Programming Assignment #2

Select one of the following datasets. Conduct a thorough analysis of these data using the methods outlined in Knafl, chapter 5.

Note: Both datasets were changed on Thursday, 8/13 @ 5:30pm. If you downloaded your data prior to that date and time please re-download and redo your analysis.

1. Internet Performance:

   The ISP dataset contains data on the latency level for a round trip transmission across the backbone of three Internet Service Providers (ISP). Latency level is the time required (in milliseconds) for a transmission to traverse the backbone of an ISP.

   Latency level was measured by special probes installed in the Internet backbone of the selected ISP's. The probes were programmed to send out "pings" (i.e. test transmissions) every 4 hours over a 24 hour period. The time required to get a response was automatically recorded for each ping.

   Do these data provide sufficient evidence to indicate a difference in latency for the three ISP's represented.

2. Human-Computer Interaction:

   The HCI dataset contains data on the time required (in seconds) to complete a specific task on three commercially available palmtop computers.

   Thirty three randomly selected college students were divided into three groups and each group was given basic training in the use of one of three commercially available palmtop computers. The students were then each assigned the same simple task set and the time required for completion was recorded.

   Do these data provide sufficient evidence to indicate a difference in task completion time for the palmtops.

You should prepare a written report (typed not handwritten). Use the format presented on 1.3.25 as a model. Ensure that you do the following:

• State the ANOVA model and assumptions.
• State all relevant hypotheses and p-values
• Present a post-hoc analysis if appropriate.
• Use meaningful variable names and dataset names.
• Use label statements where necessary to give your output sensible headings.
• Include appropriate title statements.
• Include "proc print" in your program to print input values.
• Use "proc rank" and "proc plot" to produce your normal plot.