Measuring apects of real world behaviour often reveals correlations between different variables.  Correlations indicate the strength and direction of the relationship.  For instance, there may be a correlation between the length of the page and time spent on a site. However, it is important to remember when looking at correlations that they do not prove one variable is responsible for another.  One may cause the other, the other way round or they may both be affected by a third factor.  In the case of page length and time on site, users may be spending longer on the site as there is more to read or because it is harder to find what they are looking for.  It may also be that sites with longer pages have something else in common that mean users spend more time, such as more related content to which users can be directed.