Changes in smoking prevalence over time and between countries
It should now be clear that, in order to properly assess the impact of the display ban on
smoking prevalence, I need to take account of all factors that may explain changes in
smoking prevalence over time and between countries, and not just the implementation of
the display ban. This requires using multiple regression techniques. These techniques
make it possible to estimate the relationship between two variables (e.g., smoking
prevalence and a display ban), when the variable under investigation (smoking
prevalence in our case) is influenced by many other factors (e.g., cigarette prices).
Using a multiple regression analysis, I can therefore quantify the impact on smoking
prevalence of a display ban. More precisely, I can estimate by how much smoking
prevalence would change if a display ban is introduced while all other factors influencing
the smoking rate are kept constant. And, furthermore, I can test whether this relationship
is statistically significant.
smoking prevalence, I need to take account of all factors that may explain changes in
smoking prevalence over time and between countries, and not just the implementation of
the display ban. This requires using multiple regression techniques. These techniques
make it possible to estimate the relationship between two variables (e.g., smoking
prevalence and a display ban), when the variable under investigation (smoking
prevalence in our case) is influenced by many other factors (e.g., cigarette prices).
Using a multiple regression analysis, I can therefore quantify the impact on smoking
prevalence of a display ban. More precisely, I can estimate by how much smoking
prevalence would change if a display ban is introduced while all other factors influencing
the smoking rate are kept constant. And, furthermore, I can test whether this relationship
is statistically significant.
Read the rest of this entry ... (390 words left)



