Chapter 3 Fieldwork Procedures and Response Rate
3.1 Introduction
The NTS is a continuous survey with fieldwork taking place throughout the year, which has historically operated using a face-to-face (F2F) design since its inception. The methodological approach for NTS was altered in response to COVID-19 to ensure fieldwork could continue during the height of the pandemic and beyond in such a way as to minimise the effect on data quality and the time series as much as possible. These changes have been outlined in the Technical Report for the relevant years.
Throughout 2023, traditional doorstep recruitment was used, preceded by the interviewers sending out an advance letter as usual. While F2F was the primary approach, interviewers could also offer a telephone interview as back-up in cases where COVID-19 was a concern. This fallback approach was known as the phone back-up (PB) option. The travel records (also known as diaries) were filled in by the respondents themselves, except for PB cases for whom the diaries would instead be filled in by the interviewers over the telephone.
The traditional ‘fixed travel weeks’ allocation method was in use, whereby interviewers would provide a household with a specific week in which to record their travel in the diary. This methodical allocation process was designed to ensure a relatively even spread of travel week start days over (a) the dates of the month and (b) the days of the week, due to the random element that was central to the allocation method. The first travel week available for allocation was staggered approximately 12 days after the interviewer was scheduled to start doorstep recruitment, to allow enough time to contact addresses and conduct placement interviews prior to the travel weeks starting. Interviewers were required to make 1 mid-week call for initial checking of the diaries that the F2F respondents were filling in, and 2 mid-week calls for PB cases.
The computer assisted personal interviewing (CAPI) questionnaires on the interviewer laptops were designed and implemented using the software system Blaise. A single Blaise instrument was used for the household, individual, vehicle and administrative sections of the questionnaire.
A separate Diary Entry System (DES) was written in Visual Basic. Selected CAPI variables were extracted and loaded into the NatCen field management system from where they were referenced by the DES. This process provides contextual information from the CAPI interview for those people inputting and editing travel diary data.
Note: for this chapter, only comparisons of the 2023 data against 2022 data and 2019 data are included, as data from the pandemic years (2021 and 2020) was atypical for the NTS time series.
3.2 Interviewer briefings
In 2023 there were 2 types of briefings. New interviewers were briefed by NatCen’s research team during a series of 2-day online briefings. The briefings covered all aspects of the survey, including how to carry out the survey using either methodology (F2F or PB). The sessions included the demonstration of a dummy interview in the CAPI program, as well as role-play exercises to practise doorstep technique and the placing and picking up of the travel diaries. Interviewers were also given a pre-briefing exercise. This involved completing a travel diary themselves using their own journey details for the days preceding the briefing, studying the definitions manual, and completing a short test on how the diary should be filled in and the type of probing needed to gain the required level of accuracy and specificity.
Interviewers who had completed an NTS point (also known as an assignment) over the previous 12 months attended a half-day online refresher briefing shortly before the start of the survey year. This involved training on any changes that were new for the 2023 survey, including a diary checking practice and a reminder of the fixed travel weeks allocation method.
3.3 Questionnaire and document despatch to interviewers
The Operations Department made the sampled addresses and the CAPI questionnaire available to the interviewers for download onto their laptops via a secure broadband connection before the start of each quota month. The relevant NTS materials were despatched to the interviewers by post.
Any queries about transmission or other technical matters were dealt with by a helpline run from the Operations Department during working hours, and by a team of experienced interviewers working from home outside of working hours. Laptop maintenance was handled by a separate IT support department within NatCen. The interviewers were also able to contact staff within the Operations Department who deal with the administration of fieldwork.
3.4 Interviewer start dates and points
Interviewers were assigned to start fieldwork on different dates to ensure that interviewing was evenly spread across the year. The same principle applied to the travel week start dates, ensuring that the data recorded in the travel diaries was also spread evenly across the year, which is further underpinned by the fixed travel week allocation method used.
Interviewer points were allocated evenly across a quarter, rather than by month. When allocating points to quota months it is done in such a way a naturally representative sample is distributed for each quarter. See sections 2.7 and 2.8 for further information.
3.5 Contacting respondents
Interviewers were given advance letters to send to the selected addresses ahead of their first doorstep call (see Appendix B). For NTS 2023, the advance letter underwent various revisions intended to improve response and provide clarity on certain details so that informed consent from respondents was better achieved. For example, references to COVID-19 were removed from the ‘Frequently Asked Questions’ page on the rear side of the letter, and more broadly the tone of the letter was changed to emphasise the NTS as a legitimate social research study, avoiding language that could have been associated with market research.
The advance letter gave some general background to the survey and explained its importance, some of its uses, and how the household had been selected. It also stated that each respondent would receive a £5 shopping voucher if all household members completed the survey, the details of which were clarified as part of the 2023 letter revisions to improve response. See section 3.15 for more details on the shopping voucher for respondents.
The letters included a space for interviewers to write in their name so that respondents knew who would be calling and to make the letters more personal. As has been the case since June 2004 onwards, a book of first-class stamps was included with the advance letter as a gesture of goodwill to encourage respondents to take part.
Interviewers were notified of any refusals made to the Operations Department as a result of the advance letter being responded to by the residents of the sampled address. Interviewers were not required to visit these addresses and they did not count against interviewers’ individual response rates. However, they were classified as non-response (office refusals) in calculating the overall response to the survey.
A few days after the advance letters had been sent, interviewers contacted respondents by personal visit. Interviewers were required to make a minimum of 6 calls (that is, personal visits to the address) and to make sure that the days and times of these calls varied to maximise the chance of making contact with the selected household. If there was still no contact, only then could an interviewer return a case as a ‘non-contact’. Interviewers were also given a non-contact letter (which has been in place since November 2008) to post through the door of addresses where contact could not be made after 5 or 6 attempts (see Appendix C). The non-contact letter offered the household a further opportunity to learn about the NTS and to let NatCen’s Operations Department know if they wanted to take part. The changes that were made to the advance letter for NTS 2023 were also mirrored in the non-contact letter where relevant.
As in previous years, interviewers were also given information leaflets about the study (see Appendix D) and the NTS 2021 Factsheet so that, when necessary, they could demonstrate to respondents the type of data collected by the NTS. This was intended to help boost response and make the survey easier to sell on the doorstep.
3.6 Confidentiality
Respondents were informed in the advance letter that their participation was voluntary and that any information they provided would remain confidential and would not be passed on to anyone outside NatCen or the statistics section at DfT in a form that could be used to identify them. Respondents were provided with a telephone number and an email address for NatCen’s Operations Department that they could contact if they had any queries. Any substantive queries or complaints were subsequently passed on to researchers to deal with.
3.7 The placement interview
The first stage of interviewing consisted of the placement interview, using the CAPI questionnaire. This was conducted with all household members and consisted of 3 sections.
The household questionnaire was asked of the Household Reference Person (HRP), which is the householder with the highest income, or their spouse or partner. In exceptional cases the household questionnaire could be asked of another responsible adult aged 16 or over.
The individual questionnaire was asked of each household member, including children (although proxy information was always collected for children under 11 years old). A maximum of 10 people could be included. On the extremely rare occasions when interviewers encountered a household with more than 10 members, they were instructed to select the oldest 10 to take part in the interview, and to ensure that all vehicle owners were included.
The vehicle questionnaire was asked of the main driver for each vehicle in the household. A maximum of 10 vehicles could be recorded.
The percentage of individuals in 2023 who were interviewed in-person or by proxy for the placement interview (as well as those not interviewed) is shown in Table 3.1. The table is broken down by age into the 2 main age groups:
those aged under 16 (defined as a ‘young person’)
those aged 16 or over (defined as an ‘adult’)
Comparable figures for 2022 and 2019 are shown in Tables 3.2 and 3.3.
Table 3.1 shows that a high proportion of the individuals aged 16 or over were interviewed in-person in 2023 (69%), which is very similar to the proportion interviewed in-person in 2019 before the pandemic (70%) as shown in Table 3.3. The vast majority of individuals aged under 16 were interviewed by proxy in 2023 (86%), although this is extremely similar to the proportion interviewed by proxy in 2019 (85%).
Note: figures in the tables may not total 100% due to rounding.
Table 3.1: Method of individual interview at placement, by age (2023)
| Method of interview | Aged under 16 (young person) | Aged 16 or over (adult) | Total |
|---|---|---|---|
| In-person | 11% | 69% | 59% |
| Proxy | 86% | 30% | 40% |
| Not interviewed | 3% | 2% | 2% |
| Base (individuals) | 3,032 | 14,351 | 17,383 |
Table 3.2: Method of individual interview at placement, by age (2022)
| Method of interview | Aged under 16 (young person) | Aged 16 or over (adult) | Total |
|---|---|---|---|
| In-person | 10% | 67% | 57% |
| Proxy | 89% | 32% | 42% |
| Not interviewed | 2% | 1% | 1% |
| Base (individuals) | 1,685 | 8,182 | 9,867 |
Table 3.3: Method of individual interview at placement, by age (2019)
| Method of interview | Aged under 16 (young person) | Aged 16 or over (adult) | Total |
|---|---|---|---|
| In-person | 15% | 70% | 59% |
| Proxy | 85% | 30% | 41% |
| Not interviewed | 0% | 0% | 0% |
| Base (individuals) | 3,163 | 12,920 | 16,083 |
3.7.1 The 2023 NTS questionnaire
Each section of the placement interview covered a number of different topics:
The household questionnaire included:
the household grid
questions about home deliveries and food shopping
questions about children’s travel to school
the household vehicles grid
The individual questionnaire included questions about:
disabilities that affect travel
methods of transport used
walking
cycling
e-scooter
driving licences
internet use
(self-completion section) satisfaction with transport services, and perceptions of 1’s own social isolation
education, paid work and journey planning
last paid job
income
location of work
travel to work
working at home
ease or difficulty of travelling to work
transport difficulties
road accidents involving adults
road accidents involving children
special tickets or passes
long-distance journeys
permission for re-contact for follow up
The vehicle questionnaire included questions about:
the vehicle registration number
vehicle details
parking
mileage
Note: The coding of Standard Occupational Classification was carried out in 2023 according to the 2020 system.
The questionnaire specification is provided in Appendix A3, and an overview of how to interpret the questionnaire specification is provided in Appendix A1. Additionally, the questionnaire show cards are provided in Appendix H for reference.
Long-distance journeys
The NTS collects details about any long-distance journeys, defined as trips of 50 miles or more made within Great Britain. Where journeys of this type occur during the allocated travel week, these are simply recorded in the travel diary as usual.
Additionally, respondents are asked about long-distance journeys in the placement interview itself, but relating to a different (earlier) 7-day period. The week for which respondents were asked about long-distance journeys was normally the 7 days preceding the placement interview. However, in cases where the travel week was allocated to start date before the placement interview, the 7 days were instead taken to be the week preceding the start of the travel week. In total, a maximum of 40 long-distance journeys could be recorded during the interview.
Changes in 2023
The changes made to the NTS questionnaire in 2023 were very few, as follows:
minor changes to the wording of responses to the marital status question
minor changes to the wording of the consent question asking respondents if they would be willing to take part in future research such as NTAS
All changes to the questionnaire for 2023 are detailed in Appendix A2 and a summary of all questionnaire changes since 2002 is provide in Appendix A4.
Rotated questions
From 2002, some questions were designated to be ‘rotated’, such that they would be asked every other year. For 2023, the biennial rotating questions in Module B (odd years questions) were asked, and the questions in Module A (even years questions, last asked in 2022) were taken out.
Module B (odd years) included questions about:
home deliveries and food shopping:
Items that have been delivered and how often; and usual method of food shoppingchildren’s travel to school:
Barriers of children walking to school more and what would encourage more walking to school more; and reasons for travelling to and from school with an adultdisabilities that affect travel:Use of powered or manual wheelchair and how often it is used; whether a respondent is a carer and time spent caring; and prevalence of special transport services and usage of any of these special services
walking:
Barriers to walking more and what would encourage more walking in local areasdriving licenses:
Reasons for not driving among those without a provisional licence; and likelihood to learn to drivetravel to work:
Types of road used to travel to work; are respondents the driver or passenger travelling to work and how often is a lift given or received; and place of parking when driving to workworking from home:
Days of the week normally working from home; and extent to which it is possible to work from home and how much can be completed from home
For reference, Module A (even years) included questions about:
cycling:
Barriers to cycling; and what would encourage more cyclingdriving licenses:
Reasons for not driving currently among those with a licence; reasons for no longer holding a licence; and age last droveease or difficulty of travelling to work:
Difficulties with travelling to or from work by car, van, motorbike, scooter, or moped or by public transport; how easy is it to make journey to work, not using a car, van, motorbike, scooter, or moped; and what method of transport would they use insteadtransport difficulties:
Type of transport respondent has difficulty with:difficulties when travelling to the doctor’s surgery or hospital
difficulties experienced when visiting friends or relatives or other social activities
difficulties experienced when taking children to school or university
difficulties experienced when travelling for other purposes
parking:
Location of parking for vehiclesmileage:
Purpose and miles driven outside of Great Britain
Self-completion questions
Since 2017, attitudinal questions were asked of 1 member of the household aged over 16, administered by Computer Assisted Self-Interviewing (CASI). This member was randomly selected by the Computer Assisted Personal Interviewing program among the adults of the household present during the interview (proxy respondents could not be selected). See section 2.6 for more details on how the random selection was made.
The respondent completed the section on their own, using the interviewer’s laptop. The questions asked about (a) the respondent’s satisfaction with transport services, and (b) their own perception of how socially isolated they were.
3.7.2 Harmonised questions
A number of harmonised questions are used in the NTS to allow users of the data to compare NTS data with those from other social surveys. These questions are listed in Table 3.4, with the harmonised question name shown against the equivalent NTS question name. The final column of the table shows when the question was introduced into the survey and also provides additional information if the question has changed since that time.
Table 3.4: Harmonised questions used in the 2023 NTS
| Harmonised question name | NTS question name | Year introduced |
|---|---|---|
| Sex | Sex | 1998 |
| Age | AgeIf | 1998 |
| Date of birth | Birth | 2000 |
| Marital status | MaritalStat | 2013, with answer categories amended in 2023 |
| Living arrangements | LiveWithN | 2013 |
| Ownership of accommodation | Hhldr | 2002, with answer categories amended in 2003 |
| Joint ownership | HiHNum | 2002 |
| Ethnic group | EthGroup | 2001, with answer categories amended in 2011 and headings for answer categories amended in 2022 |
| Length of residence | HLongA | 1998 |
| Relationship to head of household | RelHoH | 1998 (question now asked as relationships of household members) |
| Relationships of household members | Relation | 2013, with answer categories amended in 2013 |
| Religion | Religion | 2022 |
| Accommodation type | Accom | 2000 |
| House type | HseType | 2000 |
| Flat type | FltTypN | 2000 |
| Other accommodation | AccOth | 2000 |
| Housing tenure | Ten1 | 1998, with answer categories amended in 2013 |
| Car ownership | UseVcl | 1998, with answer categories amended in 2003 and question text amended in 2004 |
| Vehicle type | TypeVcl2 | 1998, with answer categories amended in 2003 (question was deleted in 2004, and has since been imputed using TypeVcl2) |
| Company car | PrivVcl | 1998, with question text amended in 2009 |
| General health | GenHeal | 2017 |
| Length of residence | HLongInd | 2017 |
| In employment | Wrking | 1998 |
| Training scheme | SchemeET | 1998 |
| Away from work | JbAway | 1998 |
| Own business | OwnBus | 1998 |
| Relative business | RelBus | 1998 |
| Looking for work | Looked | 1998 |
| Starting work | StartJ | 1998 |
| Inactive | YinAct | 1998 |
| Industry | IndD | 1998 |
| Job title | OccT | 1998 |
| Job description | OccD | 1998 |
| Job status | Stat | 1998 |
| Paid employment | EverWk | 1998 |
| Date of leaving last job | DtJbl | 1998 |
| Supervising employees | SVise | 2001 |
| Organisation size | EmpNo | 1998 |
| Self-employed | Solo | 1998 |
| Number of employees | SENo | 1998, with answer categories amended in 2003 |
| Full or part time work | FtPtWk | 1998 |
| Long-term unemployed | HowLong | 2004 |
| Educational qualifications | EdAttn1 | 2005 |
| Professional or vocational qualifications | EdAttn2 | 2005 |
| Highest qualification | EdAttn4 | 2005 |
| Internet access | OnlineN | 2013, with answer categories amended in 2013 |
| Well-being | LackComp | 2019 |
| Well-being | LeftOut | 2019 |
| Well-being | Isolated | 2019 |
| Well-being | Lonely | 2019 |
| Disabilities that affect travel | Heallll | 2018 |
| Disabilities that affect travel | ImpCat1 to ImpCat10 | 2018 |
| Disabilities that affect travel | LimitAct | 2018 |
3.7.3 Length of the placement call
The average length of the placement call (that is, the placement interview plus the time taken to explain and, where necessary, complete the first days of the diary) for all households was just under 53 minutes in 2023, and varied according to household size, as shown in Table 3.5. Comparable figures for 2022 and 2019 are shown in Tables 3.6 and 3.7. These tables show the average time for different household sizes and the base figures for each row are also shown for reference in the final column.
As expected, Table 3.5 shows that the mean length of the placement call in 2023 tended to be longer for larger households, although the increase in time for each additional household member was not linear. For example, the mean length for a single-person household was 41.49 minutes, and then 55.11 minutes for a 2-person household and 57.26 minutes for a 3-person household. The longest mean time was 132.20 minutes for households with 10 people.
The 2023 figures are increasingly similar to those shown in Table 3.7 from before the pandemic in 2019. For example, the mean length in 2019 was 40.41 minutes for single-person households and 54.77 minutes for 2-person households, very close to the mean lengths in 2023. There are notable differences in mean lengths between these 2 survey years for households of 8 to 10 people, but the base sizes are very small, limiting meaningful comparisons.
Table 3.5: Mean length of placement call (in minutes), by household size (2023)
| Number of people in household | Mean length (minutes) | Base |
|---|---|---|
| 1 | 41.49 | 2,212 |
| 2 | 55.11 | 2,946 |
| 3 | 57.26 | 1,094 |
| 4 | 61.88 | 922 |
| 5 | 65.74 | 282 |
| 6 | 66.97 | 95 |
| 7 | 75.60 | 30 |
| 8 | 53.17 | 6 |
| 9 | 68.33 | 3 |
| 10 | 132.20 | 5 |
| All households | 52.95 | 7,595 |
Table 3.6: Mean length of placement call (in minutes), by household size (2022)
| Number of people in household | Mean length (minutes) | Base |
|---|---|---|
| 1 | 40.37 | 1,305 |
| 2 | 54.46 | 1,704 |
| 3 | 56.51 | 602 |
| 4 | 61.93 | 548 |
| 5 | 61.64 | 142 |
| 6 | 76.9 | 51 |
| 7 | 76.78 | 9 |
| 8 | 86.89 | 9 |
| 9 | 63 | 1 |
| 10 | not applicable | not applicable |
| All households | 52.08 | 4,371 |
Table 3.7: Mean length of placement call (in minutes), by household size (2019)
| Number of people in household | Mean length (minutes) | Base |
|---|---|---|
| 1 | 40.41 | 1,879 |
| 2 | 54.77 | 2,522 |
| 3 | 60.69 | 968 |
| 4 | 64.04 | 958 |
| 5 | 65.23 | 273 |
| 6 | 69.08 | 79 |
| 7 | 71.79 | 24 |
| 8 | 69.50 | 6 |
| 9 | 80.20 | 5 |
| 10 | 89.33 | 3 |
| All households | 53.63 | 6,717 |
3.8 Allocation of Travel Weeks
Respondents were required to record their journeys in the travel record for a period of 7 days (the period known as the travel week). Interviewers were given pre-selected lists of travel weeks, and from these lists the travel weeks were allocated to households by the interviewers using a methodical and unbiased process whenever they reached a definitive outcome.
To achieve this, each quota month was assigned a non-overlapping sequence of 30 to 31 travel week start dates. Individual interviewers were given a bespoke list (the ‘Travel Week Allocation Card’) of 22 travel week start dates, each of which was selected at random. This was achieved in such a way as to ensure, as far as possible, that an equal number of travel weeks would begin on each day in that month’s sequence (that is, evenly spread across all the interviewers working on a point belonging to that particular quota month). This means that it was possible to capture not only variations in travel behaviour within a month, but also differences across days of the week.
Because the number of trips that respondents record often falls away as their travel week goes on, the selection and allocation processes have been designed so that an approximately equal number of travel weeks start on each of the 7 days of the week. If, for example, all travel weeks were to start on Mondays, results would be biased toward travel behaviour on Mondays, Tuesdays and Wednesdays, whereas Saturday and Sunday travel behaviour would be under-represented.
Interviewers working on the 2023 survey were instructed to use the traditional ‘fixed travel week’ allocation method. This meant assigning the first travel week in their Travel Week Allocation Card list to the first address at which they reached a definitive outcome. They would then assign the next travel week on their list to the second address at which they reached a definitive outcome, and so on.
This method was applied universally, with the exception that when interviewers had to allocate a travel week that was in the past in relation to the interview itself they would therefore need to skip any travel weeks that started more than 2 days before the interview, and instead allocated the first travel week available on their list that started within 2 days before the interview. This exception exists to minimise the risk that respondents forget to include journeys or forget the specific details of journeys because the travel took place too far in the past for them to remember correctly. However, typically most travel weeks would start in the future in relation to the interview, especially if the interviewer had started their fieldwork as early as possible and did not suffer any unforeseen delays to their fieldwork progress.
Definitive outcomes include all the non-productive outcome types, such as deadwood addresses, refusing households, and any other type of non-participating address. For participating households, a definitive outcome was reached (that is, for the purpose of allocating a travel week) when the interview began, and not before, so as to avoid allocating travel weeks to households who eventually break interview appointments. The CAPI program prompted the interviewer to enter the allocated travel week at the very beginning of the interview.
Allocating travel weeks in this methodical manner, including to the non-participating addresses, is the best way to ensure the travel records of the participating households cover the entire survey period evenly. It also avoids biasing the spread of diary data collection that would occur if the respondents were to choose the travel week themselves.
The ‘rolling travel week’ approach (whereby the travel week always started on the day before the placement interview) was introduced in 2020 to replace fixed travel weeks as it was not possible to control for the timing of the placement interview as part of the push-to-telephone methodology. It was retained in 2023 as a fallback option, only used in exceptional cases where the interviewer had been allocated their point late into the fieldwork period, and permission was required to use them.
3.9 Placing the travel diary and other documents
For 2023 the approach for placing the travel diary and other documents followed the traditional NTS F2F pre-pandemic approach. This approach involved the interviewer introducing and placing the documents at the end of the placement interview with the intention that the respondents would fill in the documents themselves. The 7-day travel diaries (see Appendix E) were placed for each household member, and the mileage charts (see Appendix G) were placed for each household vehicle.
For PB cases, the interviewer completed the 7-day travel diary with the respondent, collecting the information from them over the phone and populating the paper version on behalf of the respondent.
The 7-day travel diary
Everyone in the household was eligible to complete a 7-day travel diary in which their travel activity was to be recorded, completed either by the respondent themselves, by a proxy within the household, or by the interviewer (depending on the methodology used for the interview). There are 2 versions of the travel record:
one for adults (respondents aged 16 and over)
one for children (the young person’s travel diary, aged 15 or younger)
Each trip was recorded, and details of origin and destination, purpose, mode, distance travelled, time, number travelling in their party, vehicles used, tickets used and cost was collected. In addition, the adult version of the travel diary collected details about of any parking costs as well as indicating whether they were a passenger or driver.
For F2F cases, interviewers explained to respondents in detail how to complete the travel diary. They generally did this by entering the details of some typical journeys made by the respondent in the blank pages provided, often using the respondent’s previous day’s journeys. Some interviewers referred to the NTS definitions manual to help describe the level and type of details required. Since 2014, interviewers have also been provided with a list of key points to cover when placing and checking a travel diary.
For PB cases, interviewers collected information from respondents via telephone to populate the travel diaries. Therefore the need to explain the nuances of completing the diary was lessened for PB cases, since the interviewer would simply probe the respondent for the necessary level of detail required as they were populating the diaries.
Note: All survey definitions are given in the NTS definitions manual, copies of which are available on request.
Simplified pocket size diaries or memory joggers (see Appendix F), into which respondents could briefly note down their journeys, were placed with F2F respondents if the interviewer felt they would be helpful, and were placed with all PB respondents so that they could make notes about their travel which they could later refer to when explaining their travel to the interviewer over the phone. The memory joggers however did not replace the need for a diary to be completed, they were simply provided as a note-taking aid.
The travel diary was redesigned in 2007 following an extensive development study. Full details of this study are available on the Department for Transport’s website. For further detail see McGee A, Gray M and Collins D (2006), NTS Travel Record Review Stage 1; and see McGee A, Gray M, Andrews F, Legard R, Wood N and Collins D (2006) NTS Travel Record Review Stage 2.
Minor revisions were made to the diary in the following years
Table 3.8 Minor revisions to the travel diary
| Year | Minor revision |
|---|---|
| 2009 | More rows added to the diary to record journeys, and text added to remind respondents to record short journeys |
| 2013 | Removal of diary elements related to cost of road tolls, congestion charges and taxi sharing |
| 2017 | Short walking day for recording walks of under 1 mile was changed from Day 7 to Day 1 |
A full list of changes that have been made to the travel diary since 2002 can be found in Appendix A4.
For 2023, the travel diaries collected information on:
purpose of journey
time left
time arrived
origin: where the journey started (from village, town or local area)
destination: where the journey ended (to village, town or local area)
method of travel (car, bus, walking)
distance (in miles)
time travelling (in minutes)
number of people in party travelling
which car or motorcycle was used (if journey was made by private transport)
ticket type: single, return or travel card (if journey was made by public transport)
cost (if journey was made by public transport)
number of boardings: the number of trains or buses. used to reach journey destination (if journey was made by public transport)
On Day 1 of the travel week, information was collected on all walks over 50 yards. However, on Days 2 to 7, only walks that were 1 mile or more were recorded (approximately equivalent to taking 20 minutes or more).
Additionally, the adult travel diaries (for those aged 16 or over) collected information on:
whether the respondent was the driver or the passenger (if journey was made by private transport)
how much was was paid for parking (if journey was made by car or motorcycle for example)
The young person’s travel diaries (for those aged under 16) also collected additional information on the number of people in party travelling, by splitting out the party into the number of adults and the number of children (whereas in the adult travel diaries only the total number of people was recorded).
The mileage chart
In addition to the diaries, a mileage chart was placed at the end of the placement interview for each household vehicle. The driver was encouraged to keep this chart in their vehicle. The chart required the driver to record the milometer reading at the start and end of the travel week. For the small proportion of PB cases the mileage chart was provided to respondents as part of the doorstep recruitment stage.
See Appendix G for a copy of the mileage chart.
3.10 The reminder call
Once the travel diary has been placed, the next stage is to remind the household to start recording their journeys on the date allocated to them, as the allocated travel week would in most cases start on a date after the placement interview. Interviewers did this either by either making a reminder call visit to the household (or leaving a reminder card if the household was absent at the time the interviewer made the visit) or by making a reminder phone call. Reminder calls would usually take place 1 or 2 days before the start of the travel week and were used for both F2F and PB cases in 2023.
3.11 The mid-week check call
Interviewers were strongly encouraged to conduct a call halfway through the travel week for F2F cases, to motivate respondents and help them with any difficulties they might be having filling out their travel diaries. This could be either a phone call or a personal visit and was at the interviewer’s discretion, although they were encouraged to conduct an in-person check, especially for elderly participants.
For PB cases, the mid-week check was used by interviewers to complete several days of the travel diaries on the respondent’s behalf, and interviewers were strongly encouraged to conduct this over 2 calls (rather than 1 call for F2F cases). These checks were always completed over the phone for PB cases. Ideally the first mid-week call would take place on Day 2 of the travel week and the second mid-week call on either Day 4 or 5. The intention of conducting 2 calls was to help ensure that the data quality on the travel records remained as high as possible, by minimising the number of days between the respondent travelling and then having to recall their travel to the interviewer over the phone. Additionally, spreading the mid-week check over 2 calls helped to break the activity down into smaller, more manageable portions.
The proportion of households receiving a mid-week check in 2023 is shown in Table 3.9. Comparable figures for 2022 and 2019 are shown in Tables 3.10 and 3.11. These tables show the figures for each possible outcome:
a mid-week conducted by phone
a mid-week check conducted in person
no mid-week check conducted.
The table is broken down by fully co-operating and partially co-operating households.
In 2023, just under 77% of fully productive households had a mid-week check (either by phone or in-person), which is very close to 2019 (80%). Of those mid-week checks, 28% were conducted in-person in 2023 compared to 30% in 2019.
Note: figures in the tables below may not total 100% due to rounding.
Table 3.9: Proportion of households where a mid-week check was conducted (either by phone or in-person), by fully and partially co-operating households (2023)
| Type of check | Fully co-operating | Partially co-operating | Total |
|---|---|---|---|
| Mid-week check conducted by phone | 48% | 41% | 47% |
| Mid-week check conducted in-person | 28% | 13% | 26% |
| No mid-week check | 23% | 46% | 27% |
| Base (households) | 6,314 | 1,279 | 7,593 |
Table 3.10: Proportion of households where a mid-week check was conducted (either by phone or in-person), by fully and partially co-operating households (2022)
| Type of check | Fully co-operating | Partially co-operating | Total |
|---|---|---|---|
| Mid-week check conducted by phone | 56% | 40% | 53% |
| Mid-week check conducted in-person | 18% | 11% | 17% |
| No mid-week check | 25% | 47% | 29% |
| Base (households) | 3,646 | 723 | 4,369 |
Table 3.11: Proportion of households where a mid-week check was conducted (either by phone or in-person), by fully and partially co-operating households (2019)
| Type of check | Fully co-operating | Partially co-operating | Total |
|---|---|---|---|
| Mid-week check conducted by phone | 50% | 33% | 48% |
| Mid-week check conducted in-person | 30% | 22% | 29% |
| No mid-week check | 20% | 44% | 23% |
| Base (households) | 6,162 | 663 | 6,825 |
3.12 The pick-up call
For F2F cases, at the end of the travel week, the interviewer called at the household (generally within a few days) to pick up and check the travel diaries and to carry out another much shorter interview, known as the pick-up interview. The pick-up interview collected information about:
new vehicles acquired since the placement interview (for the household)
the disposal of vehicles recorded at the placement interview (for the household)
new driving licenses acquired since the placement interview (for individuals)
new season tickets acquired since the placement interview (for individuals)
mileage details (for each household vehicle)
This pick-up call was also when the diaries were collected and checked before being sent back for coding by NatCen’s Operations Department. The mileage chart was collected also, and the details transferred into the CAPI questionnaire either during the interview or later on by the interviewer at home.
For PB cases, the pick-up interview was conducted over the phone, and he remaining days of the travel record were completed by the interviewer during the same call. The data collected by the mileage chart was collected over the phone at this stage too.
During F2F data collection, the pick-up interview was completed on the laptop using the CAPI questionnaire, or on rare occasions using a paper questionnaire as a fallback option (with the interviewer ensuring that the paper-recorded details were transferred into the CAPI questionnaire afterwards). The paper version of the pick-up questionnaire was introduced in 2002 to enable interviewers do the pick-up interview on the doorstep where respondents were unwilling to let them into the property again. For PB cases, the pick-up interview was always completed using the CAPI questionnaire over the telephone.
Table 3.12 shows the mean length (in minutes) of pick-up interviews for 2023 and all previous survey years going back to 2002, for both fully co-operating households and the total achieved sample of co-operating households. On average, the pick-up call (including the interview and checking the travel diaries) lasted about 14 minutes for fully productive households in 2023. This call was ideally made within 6 days of the end of the travel week. The 2023 average length of the pick-up call is slightly lower than the last pre-pandemic figure (14.5 in 2019).
The mean length of the pick-up interview reported here is calculated using the amount of time entered by the interviewer into the CAPI program. Although the length of pick-up is also calculated within the CAPI programme, this is not a reliable source because the pick-up interview is sometimes conducted on paper with the interviewer entering the information into the program at home. In previous technical reports, the pick-up length has not been calculated in a consistent manner. The pick-up interview lengths shown in Table 3.12 replace those in earlier NTS technical reports and are based on the interviewer-reported length.
Table 3.12: Mean length (in minutes) of pick-up interview in minutes, by fully co-operating and total co-operating households (from 2023 going back to 2002)
| Year | Fully co-operating | Total co-operating |
|---|---|---|
| 2023 | 14.0 | 14.0 |
| 2022 | 15.5 | 15.5 |
| 2021 | 18.1 | 18.0 |
| 2020 | 17.1 | 16.9 |
| 2019 | 14.5 | 14.4 |
| 2018 | 15.1 | 15.0 |
| 2017 | 14.7 | 14.6 |
| 2016 | 15.1 | 15.2 |
| 2015 | 15.5 | 15.3 |
| 2014 | 15.4 | 15.3 |
| 2013 | 15.1 | 15.0 |
| 2012 | 15.9 | 15.7 |
| 2011 | 16.4 | 16.3 |
| 2010 | 16.0 | 15.9 |
| 2009 | 16.4 | 16.2 |
| 2008 | 16.2 | 16.0 |
| 2007 | 16.6 | 16.4 |
| 2006 | 16.3 | 16.1 |
| 2005 | 18.6 | 18.5 |
| 2004 | 19.0 | 18.7 |
| 2003 | 18.6 | 16.3 |
| 2002 | 18.4 | 18.0 |
3.13 Gazetteer
The gazetteer is used to code the location of where respondents work and the origin and destination of any long-distance journeys during the CAPI interview. It is also used to code the location of journeys made in the travel record using the Diary Entry System. It holds a list of locations in Great Britain which is based on 1km grid references.
After the location of the respondent’s place of work has been established through the gazetteer, a follow-up question may be asked during the interview to determine more precisely where they work in relation to a common reference point. For respondents working in larger towns and cities they would be asked whether their place of work is within a short walking distance (between 5 and 15 minutes’ walk) from the reference point for their particular town of city, such as a landmark or a main street. Likewise, for respondents working in non-urban areas they would be asked if their place of work is ‘within 5 minutes’ walk of the main shopping or business centre’. The gazetteer maps these reference points against its list of locations, details of which are provided in the lookup table in Appendix J. Respondents working in other urban areas (that is, besides the larger towns and cities for which a reference point is given) would not be asked the follow-up question.
During the interview and the data checking stage, the CAPI and Diary Entry System uses the gazetteer’s grid references to calculate reasonably precise distances between each named location using checks based on straight line distances. For trips of 15 miles or over, respondents’ estimates of distance are flagged for checking if they are not between 0.75 and 1.75 ‘as the crow flies’ miles at the data processing stage. Discrepancies in distance estimates are not flagged where respondent and ‘as the crow flies’ miles are both below 15 miles.
3.14 Outcome coding
Interviewers were required to assign an outcome code to every address in their point. The fully and partially co-operating codes (110 to 130 and 241 to 262) were automatically computed for them by the CAPI program, whereas they had to input non-productive codes themselves into the CAPI program (with a small number of codes requiring office approval to be used).
Note: these fieldwork outcome codes are different to the participation categories that are used for the purposes of weighting.
For a household to be classed as fully co-operating, the placement interview had to be fully completed, and wholly filled in travel diaries had to be collected for all household members. The placement interview needed the household section, all individual interviews (whether in person or by proxy), and at least 1 vehicle section (if applicable) to be completed.
The household was coded as partially co-operating if any of the following applied, namely:
the household section of the placement questionnaire was not completed
anyone was coded as ‘not available’ for the individual section
no vehicle questionnaire sections were complete (if applicable)
travel diaries were not collected for all household members at pick-up
any of the travel diaries were incomplete (for example, missing days)
The range of possible fieldwork outcomes is shown below in Table 3.13.
Table 3.13: NTS outcome codes
| Code | Type | Outcome |
|---|---|---|
| 110 | Fully co-operating | Fully productive: All desired respondent(s) in person |
| 120 | Fully co-operating | Fully productive: Partly by desired respondent(s) in person, partly by proxy |
| 130 | Fully co-operating | Fully productive: By proxy |
| 241 | Partially co-operating | Partial productive: Desired respondent(s) in person, with some travel records |
| 242 | Partially co-operating | Partial productive: Desired respondent(s) in person, with no travel records |
| 251 | Partially co-operating | Partial productive: Partly by desired respondent(s) in person, partly by proxy, with some travel records |
| 252 | Partially co-operating | Partial productive: Partly by desired respondent(s) in person, partly by proxy), with no travel records |
| 261 | Partially co-operating | Partial productive: By proxy, with some travel records |
| 262 | Partially co-operating | Partial productive: By proxy, with no travel records |
| 310 | Non-contact | No contact with anyone at address |
| 321 | Non-contact | Contact made, but not at selected dwelling unit (used only for addresses with multiple dwelling units) |
| 322 | Non-contact | Contact made at (selected) household but not with responsible adult |
| 328 | Non-contact | Multi-household dwelling units only: contact made at (selected) dwelling unit, but not with anyone from selected household |
| 333 | Non-contact | Household opt-in, but telephone number wrong |
| 410 | Refusal | Office refusal |
| 421 | Refusal | Information refused about number of DUs at address |
| 428 | Refusal | Contact made but information refused about number of households |
| 431 | Refusal | Refusal before interview |
| 432 | Refusal | Proxy refusal |
| 433 | Refusal | Refusal of face-to-face interview due to COVID-19, no agreement for remote interview |
| 440 | Refusal | Refusal during interview |
| 450 | Refusal | Broken appointment, no recontact |
| 510 | Other unproductive | Ill at home during survey period |
| 520 | Other unproductive | Away or in hospital during survey period |
| 530 | Other unproductive | Physically or mentally unable/incapable |
| 540 | Other unproductive | Language difficulties |
| 550 | Other unproductive | OFFICE APPROVAL ONLY: Lost productive |
| 561 | Other unproductive | Household member ill with COVID-19 or related symptoms |
| 563 | Other unproductive | Household member is shielding from COVID-19 |
| 565 | Other unproductive | Household has been instructed to self-isolate |
| 591 | Other unproductive | Interview completed but respondent requested deletion |
| 599 | Other unproductive | OFFICE APPROVAL ONLY: Other unproductive |
| 612 | Unknown eligibility: no contact made | OFFICE APPROVAL ONLY: Issued, but not attempted |
| 614 | Unknown eligibility: no contact made | OFFICE APPROVAL ONLY: Issued, but not attempted due to COVID-19 local lockdowns |
| 620 | Unknown eligibility: no contact made | OFFICE APPROVAL ONLY: Inaccessible |
| 621 | Unknown eligibility: no contact made | OFFICE APPROVAL ONLY: Inaccessible due to COVID-19 |
| 630 | Unknown eligibility: no contact made | OFFICE APPROVAL ONLY: Unable to locate address |
| 640 | Unknown eligibility: no contact made | Unknown whether address contains residential, no contact made after 6 or more calls |
| 650 | Unknown eligibility: no contact made | Residential address, unknown if occupied due to no contact made after 6 or more calls |
| 690 | Unknown eligibility: no contact made | Other unknown eligibility (no contact made) |
| 710 | Ineligible or deadwood | Not yet built or under construction |
| 720 | Ineligible or deadwood | Demolished or derelict |
| 730 | Ineligible or deadwood | Vacant or empty |
| 740 | Ineligible or deadwood | Non-residential address (for example: business, school, office, factory, etc.) |
| 750 | Ineligible or deadwood | Address occupied, but no resident household (for example: holiday or weekend home) |
| 760 | Ineligible or deadwood | Communal establishment or institution, no private dwellings |
| 770 | Ineligible or deadwood | Residential, but not eligible for survey (for example, no-1 aged 16 and over) |
| 780 | Ineligible or deadwood | OFFICE USE ONLY: Address out of sample |
| 790 | Ineligible or deadwood | Other ineligible |
| 810 | Unknown eligibility: contact made | Information refused about whether address is residential |
| 820 | Unknown eligibility: contact made | Contact made, but not with someone who can confirm the presence of a residential household |
| 830 | Unknown eligibility: contact made | Information refused about whether resident(s) are eligible |
| 850 | Unknown eligibility: contact made | Unable to confirm eligibility due to language difficulties |
| 890 | Unknown eligibility: contact made | Other unknown eligibility (contact made) |
For 2021, in response to the COVID-19 pandemic, some new outcomes were added to the survey which were also used in 2022 and (as shown in Table 3.13 above) in 2023, namely:
433, refusal of face-to-face interview due to COVID-19, no agreement for remote interview
561, household member ill with COVID-19 or related symptoms
563, household member shielding from COVID-19
565, household has been instructed to self-isolate
614, OFFICE APPROVAL ONLY: Issued but not attempted due to Covid-19 local lockdowns
621, OFFICE APPROVAL ONLY: Inaccessible due to COVID-19
Addresses which were not issued to an interviewer or which were issued but not coded out (and therefore deemed to be unworked) were defaulted to the outcome code of 612 at the end of fieldwork for the quota month.
3.15 The shopping voucher incentive and thank you letter
Each member of the fully responding household was given a £5 shopping voucher (in the form of a physical gift card) by interviewers at the end of the pick-up call. Interviewers explained to respondents that the gift card would be activated by the office within 2 to 3 working days. For PB cases, the same £5 incentive was used, but this was sent out by NatCen’s Operations Department due to the fact that the pick-up call would have been conducted over the phone.
Starting in 2019, interviewers also handed out a thank you letter at the end of the pick-up call (see Appendix I). The thank you letter was addressed from the DfT and given as a recognition of the respondents’ time and effort. For PB cases, an equivalent letter was mailed out by the Operations Department along with the gift card(s).
3.16 Response rates
Tables 3.14 to 3.17 show the response rates in 2023 for England as a whole (the national response rate), the response for Inner London and Outer London, and finally the response for the rest of England excluding London. These tables show the response rate for the Achieved Sample, which includes all sampled addresses, and also the Standard Response Rate, which includes eligible households only and is the measure by which response is monitored year on year.
The overall national Standard Response Rate in 2023 was 32%, but this was lower in Inner London (23%), and very slightly higher in Outer London (33%). The response rate for the rest of England (excluding London) was 32%.
Table 3.14: NTS National (England) response rates
| Category | Achieved Sample | Achieved Sample Response Rate | Standard Response Rate |
|---|---|---|---|
| Set sample | 21,758 | 100% | not applicable |
| Ineligible or deadwood | 1,683 | 8% | not applicable |
| Unknown eligibility | 2,641 | 12% | not applicable |
| Eligible households | 19,842 | 91% | 100% |
| Fully co-operating | 6,314 | 29% | 32% |
| Partially co-operating | 1,279 | 6% | 6% |
| Refusal to co-operate and other unproductive | 7,837 | 36% | 49% |
| Non-contact | 2,004 | 9% | 13% |
Table 3.15: NTS Inner London response rates
| Category | Achieved Sample | Achieved Sample Response Rate | Standard Response Rate |
|---|---|---|---|
| Set sample | 1,540 | 100% | not applicable |
| Ineligible or deadwood | 158 | 10% | not applicable |
| Unknown eligibility | 156 | 10% | not applicable |
| Eligible households | 1,364 | 89% | 100% |
| Fully co-operating | 308 | 20% | 23% |
| Partially co-operating | 103 | 7% | 8% |
| Refusal to co-operate and other unproductive | 601 | 39% | 52% |
| Non-contact | 214 | 14% | 18% |
Table 3.16: NTS Outer London response rates
| Category | Achieved Sample | Achieved Sample Response Rate | Standard Response Rate |
|---|---|---|---|
| Set sample | 2,112 | 100% | not applicable |
| Ineligible or deadwood | 184 | 9% | not applicable |
| Unknown eligibility | 123 | 6% | not applicable |
| Eligible households | 1,917 | 91% | 100% |
| Fully co-operating | 631 | 30% | 33% |
| Partially co-operating | 128 | 6% | 7% |
| Refusal to co-operate and other unproductive | 823 | 39% | 48% |
| Non-contact | 223 | 11% | 13% |
Table 3.17: NTS England excluding London response rates
| Category | Achieved Sample | Achieved Sample Response Rate | Standard Response Rate |
|---|---|---|---|
| Set sample | 18,106 | 100% | not applicable |
| Ineligible or deadwood | 1,341 | 7% | not applicable |
| Unknown eligibility | 2,362 | 13% | not applicable |
| Eligible households | 16,564 | 91% | 100% |
| Fully co-operating | 5,375 | 30% | 32% |
| Partially co-operating | 1,048 | 6% | 6% |
| Refusal to co-operate and other unproductive | 6,413 | 35% | 49% |
| Non-contact | 1,567 | 9% | 12% |
As mentioned in section 2.4 the NTS oversamples Inner and Outer London with the aim of achieving responding sample sizes that reflect the regional distribution without the need for corrective weighting. The degree of oversampling in 2023 was based on estimates of differences in response rates between Inner London, Outer London and the rest of England.
From 2006 onwards, weights were introduced in order to correct for non-response (see Chapter 5 for a detailed description of the weighting). Data back to 1995 have been weighted retrospectively.
3.17 Back-checking and quality control
Like all NatCen projects in the field, the NTS was back-checked to ensure that interviewers were working to the standards to which they were trained and in accordance with the specific project requirements on which they were briefed.
A minimum of 10% of the total productive interviews were back-checked, the majority (usually 90%) by telephone, but where this was not possible (usually 10%) by letter. If the responses received indicated significant deviations from the standards set, a supervisor was asked to personally revisit the address(es) concerned. Back-checking was carried out usually within 2 weeks, and always within 4 weeks, of the interview date. The back-checking exercise did not find any systematic errors in the way interviewers were working for 2023.