[SmG Dragonrage]

AP Students' Browser Usage: A Trend?
Question:
With today’s world revolving around technology, and more specifically the Internet, we thought it would be interesting to see which Internet browser was the most popular among AP students at Bloomington High School. Statcounter’s reported national usage rates showed IE (Internet Explorer) at the forefront with about 45% of all Internet users, Mozilla Firefox was in second with just over 20%, Google Chrome, in close third with just under 20%, Safari in fourth with about 12%, Opera in fifth with about 1%, and all browsers accounted for the last 2% of all reported Internet users.
In our group we discussed what we thought might be the trend at BHS. The majority of our members preferred the newest addition to the palate of Internet browsers, Google Chrome, and we thought that this trend might be true at BHS since Google Chrome is a heavily advertised virtual gaming platform, and mainly geared for today’s youth and tomorrow’s future.
Approach:
We took a multi-stage sample of each of our AP classes (except Stats). We split Bloomington AP students into five clusters based on five major academic groups (math, English, science, social science, and foreign language). Then students were assigned a number based on their place in the teacher’s roll call sheet or based on their seating position within their classrooms. We then used randomly generated numbers to choose which students would be sampled. Half of the eligible students from each class were chosen from each classrom. We took care to exclude students who had previously had a chance to be surveyed (ex: they happened to be in multiple AP classes). Every student within the chosen classes had close to a 50% chance of being selected to take our survey.
Results:

Calculations:
= pnational
= pBHS
Ho : proportion of Chrome usage for AP BHS students is equal to the national average (- = 0)
Ha: proportion of Chrome usage for AP BHS students is greather than the national average (- > 0)
SE=()(1-)n= (.2)(.8)43= 0.0609994281
z= (-)-0SE= ( 0.53488372093023255813953488372093-.2)0.0609994281 =
5.4899485349475359776288604863318=z
P(z>5.4899485349475359776288604863318)= .00000002015172204
Confidence interval (95% confidence)
pBHS + or - ME where ME= z*()(1-())n= (1.96)(2343)(1-(2343)43= 14.9084435%
So our confidence is 53.488372093023255813953488372093 + or - 14.9084435 which gives us (38.57992859,68.39681559)%. That is to say, we are 95% confident that the true proportion of AP BHS upperclassmen using Chrome is between (38.57992859,68.39681559)%.
Conclusion (no engineering):
With a p-value so small, we rejected the null hypothesis; there is sufficient evidence to suggest that the rate of Chrome usage among AP BHS students is higher than the national average.
Concerns and Follow-Up:
We are concerned with the randomness of the survey. It is possible that our sample was not representative of all AP BHS students. To improve on this, we should have assigned each AP student a number, and randomly chosen numbers from the list created. Better yet, since this is a relatively small population we are looking at, a census would have been a more reasonable thing to do. Furthermore, we tried to clarify the difference between Google the search engine and Google Chrome the web browser, however some individuals might still have mistakenly responded that Chrome was their primary browser.
With an overwhelming difference between BHS AP student usage and overall national usage, many questions are raised. A follow-up study could be done to see if the difference in usage extends to all students, not just those enrolled in AP classes. If a difference persists, a study which breaks down browser usage by age groups could be very revealing.


A Failed Proposition
Our initial proposal for a study involved trying to find whether or not there was an association between students’ SAT scores and their participation in the AVID program. After requests to teachers and administrators, we were unable to obtain SAT scores for legal reasons. We then decided that we could still conduct a similar study using students’ AP scores instead of their SAT scores. However, we encountered similar problems. Additionally, we were told our school’s office computers could not match up AP scores with participation in AVID. Our direct request by email to CollegeBoard went unanswered. Rather than resorting to CAHSEE scores, which the office said they could likely help more with, we switched to a topic of greater interest to our group members, browser usage.