Chapter 11 — Tenant records and duplicates
Lesson map
Lesson 11 published. Map, example, EX031.
flowchart TD
N0["Learning outcomes"]
N1["11.1 Why duplicates appear"]
N2["11.2 Matching rules"]
N3["11.3 Review and merging"]
N4["11.4 Prevention"]
N0 --> N1 --> N2 --> N3 --> N4
N4 --> EX["Worked example"]
EX --> Q["Exercise EX031"]
Logical drawing
flowchart LR
N0["Learning outcomes"]
N1["11.1 Why duplicates appear"]
N2["11.2 Matching rules"]
N3["11.3 Review and merging"]
N4["11.4 Prevention"]
N0 --> N1 --> N2 --> N3 --> N4 --> Q["EX031"]
[Learning outcomes]
|
v
[11.1 Why duplicates appear]
|
v
[11.2 Matching rules]
|
v
[11.3 Review and merging]
|
v
[11.4 Prevention]
|
v
[Worked example] --> [Exercise EX031]
Learning outcomes
You will (1) find duplicate tenant records; (2) apply matching rules with review; (3) prevent new duplicates.
11.1 Why duplicates appear
Tenant records duplicate when someone moves between Calder Valley homes and a new record is created instead of updating the old one, or when names are spelled differently by different staff. Ravi finds tenants with two records, each holding part of their history. Duplicates cause repeated letters, split rent accounts and incomplete records when tenants request their data. He explains that duplicates matter both for service quality and for meeting legal duties. Ravi counts the likely duplicates in the synthetic data before proposing any rules.
11.2 Matching rules
Ravi writes matching rules: exact match on date of birth plus a similar surname and first initial, or exact match on national insurance number where recorded. He normalises names by removing spaces and converting to lowercase before comparing. Matches are classified as certain or possible. Certain matches go to the tenancy steward for quick confirmation; possible matches need careful review, because merging two different people would mix their personal data and rent accounts. The tenancy steward approves the rules after testing them on known cases.
11.3 Review and merging
The steward reviews possible matches using additional information such as previous addresses. Confirmed duplicates are merged in the housing system following a documented procedure that keeps the history from both records and records who merged them. Ravi tracks the number reviewed, confirmed and rejected. About a fifth of possible matches turn out to be different people, such as relatives with similar names, which justifies the manual review step for that group. Ravi reports these figures to the steward each month so the rules can be tuned if needed.
11.4 Prevention
To prevent new duplicates, Ravi suggests that staff search for existing tenants by date of birth and surname before creating a record, with the housing system showing likely matches. The system supplier configures this. He also adds a monthly duplicate check to the library. Ms. Moreau notes that prevention again proves cheaper than repeated clean-ups, and that matching rules should be reviewed whenever duplicates slip through, since every miss reveals a gap in the rules. The supplier's change goes live within two months.
Worked example
Ravi explains why tenant duplicates appear, writes normalised matching rules with certain and possible classes, has the steward review merges and finds a fifth of possible matches are different people, and adds search-before-create prevention.
Synthetic case. Records are synthetic.
Errors and limits
Merging without review can combine different people's data.
Exercises
- EX031 — Duplicate effects. Name one effect of tenant duplicates. Rubric: 3 = repeated letters, split rent accounts or incomplete data requests;; 2 = one;; 1 = vague;; 0 = none.
- EX032 — Normalisation steps. How are names normalised? Rubric: 3 = remove spaces and lowercase;; 2 = both;; 1 = one;; 0 = none.
- EX033 — Rejected share. What share of possible matches were different people? Rubric: 3 = about a fifth;; 2 = partly correct;; 1 = wrong;; 0 = none.
Continue studying — lesson 11 (chapter 11 of 30)
← Chapter 10 · Course chapter index · Chapter 12 →
Continue to the next authored lesson.