Data
When building legal data for structured analysis, use employee ID as your primary method for identifying case members. It’s human nature to default to names, but names cause problems, especially across large or messy datasets.
Here are a few easily avoidable issues I’ve seen repeatedly over the years, many of which have taken dozens (if not hundreds) of hours to clean up:
Name spellings that don’t match
Maiden name inconsistencies
Family members with the same last name accidentally combined
Formatting mismatches (e.g. “Smith, Jason” vs. “Jason Smith”)
Mix ups like Angela Smith vs. Angel A. Smith
Even when you do have IDs, Excel will often strip leading zeros (0000555 becomes 555). You can fix this with:
=TEXT(A1, "00000000")
This keeps ID formats consistent and avoids issues later.
In cases where you have to match on names (say, between your client’s list and production data), fuzzy logic tools can help. This can even be done in Excel, with some setup.
Names can always be added back in later. But if you anchor your analysis to a clean, consistent ID from the start, you’re building on solid ground.
Let your data be your advantage.
info@fairvaluemetrics.com
Basic live models to get your wheels turning on how we can help your case: https://lnkd.in/erRkUU8u
Read about our firm: https://lnkd.in/e5xE_FWJ