ENV103 Chap.6 Data Collection, Accuracy and the MAUP
Data Collection, Accuracy and the MAUP
Data are gathered by surveys, observation, monitoring and sensors, and every route leaves its mark on quality. This chapter compares primary and secondary data and top-down and bottom-up sources, explains LiDAR, radar, sonar and satellite positioning, separates precision from accuracy, and shows how the Modifiable Areal Unit Problem lets boundary choices change a result.
Two calculations anchor the ideas: averaging GNSS fixes to separate bias from scatter, and regrouping four blocks to show a contrast appearing or vanishing with the boundaries.
What this chapter covers
- 01
Primary and secondary data
- 02
Volunteered geographic information
- 03
LiDAR, SAR, InSAR and sonar
- 04
GNSS and positioning from four satellites
- 05
Precision versus accuracy
- 06
Top-down and bottom-up data
- 07
Scale and zoning effects of the MAUP
Worked example · free
Judge precision and accuracy from two receivers' GNSS fixes
- +1Receiver A's mean position = (15.0 ÷ 5, 10.0 ÷ 5) = (3.0, 2.0), which is √(3.0² + 2.0²) ≈ 3.6 m from the mark.
- +1A's fixes all lie within about 0.2 m of that mean, so A is precise but not accurate.
- +1Receiver B's coordinates sum to (0.0, 0.0), so its mean sits exactly on the mark.
- +1B's fixes scatter an average of about 1.35 m from the mean, so B is accurate on average but not precise.
Key terms
- Primary data
- Primary data are collected in house for a specific project, through fieldwork or digitising.
- Secondary data
- Secondary data come from outside providers such as government agencies or vendors.
- Volunteered geographic information
- Volunteered geographic information is location data contributed by members of the public through apps, maps or social media.
- Accuracy
- Accuracy is how close measurements are to the true value.
- Precision
- Precision is how consistent and detailed repeated measurements are.
- LiDAR
- LiDAR is a ranging technology that times laser pulses to build detailed 3D models of surfaces.
- Zoning effect
- The zoning effect is the change in results caused by redrawing aggregation boundaries at a fixed unit size.
- Scale effect
- The scale effect is the change in results caused by aggregating the same data to larger or smaller spatial units.
Data Collection, Accuracy and the MAUP FAQ
What is the Modifiable Areal Unit Problem?
It is a bias that appears when individual data are aggregated into arbitrary areas. The same data and method can give different results if the units change size, the scale effect, or if their boundaries are redrawn, the zoning effect. Gerrymandering is the zoning effect used deliberately.
How does a GPS receiver work out its position?
Satellites broadcast their position and the exact time. The receiver measures how long each signal took to arrive, converts that into a distance, and combines distances from at least four satellites to fix latitude, longitude and altitude. Modern devices often combine several systems such as GPS, Galileo, GLONASS and BeiDou.
Are bottom-up data less trustworthy than census data?
They trade reliability for volume and speed. Census data are standardised and comprehensive but slow and inflexible; social media and app data are cheap and real-time but raise quality, privacy and access problems. Neither is automatically better; the right choice depends on the question and how carefully the data are checked.
What do LiDAR, radar and sonar have in common?
All three are ranging technologies that send out a pulse and time its return, much as you might call out in a canyon and time the echo. LiDAR uses laser light, radar uses radio waves and sonar uses sound underwater. The return time is converted into a distance, which builds up a model of the terrain, seafloor or object.
Assessment move
Organise this chapter as a set of contrasts, because that is how it is usually questioned: primary versus secondary, top-down versus bottom-up, precision versus accuracy, and the scale effect versus the zoning effect. For each pair, write a one-line definition of both sides and an Auckland example.
For the sensing technologies, make a small table of signal, platform and typical use for LiDAR, radar and InSAR, sonar and GNSS, and remember the canyon-echo analogy that links them: send a signal, time the return, convert to distance. Practise the precision and accuracy idea with numbers, as in the worked example, until you can explain why a tight cluster can still be wrong.
For the MAUP, redo the four-block zoning calculation with your own values and then weight the blocks by unequal household counts; seeing the answer move is the fastest way to understand the problem. When you work with census data in the labs, note the units each layer uses, meshblock, statistical area or suburb, and say so in any interpretation.
Finally, open the location settings on your phone, as the lecture suggests, and list which apps can see where you are; it makes the ethics discussion in Chapter 12 concrete.