To test their theory, they looked at a sample of 3,066 men and women who had been interviewed and tested for depression once in either 1987 or 1988 and then again five years later. In the interviews, they were asked about the quality of their marriage (if they were married).
On average, controlling for differences in depression, subjects who had gotten married over the five-year span between the two interviews reported improved psychological well-being in the second interview--scoring an average of 3.42 points lower on the 84-point depression scale--than their counterparts who did not marry.
When they teased apart how marriage affected those who had been depressed at the start of the study to those who had been happy, however, they came across something unexpected. The depressed who married scored an average of 7.56 points lower on the depression scale than the depressed who did not marry, while those who were happy and got married scored only 1.87 points lower on the scale.
