Home Page Download Times Study

Abstract: Based on an investigative survey of homepage download times,

the OurCampus! Quality Response Team has been able to make some useful
deductions about the pattern of those times and is suggesting some possible
action items for the administrators of the OurCampus! Web site.
History: Last year, the OurCampus! Quality Response Team was constituted
and charged with investigating the download times of the main home page
of OurCampus! In our first meeting, Cindy Geozak and Wilson Fremont were
selected as co-Team Leaders. The QRT then held a series of five morning
meetings to discuss its methodology. We agreed that we would survey
download times in-house and not at remote locations. Katie and Matt met
and interviewed various officers of our company and the entire QRT team
decided to hold an one-day conference. When attendance was less than
expected, we scheduled an open forum in which interested employees asked
questions. Based on that feedback, the team agreed to ask our IT
department to develop a custom application that would automatically record
download times.
Methodology: When we were ready to begin our data collection, we loaded
the latest version of Microsoft Internet Explorer and the custom download
time-tracking application onto 5 identical personal computers in the Technical
Administration Department. Using standard 10/100 Mbps Ethernet
connections, we attached the PCs to the primary Web server for the main
OurCampus! portal server. We opened Internet Explorer, triggered the
custom application, and entered the URL of the TriCities OurCampus! home
page into the browser. Our custom application automatically detected when
the home page was finished downloaded and it recorded in a log file the
elapsed
time
along
with
a
date
and
time
stamp.
We then analyzed the collected data and present for the first time publicly
that data in the next section of this report.
Conclusions: We were surprised by the many different download times that
were recorded.

We were pleased to see that the mean time was only 12.8596 seconds. We were
pleased to discover that 50 percent of download times were less than the mean
and that 50 percent were greater than the mean. This fact allowed us to use the
normal distribution to further analyze our data. Using the principles of this
distribution, we can conclude the following:
A 15-second download is less likely than a 14 or 13-second download. (Recall
this was a question raised in our open conference when we began our
investigation.)
If we can strive to eliminate times greater than 22.7 seconds, then more times
will fall within 3 standard deviations.
One time out of every 10 times, an individual user will experience a download
time that is greater than 17.06 seconds.
Since over 99 percent of download times fall within plus or minus 3 standard
deviations, our home page download process meets the Six Sigma benchmark for
industrial quality. (Recall that senior management held a meeting last month on
the importance of the Six Sigma methodology.)
We would like to suggest that the funding for the QRT be continued into next
fiscal year so that we can undertake a more complete sampling of home page
(and other Web page) download times.

Respectfully submitted,
The OurCampus! Quality Response Team

Assignment Questions
1. Can the collected data be approximated by normal distribution?
2. Review and evaluate the conclusions made by OurCampus! QRT. Which conclusion are correct? Which
ones are incorrect?
3. If OurCampus! could improve the mean time by five seconds, how would the probabilities change?