As much as I would like this survey to be all nice and perfect, there are some limitations that are unavoidable.

---------------------------------------------------------------

The survey at times is not representative of the whole population. For example, 200 people as the sample size may not provide as accurate results as I would hope, but it is enough to give a fairly accurate trend, and the trends are what I aim to find out.

Another example would be the weakness of the 2nd hypothesis, on the speed of the Internet connection affecting the music downloads. Only 7 out of 200 respondents had dial-up connections, the rest have broadband connections. I have reason to take this optimistically though. Dial-up connections are considered as good as obsolete, thus observing trends which are affected by dial-up connections would not be useful for the future which promises super high speed broadband. It is however useful for measuring how the difference in speed affects the downloading.So it is definitely use a comparison between dial-up and broadband, then compare it again with broadband against super high speed broadband and see whether the trend of faster Internet connections downloading more music still exists.

Disaster also struck in the week prior to the submission of the web-report. My online server, Ed-Host, crashed, and along disappeared all my data. Luckily, I saved most of the project into a thumb-drive already. I had to re-tabulate all the lost data into SPSS, re-upload it, and re-analyse it.

Frustrating!

The term test that clashed with the workshop Mr. Lee Teck Kong organised was also a headache, as the Year 3s and Year 4s were forced to miss all but 10 minutes of the workshop.