Chapter 6 — Data quality basics
Lesson map
Lesson 06 published. Map, example, EX021.
flowchart TD
N0["Learning outcomes"]
N1["6.1 Dimensions"]
N2["6.2 Applying to counts"]
N3["6.3 Choosing checks"]
N4["6.4 Quality for a purpose"]
N0 --> N1 --> N2 --> N3 --> N4
N4 --> EX["Worked example"]
EX --> Q["Exercise EX021"]
Logical drawing
flowchart LR
N0["Learning outcomes"]
N1["6.1 Dimensions"]
N2["6.2 Applying to counts"]
N3["6.3 Choosing checks"]
N4["6.4 Quality for a purpose"]
N0 --> N1 --> N2 --> N3 --> N4 --> Q["EX021"]
[Learning outcomes]
|
v
[6.1 Dimensions]
|
v
[6.2 Applying to counts]
|
v
[6.3 Choosing checks]
|
v
[6.4 Quality for a purpose]
|
v
[Worked example] --> [Exercise EX021]
Learning outcomes
You will (1) name common data quality dimensions; (2) apply them to passenger data; (3) pick the right check for each.
6.1 Dimensions
Mr. Osei introduces quality dimensions in plain words, based on Wang and Strong's research with data users. Completeness asks whether all expected data are present. Validity asks whether values follow the rules. Accuracy asks whether data match reality. Timeliness asks whether data arrive when needed. Consistency asks whether different systems agree. Marta writes an example from Linkway for each dimension in her workbook to help her remember them. Mr. Osei checks her examples and suggests one change for timeliness, which she makes so that it matches how Linkway actually uses counts.
6.2 Applying to counts
For passenger counts, completeness means every scheduled trip has a count. Validity means counts are whole, non-negative and below capacity. Accuracy means counts match the real number of passengers, which can be checked by a person counting on a sample of trips. Timeliness means counts upload by the next morning. Consistency means count totals roughly agree with ticket sales on the same trips. Marta writes these definitions in a table in her workbook, with one row per dimension and a column for the check that tests it at Linkway.
6.3 Choosing checks
Marta matches checks to dimensions. Her Python script covers validity. A comparison with the timetable covers completeness. A morning check of upload times covers timeliness. A monthly comparison with ticket sales covers consistency. Accuracy needs manual counts, which Mr. Hughes arranges on ten trips each month. She notes that no single check covers everything, so several simple checks together give a fuller picture. Mr. Hughes agrees to run the checks on a regular schedule, with the results shared each month with the operations manager and planners.
6.4 Quality for a purpose
Mr. Hughes explains that the authority cares most about completeness and accuracy, because funding depends on total passengers. Planners care more about timeliness, because they review busy trips weekly. Marta learns that quality depends on what the data are used for, so checks should focus on what matters most to each user. She adds a column to her table showing which users care about each dimension. Mr. Hughes uses the table to explain to the operations manager why manual counts are worth the time they take each month.
Worked example
Marta learns five quality dimensions, applies them to passenger counts, matches checks to each including monthly manual counts for accuracy and notes which users care most about each.
Synthetic case. Examples are synthetic; Wang and Strong (1996) is real.
Errors and limits
Checking only validity misses missing or late data.
Exercises
- EX021 — Accuracy check. How is count accuracy checked? Rubric: 3 = manual counts on sample trips;; 2 = partly correct;; 1 = vague;; 0 = none.
- EX022 — Consistency comparison. What are counts compared with for consistency? Rubric: 3 = ticket sales;; 2 = partly correct;; 1 = other;; 0 = none.
- EX023 — Authority priority. Which dimensions matter most to the authority? Rubric: 3 = completeness and accuracy;; 2 = both;; 1 = one;; 0 = none.
- EX024 — Planner priority. Which dimension matters most to planners? Rubric: 3 = timeliness;; 2 = partly correct;; 1 = other;; 0 = none.
Continue studying — lesson 06 (chapter 06 of 18)
← Chapter 05 · Course chapter index · Chapter 07 →
Continue to the next authored lesson.