Price differences of iMac configurations

This may be a little difficult to understand unless you know something about regression analysis.

I'm doing a regression analysis example for my statistics class and found something very odd about Apple versus PC makers. I went out and priced various configurations of computers, including Apple (iMac), HP, Dell and Gateway) and regressed the different available components to the different prices to forecast a fair price to pay for a computer based on components wanted.

I figured that if I regressed the various builds of a computer from kind of computer from one company, like Dell XPS, the standard configuration would be 0, meaning the price for each upgrade in components is static. But with Apple I found a huge standard deviation between builds/upgrades. The only thing I can determine from this is that Apple doesn't uniformly price their upgrades, but instead assigns prices to upgrades which cost the consumer more; basically the more you upgrade, the less value and the more premium you are charged. But with PCs you are getting the same relative value for each extra dollar you spend.

What was even more interesting was finding that regressing among maker brands, like Dell Dimension vs. Dell XPS, a small deviation appeared as expected since all things being equal, XPS is supposed to be in some why superior to the Dimension, but it wasn't even as big as the deviation between the various iMacs, which are relatively the same outside of component upgrades.

From my analysis it seems like the harsh claims of Apple charging unreasonable premiums may have some merit to them. If anyone out there has any experience doing regression analysis and want to check out my data, I'd be happy to send it to you.