This is the technique I tend to use, together with a 'tentative' flag to condition the presentation on the diagram to highlight the status (different box outline colour etc). I tend to work off the diagram, so this works for me. John > I can see a possible benefit in being able to record the number of children > / spouse an individual has but if I suspect a couple had children (ie > deceased children from the 1911 census), I think I would add a child just > named (for example) "Child 1", "Child 2" etc. and put a note explaining the > assumption rather than just entering a number. Brian