AP Statistics / Mr. Hansen 1/14/2006

Name: _________________________

Summary Results from Project 1:
Exploratory Data Analysis

Here, recorded for posterity, are the results of the group exploratory data research from the first quarter.

Group	Members	Report Title	Synopsis	Comments
1	Glenn K., Ben J., Andrew H.	A Study of GPA, TV Watching, and Political Affiliation	Weak negative linear relationship (r = –0.305) found between TV hours and GPA. Republicans (20 of 52 subjects sampled) had mean GPA of 87.3, while Democrats (29 of 52 subjects) had mean GPA of 85.5. Remaining 3 students had mean GPA of 88.7 but were too few to permit any conclusions regarding them.	Sample was self-selected from mailbox postings of surveys. Researchers cited response bias (but not voluntary response bias) as a likely source of error.
2	Paul J., Andrew B., Jeffrey P.	Perception of Time vs. Type of Music	No real trends seen for rock, rap, or classical. Techno seemed to exhibit a bimodal time-perception pattern, suggesting an area for future research. Subjects listening to techno tended either to underestimate or overestimate elapsed time by about 20%. For other genres, subjects showed a more typical unimodal response pattern with fairly accurate central tendency.	Methodology involved 60 subjects and 15 data points for each genre of music. Subjects wore a blazer over their heads and indicated when they thought 90 seconds had elapsed. Researchers used a single musical composition to represent each genre. Although this approach is common, one could say that a single piece of music cannot portray its entire genre, any more than a single person (e.g., Bill Clinton or George W. Bush) can represent all Americans.
3	Henry P., Chris R., Jansen S.	[QB Rating vs. Draft Number]	No correlation found between career QB rating (a passing-only performance statistic adopted by NFL in 1973) and player’s position in draft.	A pattern not identified by researchers was that QBs near middle of draft order (approx. 35-80) show a much lower s.d. of QB rating than those whose draft position was either very high or very low.
4	Will K., Christian C., Michael L.	Does Satisfaction with St. Albans Vary Along [sic] [w]ith Time at St. Albans?	No correlation found between satisfaction with STA (as measured on a 5-point scale) and length of time enrolled.	Correlation coefficient r was not calculated correctly, but conclusion remains unchanged. Six pages of color graphics should have been replaced by one line graph and two tables.
5	Greg C., Clay B., Alex W.	Sports vs. Smarts: A [B]attle of Varsity Sports and AP Tests in the Senior Classes of [STA and NCS]	No correlation found between number of varsity letters and number of AP exams, either at STA or at NCS. Sample mean for varsity letters was 4.1 at STA (n = 35) vs. 2.7 at NCS (n = 28). Both samples showed a mean of about 4.3 AP exams.	AP numbers were total of current-year classes and prior-year exams taken. Random samples from each school’s senior class were used. Response rates of 78% (STA) and 70% (NCS) were acceptable.
6	John W., Kenny K., Michael M.	An Analysis of the Effect of Morning Commute Time on the Amount of Caffeine Consumed by St. Albans Students	No correlation found between morning commute time and number of weekday mornings per week that student consumes caffeine, where “morning” was defined as “the time you wake to the end of first period.”	Researchers expressed confidence that a larger study would uncover a correlation, especially if lurking variables (e.g., no-caffeine policy for cross country runners) could be controlled.