Paul Froes recalls his experiences of Road-Side Interviews – from banana skins to encounters with celebrities – and contrasts them to the far more scientific world of mobile network data, which provides far more robust insights at a fraction of the cost.
An insider's view to roadside traffic data collection
Paul Froes, Head of Operations at Citi Logik, previously worked for Jacobs and Tracsis, two firms that regularly collected data on traffic movements by conducting Road-Side Interviews (RSIs) with drivers and public transport passengers. This took him and his colleagues nation-wide into a strange world of banana skins, hexagonal pencils and an encounter with the Hollywood actor Guy Ritchie.
“Traffic flows at specific points and journey times along required links could be collected using non-invasive methods. However, in order to collect data about where traffic was going from and to the reasons why, we would need to ask the drivers and passengers about this,” Paul said. “The questions would depend on the client and overall requirements, but basic questions would include why and how often they made the journey.”
The location for traffic interviews
Identifying suitable sites to stop traffic for interviews was a big challenge. Vehicles had to be able to stop safely and the sites had to have toilets and refreshments nearby to avoid having to set up mobile toilets or lay on shuttle buses for staff. Seeking approvals from local and highway authorities typically took months to organise, and the operation was an expensive one: for example, each of the 100+ sites for a project involving the London North and South Circular ring roads cost several thousands of pounds each to run.
At some locations, we would set up a line of cones to divide the stopping site from the main carriageway, and I’ve seen people drive over the cones to re-join the main carriageway and avoid being interviewed.
Police stopped the cars, and although drivers were legally obliged to halt, they did not have to co-operate. “Most people answered questions, but we often got some abuse, especially if they were on their way to work or on their way home. Even a delay of two to three minutes would frustrate them,” Paul said. “At some locations, we would set up a
line of cones to divide the stopping site from the main carriageway, and I’ve seen people drive over the cones to re-join the main carriageway and avoid being interviewed. This obviously put themselves and other road users at risk, and often the police would intervene. ”
Random encounters with celebrities
Occasionally, they encountered celebrities as part of their interviews: one of the most memorable was when a member of Paul’s team interviewed Guy Ritchie. “He was very friendly and co-operative – he had his and Madonna’s kids in the back, as he was taking them to school,” Paul said.
Secrecy was the key for planning roadside interviewing
Secrecy to planned roadside interviews was paramount to avoid skewing the results. “We tried to avoid the interview locations being publicised, because if it was mentioned on local radio, people may take a different route to avoid delays. When I started there wasn’t social media, but nowadays people would tweet about it as soon as they are stopped, or put something out on their local Nextdoor or Facebook group and people avoid the road where the stop site is located, which impacts results considerably.”
We tried to avoid the interview locations being publicised, because if it was mentioned on local radio, people may take a different route to avoid delays.
The reliance on good weather
When no suitable bypass or lay-by site could be found, or the weather was poor, the team would distribute self-completion postcards instead of interviewing drivers. Many of the postcards came back with abuse scrawled across them or joke responses. “One was filled in by a Mr Monkey, who said he was on his way to the grocery store to get bananas,” Paul said. “He even enclosed an empty banana skin in the envelope.” Postcards were more likely to be returned by people with strong opinions “positive or negative – usually negative”.
Collecting data on the move
Sometimes Paul interviewed people on public transport, and bus operators told him they had to use hexagonal pencils to write down responses, as they were less likely to roll along the floor if dropped and potentially cause accidents than round ones. Health and safety concerns have led some service providers to pull out of roadside interviews due to the risks of employees working near moving traffic.
The alternative: Collecting transport data using MND
Nowadays, there are much more efficient and accurate ways of collecting transport data, such as Mobile Network Data (MND). Some clients still insist on conducting to face-to-face interviews simply because that’s the way it’s always been done. The use of MND requires a slightly different way of thinking and extrapolation for use in transport models.
Origin, destination, mode and purpose with MND
“There are pros and cons with all different data types,” Paul said. “MND, you get the devices’ origin and destination and you can accurately infer the purpose based on the devices regular day and night time location, and zoning based on land use, and mode on the route and speed. For example, using MND you may infer a journey purpose as shopping if the destination is a retail location, while if you speak to someone by the roadside you can find out exactly why they are going there along with other information such as number of occupants, detailed demographics, where they parked etc.
There are pros and cons with all different data types, with MND, you get the devices’ origin and destination and you can accurately infer the purpose based on the devices regular day and night time location, and zoning based on land use, and mode on the route and speed.
Collecting data over longer timeframe
Additionally, roadside interviews are traditionally conducted over a 12-hour period on one day at each location and give a snapshot in time, whereas MND covers one or two months on average, enabling trends and changes in patterns to be identified far more easily. A problem or incident on the road network can invalidate all RSI’s on that day, whereas using MND, specific dates or times can be removed from the overall analyses easily. It is also far less expensive, more controllable and the sample rate is much higher with MND.
Mobile network data gives you more of a sense of what the population are doing as they move around the road network in a more consistent way, which is really much more useful tool for planners and decision-making.
"With roadside interviews, you get more detailed information on the specific person you are interviewing and the occupants of their vehicle, but that person and journey may not be representative of a ‘normal’ day. Additionally, the detail you’re able to gather depends on the skills of the interviewer and how willing the interviewee is to answer the questions in full. Many people may have changed their route to avoid the delay caused by the interviews. It's dangerous to extrapolate from that to a wider trend. Mobile network data gives you more of a sense of what the population are doing as they move around the road network in a more consistent way, which is really much more useful tool for planners and decision-making.”