Chapter 3

Question 1

Define statistics

Answer 1

gathering, analysing and interpreting data in order to obtain the maximum
quantity of useful information.
• Concerned with decision making under uncertainty;

Question 2

Two difficulties in making correct inferences about sampled population:

Answer 2

• how to ensure a representative sample; and

* how to extract valid conclusions from said representative sample

Question 3

sample vs population of interest

Answer 3

Sample - collection of units especially selected to represent a larger population
Population of interest - group from which researchers try draw conclusions; subset of general population

Question 4

Sampling methods:

Answer 4

• Simple random sampling - equal and non-zero probability of being selected
• Stratified random sample (i.e. income groups / minority groups) - target population separated into mutually exclusive, homogenous groups and then a simple random sample chosen from each group.
• Choice based sampling (based on a previous choice made i.e. mode choice) (may not be random thus risk of bias)

Question 5

List 5 issues within SA regarding data collection:

Answer 5

• Digital literacy and access
• Language barriers
• Technology (tablets to collect data, drones / videos)
• Actual addresses / none registered population
• Private vehicles / Minibus taxi operators (illegal / non‐registered)
• Gated communities

Question 6

Differentiate between sampling error and bias:

Answer 6

Sampling error
• Function of sample size
• Due to only having a sample and not the population. Always present. Variability (deviation) could be different from population and sample, which influences confidence.

Sampling bias
• Due to incorrect definition of the population (i.e. youth and acceptance of pay as you drive insurance) , sampling method (random, but ignores gated estates)
• Distorts actual parameters
• Can be alleviated by better planning of data collection techniques and defining your sample

Question 7

Why is sample size a trade-off?

Answer 7

trade‐off between too large / too small sample: Cost, time, accuracy implications

Question 8

List some practical constraints in data collection:

Answer 8

• Length of study (i.e. transport time of year makes very big difference)
• Study horizon (Type of planning? Tactical (short term) or strategic planning with long study
horizon
• Limits of study area
• i.e. Stellenbosch municipality / broader Stellenbosch area – what will be the border in 20 years time?
• Study resources
• Issues with map reading in ZA

Question 9

What is a data collection plan

Answer 9

very important, costly and time consuming exercise

• information on recruiting, training, questionnaire design, supervision and quality control.

Question 10

What are some ideal dataset characteristics?

Answer 10

• Inclusion of all modes of travel, including non‐motorised trips.
• Trip purposes at disaggregate level
• Coverage of the broadest possible time period
• Data from all members of the household.
• High‐quality information robust enough to be used even at a disaggregate level

Question 11

List and explain some types of surveys

Answer 11

• Household surveys - Main input to classic four‐step modelling approach; Provides personal / individual characteristics of individuals and households; Provides travel information (travel time, travel cost, O‐D); Perceptions of transport
• Intercept surveys / external cordon surveys - People crossing a border (PT boardings / alightings)
• Intercept surveys / internal cordon / screenline surveys - People crossing area / road / railway within the study area
• Traffic and person counts: low cost; used for calibration, validation and for further checks to other surveys.
• Travel time surveys: for calibration and validation. Private & public transport (use GPS tracking)

Question 12

Level of detail of network and zoning systems are a trade of between accuracy and cost, details depend on:

Answer 12

• the schemes to be tested,
• the type of behavioural variables to be included,
• the treatment of time, etc

Question 13

List some zoning criteria

Answer 13

• Zoning size ‐ if too big, then assume activities / population in centroid?.
• Compatible with other administrative divisions, particularly with census
• homogeneous as possible in their land use and/or population composition; census zones with clear differences in this respect (i.e. residential sectors with vastly different income levels) should not be aggregated, even if they are very small.
• Zone boundaries must be compatible
• The shape of the zones should allow an easy determination of their centroid connectors;