Technical Appendix 1 - Alternative Forms of Quantitative Indicators

Technical Information On Alternative Forms of Quantitative & Qualitative Indicators

Detailed Discussion of Accessibility Indicators

The following document should be read in conjunction with section 3 of Part A of the Technical Annex.

Access Measures

Access measures are one of the easiest of measures to both use and interpret. They have the least data requirements of any measure and can be utilised to assess the ease of access to the public transport network.

Access measures can be used by partners to build up a picture of the variation in accessibility within the study area and to identify potential gaps in the provision of public transport services, by varying the bus service frequency displayed or examining alternative time periods. Figures 1 and 2 depict an example of the use of transport maps in conjunction with access measures during the Tyne & Wear access to employment pilot. Here partners found it useful to overlay the home locations of jobseekers who were successful in finding employment with the spatial locations of their places of employment. By overlaying information on the spatial coverage of the bus network of a defined frequency (e.g. 2, 4, 6, 8 departures per hour), and producing walk buffers around the bus network of a defined frequency Nexus PTE were able to identify gaps in bus service provision and clearly show how bus service provision varied within Tyne & Wear. Nexus PTE were also able to highlight potential linkages between employment outcomes (i.e. jobseekers gaining employment) and public transport provision.

As well as using individual address information, it is possible to map census data, jobseeker data and data produced by partners thematically at a small area level (e.g. output area, super output area, ward level) and overlay the thematic map with public transport service frequency and transport coverage information as shown in figure 3. Here Devon County Council have utilised census data depicting the proportion of the resident population of college age to identify where there are relatively large numbers/proportions of the resident population with poor access to a suitable bus service.

Figure 1: Application of transport maps in the Tyne & Wear access to employment pilot

Figure 2: Application of access measures in the Tyne & Wear access to employment pilot

Figure 3: Application of access measures in the Devon access to learning pilot

The principal benefits of access measures are that they do not require extensive sources of data, typically requiring information on:

• the location of public transport stops and stations; and
• frequencies of public transport services at these nodes.

The data requirements of access measures are the least of any of the quantitative measures of accessibility listed above. In addition, they do not require the use of sophisticated software, and can be applied using a calculator or a simple spreadsheet. Access accessibility measures are particularly useful for building up the underlying story of accessibility within a local area and form a natural part of the hierarchical approach to accessibility analysis described in section 6.3 of Part A of the Technical Annex.

Access measures can be used to assess the general effectiveness ¹ of a proposed scheme or policy intervention. However, it is important to appreciate that access measures cannot by themselves be used to assess the economic-efficiency of a particular scheme, policy or action. In such instances the NATA discussed in section 6.6 of Part A of the Technical Annex should be used.

A fundamental assumption underpins the use of access measures; namely that once an individual has accessed the public transport network he/she is able to travel to any location within the study area, and that the characteristics of the onward journey by public transport are not relevant. This can be a significant weakness as it neglects differences in the distribution and coverage of public transport services that affect accessibility. In addition, it is important to note that the use of this form of indicator can encourage the development of certain types of policy intervention (such as increasing frequencies) that may not directly fulfil the actual needs of the communities under consideration.

Access measures do not allow the benefits arising from more flexible forms of transport provision, such as demand responsive transport, community transport, taxi buses and dial-a-ride services, to be evaluated. Also, the omission of the destinations from access measures means that they are unable to reflect the benefits arising from improved facilities and non-transport services, improvements in the location of opportunities/services and are unable to evaluate the benefits arising from alternative forms of service provision such as home delivery, home or community visits or e-services.

Threshold Measures

Threshold based measures are the most popular and easily understood form of accessibility measure. They are intuitive requiring no specialist knowledge on the part of the user. In contrast to access measures they incorporate characteristics of the entire journey. For instance journey time based assessments by public transport, incorporate the walk time from the origin to the boarding point (e.g. bus stop, railway station), the wait/interchange time(s) for the public transport service(s), the travel time onboard the public transport service and the walk time from the alighting point to the destination.

In addition to travel time, distance or cost, the generalised cost of travel can be utilised within threshold measures. The generalised cost of travel is a weighted combination of the travel time, the wait or interchange time, the distance travelled and the cost of travel between the origin location and the destination and is particularly useful when evaluating the benefits of transport schemes. Usually utilised in combination with continuous measures generalised cost of travel can equally be used within threshold based measures and are recommended for use in EIRs (refer to section 6.5 of Part A of the Technical Annex).

Examples of threshold based indicator are "the proportion of the elderly within a 10 minute walk of the nearest GP" or "the proportion or number of jobs that jobseekers can access within 60 minutes by public transport". Threshold measures like access and measures can be presented in absolute terms e.g. "1,500 households located within 30 minutes travel time of their nearest major centre" or in relative terms e.g. "30% of pupils are able to access their nearest primary school within a 15 minute travel time".

Threshold based measures must be used with some care, as they have important limitations, the principal one being the choice of threshold. For instance, in the above access to employment example, jobs located at a five minute public transport journey time from the individual's home are allocated the same weighting as jobs located at a 60 minute journey time, with jobs located 61 minutes or more minutes travel time from the home location are neglected. If used inappropriately to monitor progress and identify changes, they can omit accessibility changes that occur above the threshold. These limitations can be overcome in part by the use of multiple thresholds.

Thresholds can be used effectively when the threshold describes:

• A service or sector related management criterion e.g. JobCentre Plus do not provide jobseekers with details of job vacancies located more than an one and a half hours travel time away from their home, thus 90 minutes represents the service related management criterion (or reflects an administrative boundary which impacts upon the availability of a service). An example of this is that school pupils living over 3 miles travel distance from the nearest appropriate school are entitled to statutory free school transport. The 3 mile threshold or travel time equivalent can be used to assess the effect of this policy on those living within the 3 mile threshold who are excluded from free school travel but may have a need, such as children in receipt of free school meals or low income households.
• Travel behaviour (most notably walking and cycling) where the physical effort involved can be used to define behavioural thresholds for specific population segments. However, there is little evidence of the existence of such thresholds for motorised forms of travel.

It is important to note that use of a threshold in the absence of a behavioural or policy rationale can implicitly attach a policy significance to that threshold. For example, it can imply that the threshold denotes an appropriate journey time that reflects the concerns and priorities of the local community and at risk groups. Threshold based measures are most useful at a local level where local characteristics, management and service delivery requirements can be identified. However, information on local circumstances and the local pattern of travel behaviour should ideally be taken into account when setting appropriate thresholds within local indicators. For example, local travel diaries can be analysed to determine whether thresholds can be identified for the population groups in question, or targeted surveys of, and consultation with, the local community can be undertaken.

Threshold measures require transport, service and socio-demographic data as well as the use of accessibility or transport planning software. They are, in all but the smallest of local study areas, unsuitable for hand-based calculation or spreadsheet analysis. Threshold measures are better able than access measures to reflect an element of the degree of choice available to an individual when utilised in conjunction with facility/service data or census or other forms of household or individual data. However they are unable reflect the full extent of degree of choice available to an individual, which can only be achieved from use of continuous measures.

Threshold measures do not involve any fundamental assumptions that may affect their applicability to different areas and can be applied equally to both urban and rural areas. Threshold measures that consider access to/from the nearest opportunity or the average journey time, distance or costs to a series of destinations should not by themselves be used to assess the economic-efficiency of a proposed action, scheme or policy. In such instances the NATA discussed in section 6.6 should be used. However, these measures can be used to assess the general effectiveness of proposed interventions.

In the case of average travel time/distance/cost, care must be exercised, if the number of destinations utilised in the scenario evaluation is reduced; misleading (or simply incorrect) results can be obtained in which the average journey time may reduce implying an improvement in accessibility has arisen as a consequence of the removal of one or more distant facilities or services.

Threshold measures allow some of the benefits arising from flexible forms of transport provision to be evaluated, provided that information on these services is available. They are able to reflect the benefits arising from improvements in the location of facilities and services and are, to a limited degree, able to evaluate the benefits arising from alternative forms service provision such as home delivery, home or community visits or e-services.

Threshold measures are particularly useful for building up the underlying story of accessibility within a local area and form a valuable component of the hierarchical approach to accessibility analysis outlined in section 6.3.

Network accessibility maps can be produced from the journey time, distance, cost or generalised cost calculations produced during the course of a threshold based assessment. Such maps can be produced incorporating one or more destination; however, as a general rule the more destinations reflected within a network accessibility map the more difficult such maps will be to interpret. Figure 4 demonstrates one method of utilising multiple destinations within network accessibility maps as used by Surrey County Council and shows public transport journey times in Surrey to the nearest secondary school during a weekday am peak period. This type of accessibility map can be used to identify regions and locations where individuals or communities are not able to access key services within a defined threshold.

Figure 4: Public transport journey times to Surrey secondary schools & colleges - week day am peak

Source: Surrey County Council. © Crown copyright Ordnance Survey

An alternative method of considering multiple destinations is to calculate the average journey time, distance, cost or generalised cost to all service locations of a particular type e.g. hospitals, major centres, schools. Figure 5 demonstrates the use of this technique by Nexus PTE. Here the average public transport journey time from all residential locations to all destinations within Tyne & Wear is calculated enabling areas to be identified where individuals or communities are not able to access key services within a defined average travel time threshold.

Figure 5: Average public transport journey time from households in Tyne & Wear to all destinations

Source: Newcastle City Council, Tyne & Wear LTP, Automatica Ltd and Nexus PTE. © Crown copyright Ordnance Survey

The results of an average journey time, distance, cost or generalised cost network accessibility calculation can be used to compute the numbers of a particular type of population group or household type situated within a defined threshold. Figure 6 demonstrates how Nexus PTE have utilised this technique to compute the number of properties located within a 40 minute average public transport journey time of all destinations. By so doing Nexus PTE were able to identify at a glance where individuals or households are not able to access key services and locations within a defined travel time threshold.

Figure 6: Number of properties within 40mins public transport journey time of all destinations

Source: Newcastle City Council, Tyne & Wear LTP, Automatica Ltd and Nexus PTE. © Crown copyright Ordnance Survey

The hierarchical application of threshold based indicators is a powerful way of utilising threshold measures, particularly when a combination of percentages and absolute numbers are used during the accessibility audit. The use of multiple thresholds is particularly useful for building up the underlying story of accessibility within a local area and can form a valuable component of the hierarchical approach to accessibility analysis.

Figure 7 outlines the benefit of the progressive use of thresholds during a hierarchical analysis. In this instance, a ward-based threshold indicator of access to the nearest secondary school is presented as a series of thematic maps for the county of Surrey. By analysing each threshold in turn, starting with the lower threshold first (e.g. the 20 minute threshold map followed by the larger 45 minute threshold map, it is possible to build up a picture of where accessibility to the service is highest and lowest. In this particular instance, southern Tandridge and southern Mole Valley are shown to offer particularly poor public transport access to secondary schools for resident school children, whilst the more urbanised parts of the county offer better levels of accessibility to secondary schools.

Figure A7: Percentage of children within 20mins & 45mins public transport travel time of their nearest secondary school

Percentage of children within 20mins public transport travel time of their nearest secondary school

Percentage of children within 45mins public transport travel time of their nearest secondary school

Source: Surrey County Council. © Crown copyright Ordnance Survey

Threshold based measures are traditionally presented in tabular or graphical format; however, they can be presented spatially by an accessibility map as shown above. In this instance, thematic maps can be used to map the spatial variation in accessibility at a small areas level (e.g. output area, super output area or ward level). For strategic accessibility audits, it is best to analyse these maps at a ward level with output areas being utilised for local accessibility audits.

Continuous Measures

Continuous measures are the most behaviourally robust form of accessibility measures. They incorporate travel characteristics, characteristics of the facilities or services and a function reflecting the deterrent effect of time, distance, cost or generalised cost on travel. Continuous measures do not utilise thresholds to define separation, but instead use a continuous function to reflect the deterrent effect on likely travel of increasing time. Expressed as the "equivalent number" of jobs, schools, general practitioners, hospitals, major centres, destinations that are located closer to an individual or household contribute a greater proportion to the measure than the equivalent destinations located further away. With continuous measures the larger the indicator, the better is the accessibility; and the smaller the figure, the poorer is the accessibility. An example of a continuous measure is "the equivalent total number of jobs available throughout the entire study area from a given origin", which provides a measure of the level of choice available from that location. As depicted in figure A1 continuous measures like access and threshold measures alike can be presented in absolute terms e.g. "1,200 equivalent jobs" or in relative terms e.g. "25% of available jobs". Figure 8 demonstrates a simple worked example of the calculation of an origin based access to employment continuous measure for a single origin location. A process that is repeated for each alternative origin location.

In contrast to threshold measures continuous measures do not just consider the nearest facility or service location but all service locations under consideration, and thus provide a good indication of the potential level of choice available to the individual. The principal weakness of continuous measures is that they are not easily understood by non-modellers.

A range of continuous deterrence functions can be used within these indicators; however, the negative exponential type formulation e.g. ai = ¿j ajexp(-¿tij) a derivative of which is utilised in equation 1 (refer to section 10 of Part A of the Technical Annex) is among the most commonly used. The deterrence function (e.g. exp(-¿t)) and related deterrence parameter (¿) (refer to section 10.4), jointly reflect the aspect of travel behaviour that an employment location located 2kms away from an individuals' home will be less attractive than the same opportunity located 100m away. The deterrence function and associated deterrence parameter are based on a given person group's observed willingness to travel to reach the type of service under consideration and can be estimated from analysis of travel diary data. For access to employment, census travel-to-work data can be analysed to derive the associated deterrence parameters. However deterrence parameters are best defined through local calibration using local data sources. However in the absence of such data, those calculated nationally by the DfT can be utilised. Section 10.4 of Part A of the Technical Annex and Technical Appendix 6 on the accessibility planning website provides advice on how deterrence parameters can be calculated.

Figure 8: Example of calculation of an origin based Hansen/Gravity continuous measure

Whilst journey time, distance or cost can be utilised within continuous measures it is important to note that the generalised cost of travel (a weighted combination of the travel time, the wait or interchange time, the distance travelled, the cost of travel between the origin location and the destination) offers particular advantages when evaluating the benefits of transport interventions. Economic Impact Reports discussed in section 6.5 utilise generalised cost based continuous measures to evaluate the benefits of transport schemes in regeneration areas.

There are a number of alternative formulations of continuous measures ranging from simple origin or destination based measures to weighted measures that take into account the number of people affected to relative measures that consider relative rather than absolute changes. The mathematical formulations of the continuous indicators contained within the DfT accessibility planning software tool are available in Technical Appendix 2 on the accessibility planning website

One of the principal advantages of continuous measures over threshold measures is that they consider not only the nearest opportunity but all opportunities of a particular type, and thus provide a good indication of the potential level of choice available to the individual. The principal weakness of continuous measures is that they are not easily understood by non-modellers and partners from non-transport sectors. In addition, the units of measurement are not intuitive to non-modellers being expressed as the "equivalent" numbers of opportunities. However, this aspect of their ease of interpretation can be partly overcome by presenting the results as percentile based indicators using deciles or quintiles as outlined in section 10.2.1.

Figure 9 depicts and origin based continuous measure in this case bus accessibility to health care facilities within Merseyside. The figure shows that the more central regions of Merseyside have higher levels of accessibility to health care facilities in contrast to the peripheral regions of Merseyside.

Continuous measures do not involve any fundamental assumptions that may affect their applicability to urban and rural areas. They capture a number of the benefits arising from flexible forms of transport provision. They are able to reflect the benefits arising from improvements in the quality and number of services and opportunities and can, to a limited degree, enable an evaluation to be made of the benefits arising from alternative forms of service provision such as home delivery, home or community visits or e-services.

Figure 9: Bus accessibility to health care facilities within Merseyside - weekday am peak period

Origin and Destination Based Measures

The quantitative accessibility indicators discussed in sections 1.1-1.3 of this appendix can be expressed as either origin or destination based measures. In particular threshold and continuous measures can be expressed as both origin and destination based measures, whilst access measures are specifically origin in nature and cannot be expressed as destination measures. Depending upon the nature of the contributing indicators, comparative and composite indicators can be either origin or destination in nature.

Origin measures describe the ease of reaching opportunities (or the public transport network in the case of access measures) and reflect the user/individual/household perspective. Examples include "the number of jobs that a jobseeker can access within 60 minutes" or "the number of primary schools that a pupil can access within a 15-minute walk".

Origin measures can be used in both strategic and local accessibility audits and are particularly useful for comparative analysis within a study area and for the development of action plans. They can be presented in tabular, graphical or map format. They are ideally suited for reflecting the benefits/disbenefits of transport based policy interventions, but can also reflect the advantages and disadvantages of non-transport changes, such as proposed new services or opportunities, or the relocation or removal of existing services and opportunities.

Destination based measures describe the ease with which facilities can be reached, and reflect the perspective of service providers such as health centres, hospitals, employers, schools, colleges, shops, leisure centres. Examples include "the proportion of elderly within a 10-minute walk of their nearest GP" or "the number of secondary school age children within a 30-minute journey time by public transport of the nearest secondary school".

Destination based indicators have traditionally been more commonly used than origin based measures in accessibility modelling, reflecting the role played by the commercial sector in the use of accessibility modelling software.

Destination measures can also be used during both strategic and local accessibility audits and are particularly useful for comparative analysis within a study area, and over time, as well as for the development of action plans. They can also be presented in tabular, graphical or map format and like origin based measures are ideally suited for reflecting the benefits/disbenefits arising from the introduction of new services or opportunities or the relocation or removal of existing services and opportunities, as well as for highlighting the advantages/disadvantages of transport based policy interventions.

The choice between origin and destination based measures will, in essence, be dictated by the focus of the accessibility audit, the needs and priorities of the local community and partners and by the potential actions under consideration. During all pilots, both origin and destination based measures proved to be useful to partners. It is recommended that, where possible, authorities and their partners make use of a combination of origin and destination based measures. This should provide the partners with a fuller understanding of underlying pattern of accessibility within an area, as well as enabling both the user and service provider perspectives of proposed actions within the accessibility plan to be evaluated.

Qualitative Measures

Qualitative surveys can be used to inform an accessibility audit and the development of potential solutions and allow for aspects such as awareness of travel options, available services, travel information, individual perceptions about the quality of services and transport, crime and the fear of crime, travel horizons, barriers experienced in accessing local services etc to be captured. Depending on the sample size, the results of such surveys may be mapped; with smaller sample sizes they are best presented in tabular or graphical format. Qualitative surveys are particularly valuable in that they enable important factors not incorporated within accessibility maps and indicators to be considered. Such surveys can be undertaken at an authority level or in targeted localities.

Authorities are encouraged to consider incorporating specific questions on the accessibility of local facilities and services during general public consultations such as a) Local Transport Plans; b) Transport projects and service changes; c) Citizen's Panels and benchmarking or tracking surveys.

Qualitative surveys can target the wider population as in the aforementioned surveys or specific at-risk or disadvantaged groups, end users of local services, frontline professionals. However, care is needed when considering socially excluded groups, as their opinions and experiences are often under-represented in broader consultation exercises and targeted surveys may be required to adequately capture their views and experiences.

Targeted surveys and local consultations with end user groups and deprived communities can be relatively small scale, addressing specific local issues, which are not appropriate for more general consultations. In particular qualitative surveys can be particularly beneficial in identifying whether a) travel horizons are restricted by the cost of travel; b) scheduling constraints and household responsibilities impact upon the access of local services, c) information, awareness and perceptions of travel and service options impacts on the accessibility of local services, d) transport reliability and availability impacts on the accessibility of services

Qualitative surveys may provide valuable sources of information to aid development of local indicators and potential solutions / actions. For instance, qualitative surveys undertaken during the Nottinghamshire access to employment pilot revealed that the frequency and timings of public transport services were important issues for access to employment. The results of such qualitative surveys can be used to inform the choice of local accessibility indicators, which should ideally reflect the issues of concern to the local community and the resulting actions proposed during the development of the action plan.

Qualitative surveys are often perceived as being expensive however this need not necessarily be the case if a carefully targeted survey is undertaken focussing on specific issues and targeting specific groups or communities.

The results of household surveys undertaken within an authority may provide valuable information on issues of importance for accessibility planning. Analysis can be undertaken of the qualitative responses of specific population groupings to particular questions within the survey or by collating information collected by partners as part of their day-to-day activities. For example the Middlesbrough local Crime and Disorder Reduction Partnerships (CDRP) have been able to identify crime hotspots in the vicinity of the transport network by mapping reported crime at the street level. Figure 10 shows actual reported crime hotspots for Middlesbrough displayed as isochrones with the local highway and bus stop locations. The figure highlights that reported crime is greatest in the vicinity of Gresham, Middlehaven, Pallister and the University. Hotspots highlighted by such maps can be used to develop a particular package of actions as outlined in Chapter 6 of the main guidance document.

Figure 10: Local hotspots of disorder incidents in Middlesbrough

¹ Effectiveness is a measure of the extent to which a project, programme or policy achieves its objectives. Economic-efficiency is achieved with a scheme/policy when nobody can be made better off without someone else being made worse off.