Just because variable A (e.g. owning a computer) and variable B (e.g. buying less music) occur together does NOT mean that either one necessarily caused the other. This study proves nothing useful, regardless of its validity. Even if you could show that people who actually download music tend to pay for it less, you still have not proved that downloading harms music sales. This is because correlational studies (such as the one FTA) do not rule out additional variables (e.g. socioeconomic status).
Slashdot Mirror
Just because variable A (e.g. owning a computer) and variable B (e.g. buying less music) occur together does NOT mean that either one necessarily caused the other. This study proves nothing useful, regardless of its validity. Even if you could show that people who actually download music tend to pay for it less, you still have not proved that downloading harms music sales. This is because correlational studies (such as the one FTA) do not rule out additional variables (e.g. socioeconomic status).