It is usually a result that has a statistical
significance that is rather low so that you might have a 20 percent difference (UNCLEAR (4.17) or a
relative risk
of 1.2 or something like that; and so it is easy when you have risks that are so small or the differences are so small to have one
study show one kind
of result and another
study show a different kind
of result.
Also, because our
study treated age as time - invariant and contrasted middle - aged and older adults, our findings
of a
relative lack
of age variations may indicate that little cohort difference exists in the
significance of perceived spousal support in relation to health.