student Voting (Qs 15-23) M4M7M9S5S7S9

An open lab will be available on Friday, April 19 from 8 a.m. - noon for those people needing extra time on SPSS. NOTE: Don't wait until Friday to begin analyzing the data. You won't have time to finish. SPSS is available on most of the computers at UT labs and the library. (NOTE: Scroll down the page to where it says "SPSS 10.0.7.")

In order to submit your projects to The Daily Beacon in time for consideration for publication prior to the end of the semester, your group will submit a copy of your final written report to the digital dropbox as a MICROSOFT WORD attachment by midnight on Monday, April 22 (Tuesday, April 23 for the NIGHT CLASS). Late assignments will lose a letter grade per day. You will have the opportunity to evaluate the contributions of your group members on the final group probject from Wednesday, April 24 to Monday, April 29 on CourseInfo's Announcement page.

Of course, if you have questions prior to the due date for the final project, you may discuss these with your instructor during regularly scheduled group meetings or office hours. E-mail queries are always welcome as well.

Technical Report

The technical report (worth up to 100 points) should provide details of the data under consideration and the hypotheses used to test that data. You should only report significant results. If you do not find significance, your group MUST TEST ANOTHER HYPOTHESIS. To see planned deductions for the technical report, click here.

As previously noted, to make it easy for you to have the correct format, click here for the required template.

An introductory paragraph should indicate the method, location and dates of the survey, as well as indicate the sample size, population, and selection procedure for the Spring 2002 SPSS data set. Your group is required to run four of the following inferential statistics: The title of each statistic MUST be used as a subheading, with the following information under each subheading. Do NOT deviate from the format listed below. REMEMBER: A hypothesis is a statement of relationship between two variables. Two hypotheses must examine variables pertaining to your assigned topic. You may choose any topic for the other two hypotheses.

FOR EACH VARIABLE you plan to use in your hypothesis, provide:
  1. The variable name and questionnaire item number.
  2. The number of both valid and missing cases.
  3. The level of measurement of the variable
  4. The name, value, and explanation of what the measure of central tendency represents. An example for nominal data (e.g. the mode of 1 for gender indicates that more males were interviewed than females). Points will be deducted for statements like "It's the largest category" or "It lies between X and Y."

    On Questions 1-4, you should provide information on both of the variables you are testing. These answers will be available by running descriptive statistics. Report the correct measure of central tendency for your independent variable and for your dependent variable. Then, report the value of the the correct measure of central tendency for BOTH your independent and dependent variables by interpreting your SPSS output. Finally, explain what the measure indicates IN RELATION TO THE QUESTION. After you have detailed the variables your group plans to use, provide the following:

  5. A statement of the hypothesis
  6. An indication of whether the hypothesis is one-tailed or two-tailed.
  7. An indication of any recoding that you have done to handle missing data, collapse categories, or select cases. You should not recode "99" as the SPSS program already  has been instructed to consider it as missing data. Explain why you did it and the level of measurement of the resulting variables.

    At this point you should run the appropriate inferential test. Once again, if no significance exists, TEST ANOTHER HYPOTHESIS. You should provide the following:

  8. An indication of whether the result is statistically significant or not, and an explanation of how you determined that. Report the exact significance value from the output. If there's more than one significance level reported for the test, explain your choice (remember, for example that the t-test for independent samples reports three significance levels, but only one will be right) If the significance level printed on the output needs correction, do the correction and explain why (remember that sometimes you need to correct two-tailed significance levels for one-tailed hypotheses).
  9. An inserted copy of the SPSS output.
  10. An interpretation of the results of the inferential test. Remember: interpretation of Chi-square should indicate differences in percentages; interpretation of t-tests or ANOVA should indicate differences in means; and interpretation of correlation should indicate positive or negative relationships between the variables by interpreting the correlation coefficient.
NOTE on Correlation: When running correlation, interpretation is difficult when comparing more than two variables. If you report information from correlation in your news release, you should discuss relationships  between the data.

Pay attention to the way that the data are coded. Sometimes postive things are coded with low values, and sometimes with high values. You need to know about these things to interpret correlation, e.g., if you plan to run correlation using Q3 as a variable, you will often encounter a "negative" relationship. You should indicate WHY this is so.

News Release

The news release (worth up to 50 points) highlights findings of the survey and technical report. Though the group's news release will focus on their assigned topic, it may also incorporate other topics with news value.

You'll need to know the basic frequencies for your news release. These have been provided for you here. However, you may need to run other frequencies if you've recoded data, or if you report data from subsamples. If you do such things for your news release, be sure to PROVIDE THE OUTPUT in your technical report.

This portion of the assignment is similar to Group Project 2. However, you now know how to run and report inferential statistical data. Like the professionals whose work you reviewed in the Group Project 1, you don't need to report all the information available in the data. Your group should report newsworthy aspects of the data and ignore others.

The length of the news release MUST be between 750 and 1,500 words long. It must be written in a professional manner (with no mechanical errors of grammar, spelling or style), and adhere to AAPOR standards and AP style. To see planned deductions for the news release, click here.The release MUST have the following:
  1. An appropriate embargo date and contact information.
  2. A quality news lead (35 words or less) based on news values and incorporating your analysis.
  3. A coherent explanation of the findings of the study. Generally, you should NOT speak of significance in your news release. You will ONLY report significant findings.
  4. Appropriate use of numbers (e.g. percentages and means).
  5. A boxed sidebar containing information that adheres to the AAPOR Standards of Disclosure.
  6. A -30- or ### to indicate the end of your release.
You can increase the possibility of receiving the bonus points by presenting the following in your news release:
  1. graphics that elucidate the findings
  2. interviews from appropriate sources that exemplify the findings of your release
  3. evidence of review of appropriate literature from such sources as newspapers, magazines, Web sources, books, etc.
  4. gathering of additional statistical data, e.g. percentages of reported campus crimes
Pertinent Information for AAPOR standards

  1. Sponsor: College of Communication, University of Tennessee
  2. Exact Wording: Exact question wording can be found in the codebook at cbs02.html.
  3. Population: currently enrolled UT undergraduates
  4. Selection procedure: A fixed interval sample was used to assign each interviewer a page from the Spring 2002 timetable. As part of a class project, students in three Mass Communication Research classes conducted interviews with either upperclassmen (juniors and seniors) or underclassmen (freshmen and sophomores), depending on their group assignment. Intercept interviews were used to find students coming or going from classes chosen from the timetable pages. A quota sample was also used to ensure a balance of respondents by gender.
  5. Sample size: 489 students
  6. Precision of Findings: +/- 5 percent at a confidence level of 95 percent
  7. Subsamples: The error rate for smaller samples may be even larger than the 5 percent overall error rate for the entire sample.
  8. Method, location and dates: Face-to-face interviews conducted in various academic classroom buildings on the University of Tennessee-Knoxville campus from April 2-12, 2002.
Technical Report Deductions
  • not following format (-5)
Specifics for Introduction
  1. no introduction (-5)
  2. lack of detail in intro (-2)
  3. incorrect date (-1)
  4. not indicating survey refers to undergraduates only (-1)
Specifics for Descriptive Statistics
  1. no indication of question's level of measurement (-3)
  2. incorrect level of measurement for question (-1)
  3. measurement of central tendency doesn't match question (-2)
  4. no measure of central tendency provided (-2)
  5. incorrect explanation of central tendency value, e.g. just saying it's a five on a 10 pt scale rather than explaining what the five means in relation to the scale being used (-1)
  6. using name other than variable name when reporting dependent variable (-1)
Specifics for Hypotheses
  1. no indication of variables used (-1)
  2. using wording inconsistent with survey question which implies a test of variables that don't exist, e.g. stating hypothesis "Students rate quality of social life higher than quality of education." In actuality, the variables are "social life satisfaction" and "education quality satisfaction" (-1)
  3. incorrectly naming statistical test (-4)
  4. hypothesis doesn't match given variables (-3)
  5. variable names indicated don't match those in questionnaire (-3)
  6. no indication of how recoding accomplished (-2)
  7. incorrect direction [1- or 2-tail] (-1)
Specifics for Output
  1. not inserting SPSS output (-5)
  2. incomplete SPSS output (-2)
Specifics for Interpretation
  1. rounding errors (-1)
  2. no indication of what to do with 2-tail significance level when there’s a 1-tail hypothesis (-1)
  3. not specifying significance value; e.g. reference to significance without specific value "the significance was above 0.05 on the test, but because it was a one-tailed hypothesis we divided it by two and it was significant." (-2)
  4. interpretation mistakes (-3)
Specifics for Chi-Square
  1. incorrect interpretation of table warning due to insufficient cell counts when selecting only certain segments of a population (-2)
  2. reading the wrong rows when intepreting the crosstabs output (-5)
  3. running two chi-squares (-4)
  4. running chi-square on interval data (-3)
  5. reporting differences in means when running chi-square (-3)
Specifics for t-tests
  1. reporting wrong significance output, thus wrong interpretation but correct means (-3)
  2. dividing t-value rather than significance of t-value (-2)
  3. reporting significance of Levene's test rather than significance of t on Independent samples (-2)
  4. likelihood statement w/ t-test (-1)
  5. running two of the same type of t-test (-4)
  6. running t-test with nominal data (-3)
  7. reporting differences in percentages when interpreting means (-3)
Specifics for ANOVA
  1. running ANOVA w/ 2 dependent variables (-1)
  2. not providing significance of F, but giving significance between groups (-1)
Specifics for Correlation
  1. Not indicating how differences in scales make inverse relationships (-3)
  2. Not providing correct interpretation of correlation coefficient (-4)
News Release Deductions

Quality of Writing
  1. grammatical errors (-1 per mistake, up to four mistakes)
  2. AP style mistakes (-1 per mistake, up to four mistakes)
  3. writing in present tense (-2 per mistake, up to four mistakes)
  4. writing in broadcast style for news release (-3)
  5. writing in first person, i.e. "I" "We" "Us" "Me" (-2 per mistake, up to four mistakes)
  6. writing in second person, i.e. "You" (-2 per mistake, up to four mistakes)
  7. FATAL errors, e.g. misspelling of names, inferential statistics, etc. (-4 per mistake, up to four mistakes)
  8. lack of embargo date, contact information (-4)
  9. unsupported assumptions (-3)
  10. Lack of transitions (-3)
  11. awkward attribution (-3)
  12. not meeting minimum length requirements (-5)
Errors in Lead
  1. lead longer than 35 words (-2)
  2. mentioning AAPOR standards in lead (-2)
  3. lead not based on news values (-5)
  4. poorly worded lead (-5)
Quality of Reporting
  1. not using Media Research instead of/in addition to Comm 300 in News release [readers outside of communication will not know what Com300 is] (-1)
  2. not indicating survey pertains to undergraduates only (-1)
  3. reporting information not in technical report (-3)
Errors Reporting Numbers
  1. rounding errors (-1 per mistake, up to four mistakes)
  2. too much detail in reporting percents, e.g. 67.7 percent (-1 per mistake, up to four mistakes)
  3. AP style mistakes concerning numbers (-1 per mistake, up to four mistakes)
  4. using biased terminology, e.g. "surprisingly good," "comforting," etc. (-4)
  5. reporting sample as percentage of the known UT population (-2)
  6. reporting percentages of interval scales without indicating recoding (-3)
  7. not providing appropriate context when reporting subsample results (-2)
Errors Regarding AAPOR Standards
  1. Leaving out one standard (-3)
  2. Leaving out two standards (-5)
  3. Leaving out three or more standards (-8)
Extra Credit:

GROUP POINTS

Your group is competing with five other groups — from your section and the other two sections — for bonus points, i.e. each member in the group will receive the bonus. The best news release for each topic will receive 35 bonus points. The second-best release for each topic will receive 20 bonus points. The top three releases will then be submitted to the The Daily Beacon for their consideration for publication.

For examples of articles that received bonus points last semester, check the articles on these topics: INDIVIDUAL POINTS

For up to 25 points extra credit, individuals may write a stand-alone news story of no fewer than 500 words. The individual articles should focus on topics not covered by any of the groups. It is governed by the same due dates as the rest of the assignment.

The article should have a newsworthy lead of 30 words or less, and conform to Associated Press style and AAPOR standards. It is based on individual analysis, and should also be accompanied by a technical report of the variables used for analysis, including the SPSS output that supports your findings. The technical report should adhere to the form of that turned in by the group.

Duplicate articles will NOT be accepted. There should be enough differences in individual sidebars that there is NO QUESTION concerning authorship — in other words, individual articles should have different angles just as different reporters will cover different elements of a story.

High-quality individual projects may also be submitted to the The Daily Beacon for publication consideration.

If you don't understand something in this Web note, please e-mail Dr. Sitton.

INDEX SYLLABUS SCHEDULE e-MEDIA COMM-STOP
©M. Mark Miller & Ronald W. Sitton 2002
Revised 092811 — http://www.uamont.edu/FacultyWeb/sitton/crz/mrea/finalproj.html