...
Research the easy lines and use them to learn your way around the research tools and resources available to you. What they can tell you and what they can't. And start to get the feel of the pattern of how people lived and worked in different parts of the country at different times in the past.
Then take this knowledge and experience, and apply it when researching your more difficult lines.
Good advice, jbml, thanks! Based on things I've recently learned through Rootschat, I'm again checking the completeness and reliability of the data I already have. Some new tools helped to highlight possible inconsistencies. I'm (again) re-checking the records and 'evidence' I have that supports the identity of key people and relationships in my tree. I'm also checking for any records/evidence that may contradict these. I've gone through a similar exercise in the past but I'm being more systematic and meticulous about it this time around. It's a good learning exercise!
All the relevant records in England and Wales are readily available back to 1837 (BMD) and 1841-1911 (census). So I'm mostly filling in any gaps I discover in my record-keeping or in the links between records and the relevant people/relationships.
Although I've added 'research notes' sporadically in the past, I'm adding these more systematically now to summarise how I arrive at conclusions based on the available evidence. This is useful for me in the future and also for others using my tree data. I'm also adding 'To-lists' to people to collect additional records, resolve inconsistencies, etc. So I am in way developing my own 'workflow' based a checklist of facts to find and things to check for both individuals and relationships.
I've also made the reliability of data visible in the tree by using icons (OK, ?, !). This helps me see what needs to be done first at each level in the tree before going on to research higher levels.
I was interested in how other people approach this kind of work.
Mike