That's not enough. One of the often espoused counterarguments is "do happy people marry more, or does marriage make people happier?". You'd have to continuously check happy people and keep a control group from marrying. Checking just the widowers doesn't account for the severity of the loss.