Enhancing Internal Trip Capture Estimation for Mixed-Use Developments (2011)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

C-1 This chapter describes a recommended procedural frame- work for conducting internal capture data collection at MXD sites. The framework collects the independent variable and internal trip making information required by the estimation methodology presented in Chapter 3. The audience of this chapter is the potential collector of internal capture data (whether typical traffic consultants, researchers, or public agency staff). The recommended framework consists of six steps, start- ing with the definition of the specific purpose of the data col- lection effort and concluding with the proper processing of the on-site interview survey data. These steps are described in detail later in this chapter. For the internal capture estimation method presented earlier in Chapter 3 to be effective, it must be based on con- sistent and correctly applicable data. Therefore, it is essential that there be consistency in the definitions used and the means by which internal capture data are collected. The data collec- tion framework is structured to be straightforward, easily replicated, and adaptable to any potential mixed-use land use and development type. The field data collection can be conducted with an experi- enced survey supervisor and low-cost or temporary personnel who are given specific training prior to initiation of the sur- vey. The data collection procedure described in Steps 4 and 5 may at first appear to be onerous. However, all data listed will be needed for a typical internal capture trip generation analy- sis. Special or limited studies may require more, less, or differ- ent data. Prior to collection of any data the desired outputs should be examined and the necessary field data determined. Even for such special studies, the recommended framework presented in this chapter will provide a good foundation from which to work. However, if the resulting data are to be consis- tent with other data collected in accordance with NCHRP Project 8-51, the procedures described in this chapter should be followed. Any deviations to add more data should not change the basic data described herein. The list of data to be collected for a typical analysis has been streamlined so that no extraneous data are collected. There are numerous types of information that could be interesting descriptors but that do not provide direct relevance to esti- mating internal capture. These extraneous data have been excluded from the data collection plan because requiring them would expand the volume of data to collect (and the cost), could intimidate or discourage a potential data collector, and could thereby hinder the collection of the important and rel- evant data. However, the entity conducting the survey may have other reasons to collect additional data. Need for Quality Assurance and Control An important component of the data collection effort is adherence to a quality assurance/quality control (QA/QC) program. The exact nature of the program should be at the discretion of the agency that is funding or conducting the data collection. However, at a minimum, a QA/QC plan should be developed at the outset and checks should be undertaken dur- ing each of the six framework steps. An important consideration in the QA/QC process should be definition of the level of precision desired. This should be one of the first things determined for each survey. It is critical that the internal capture data be compatible among mixed-use developments. One quality assurance action is to carefully digest the definitions and descriptions of both the develop- ments and the data to be collected and applied. Methodology Framework Step 1: Define Purpose of Data Collection Step 1 provides the structure and scope for the survey. It is used to identify what is to be collected, how the data are to be used, and where to collect it. Procedures for Internal Capture Surveys A P P E N D I X C

Use of Data The first step is to clearly specify the purpose of the inter- nal capture data collection effort. There are two basic choices: (1) to study specific land use pairs in MXDs or (2) to determine internal capture rates for a development that is similar to a pro- posed MXD under consideration. In either case, the purpose may be to enhance the existing internal trip capture database or to establish internal capture rates for a similar MXD. Under both choices, the data to be collected, the survey instrument, and the interview procedures remain the same. The only difference occurs in Step 2, when a data collection site is selected. Also important is how those data will be used. Is it to assess traffic impacts of a proposed development on roads in an area that already experiences congestion during certain periods, or will the data be used in a special generator estimate of trip generation for a regional forecast of daily travel? The specific use will influence selection of the study site as well as the season, day of week, and time-of-day when surveys should be conducted. Site Selection At first glance, it may seem that any MXD could be selected for data collection. However, mixed-use sites are rarely iden- tical and often are very different from each other. Their differ- ences may, in some cases, cause only small changes in internal capture. However, some seemingly minor differences (for example, in the proximity of uses or in an area with a differ- ent nearby land use mix) can cause substantial changes in internal capture. Therefore, it is important to select a develop- ment that is similar to the one to be analyzed or represented in the resulting database. It is also important to collect the complete set of data to help identify differences that could explain the need to interpret the comparable sites for slightly different characteristics. In other words, although two sites may appear the same, when individ- ual parameters are examined (e.g., actual walking distance between buildings), slight, yet important, differences may be revealed. Site selection should consider: • types or styles of development that the data will be used to analyze; • development land uses and mix; • size range of development; • development maturity (is it fully occupied and sufficiently vibrant?) • external conditions; • representativeness of the development in relation to sites the data will support analysis of; • external conditions, including competing opportunities, modes of access, economic strength of the area; and • willingness of the development(s) owners and/or managers to permit the surveys in a manner needed for the surveys. Timeframe An important element to establish when defining the data collection purpose is the timeframe for which internal capture data are desired or required. Internal capture rates at a mixed- use site may vary by the time-of-day, day of the week, season of the year. Therefore, select the following: • time-of-day such as the morning peak hour for the site, morning peak hour for the adjacent street, evening peak hour for the site, and evening peak hour for the adjacent street, and other peak hour of generator if it may be subject to traffic impact analysis; • day of the week (weekday, Saturday, or Sunday); and • season or month of the year (e.g., typical month, holiday shopping season, summer, school-in-session). In terms of data that would be useful for the enhancement of the overall internal capture database, refer to Step 2 for suggested timeframes for particular land use pairs. Step 2: Select an Appropriate Site If the purpose of the data collection effort is to enhance the existing internal capture database, selection of an appropri- ate mixed-use site should be based on the following criteria. • The site should be of a density and magnitude for which the potential for intra-site walk trips is significant. • Individual land uses should be totally accessible internally either by pedestrian pathways or by streets completely within the development being surveyed (i.e., no vehicular travel required to make trips between internal points on streets on or beyond the periphery of the development). • The mix of land uses should be representative of current or anticipated trends in mixed-use development. • The land uses at the site should interact with each other. If one component of the mixed-use site does not have definitive synergy with any other on-site use (i.e., the num- ber of on-site trips to or from that land use are miniscule or unlikely), the overall mixed-use site should be rejected because it really does not act like a true mixed-use site. Table C-1 shows the land use pairs the researchers con- cluded are best suited to both produce significant inter- nal trip capture based on data reviewed to date and exist C-2

in significant quantity in current and anticipated MXDs. Trip capture data collection should be prioritized for these uses. • The mix of land uses should be transferable. If a particular mixed-use site has a truly unique land use or tenant or set- ting, the internal capture data may not be applicable to other sites. • The site should be fully occupied (or nearly so), mature (at least three years old), and considered successful locally. • The area in which the development is located should also be mature and mostly built-out with a pattern of develop- ment normal for that type of area. • Buildings are conveniently accessible to each other, both by distance and by accessibility. • Parking is shared between land uses; the percentage of reserved spaces should be minor. • The data collection program should be able to isolate the trips to, from, and within the development. – There should be locations where representative samples of trip making to and from each individual land use can be surveyed. – To that end, it is essential that through traffic not com- plicate data collection at the site. Ideally, there should be no through traffic. – Where tube traffic counters are to be used, the design of external access points should be such that mechanical counting techniques will produce accurate vehicle counts (e.g., short driveway throats make it difficult to place tube counters to work properly), or if not, manual or video counts should be employed. If the purpose of the data collection effort is to determine internal capture at a site similar to a proposed MXD, the ana- lyst should take a slightly different approach. First, the ana- lyst must define the proposed MXD in terms of the indepen- dent variables collected in Step 3. In other words, compile the descriptive data for the proposed MXD as if it was the data collection site. Armed with that information, selection of a similar site may be possible. Identify a mixed-use site (1) with the same land uses, (2) a similar balance of land uses, (3) with similar site lay- out characteristics, (4) that is at least three years old, and (5), if possible, that is located near enough so that competing oppor- tunities are similar. In addition, follow the previous criteria. When data are to be collected for a similar development, it is always valuable to verify acceptance of transferability with the agency that will review and decide whether to accept the results. Advance concurrence with site selection and procedures usu- ally alleviates the possibility of having to collect data elsewhere. Step 3: Obtain Permission to Collect Data at Study Site After an appropriate MXD site is selected for the data collec- tion, it will be necessary to obtain the permission from the site owner or property manager. It is not possible or appropriate to collect the necessary data (especially the on-site interviews of site visitors, patrons, and workers) without their permission and cooperation. In most cases, the owner or manager will communicate with internal businesses, landlords, etc. In some cases, the survey supervisor may need to make direct contact to gain full permission. A primary objective of property management is to keep property ownership and property tenants content by, if pos- sible, maintaining the status quo. One means of achieving this objective is to prevent the occurrence of any problems for the customers, visitors, workers, etc. of their property tenants. To that end, the analyst should contact property management by phone and mail/email, and then meet as necessary to discuss the purpose and procedures of the data collection effort. During each contact, the analyst should convey an under- standing of the need (1) to not impede patrons and (2) to not C-3 Land Use Land Use Retail - Convenience Retail - Other Restaurant Office Residential Hotel Entertainment Retail (Convenience) Retail (other) Restaurant Office Residential Hotel Entertainment Table C-1. Priority land use pairs for data collection.

divulge proprietary or sensitive information. An incentive for property management to cooperate is to offer to include a site-specific question during the interview process (and to offer the opportunity to receive the survey results or a copy of the study report). If a good working relationship can be developed, property management can often help tailor the intercept sampling procedure for the site and to interpret the survey results. Step 4: Compile Descriptive Data on Characteristics of Site After a subject site is selected, all information listed in Table C-2 needs to be collected and compiled. Most of these data will quantify the independent variables that have been demonstrated to affect internal capture at the mixed-use site. Step 5: Collect Internal Trip Capture Data The on-site internal trip capture data collection effort must be comprised of at least two components. 1. Counts of people entering and exiting each establishment where interviews are being conducted. These counts are used as controls for expanding interview samples (since complete interviews will not be obtained from every person entering and exiting) to represent all people entering and exiting the establishment. 2. In-person intercept interviews of people as they enter/exit a building (or significant use within a building) to determine the origin/destination, mode and purpose of trips internal to the mixed-use site. Other data collection options such as mail-back questionnaires, employee surveys, and visi- tor surveys do not obtain all the information required to understand and accurately quantify internal capture at the study site. It is highly recommended that cordon counts of all persons by mode entering and exiting the survey site be made during the survey. This will provide information on mode of ingress/ egress as well as the number of external trips being gener- ated. This also provides the basis for an approximate check of expanded interview data. Step 5 is subdivided into eight specific steps/decisions that need to be completed to conduct a successful field survey. Step 5A: Specify Purpose of Internal Capture Data Collection Step 1 in the overall data collection framework requires the analyst to define the specific purpose of the data collection effort. It should be repeated here and with specific reference to the following questions and issues. • Within the specific MXD, is internal capture to be mea- sured between selected pairs of buildings or throughout the entire site? • Specify the timeframe of interest for determining internal capture. Plan to collect internal capture data for one or more of the following periods: – street peak hour – collect for at least one-half hour before to one-half hour after the known peak hour (i.e., for at least two hours total) to make sure the peak hour during the survey is actually covered. Check current ITE definition for the complete street peak hour definition to ensure the correct peak hour is selected (the weekday street peak hour is currently the highest 60 minutes of site plus adjacent street traffic within 7 A.M.–9 A.M. and 4 P.M.–6 P.M.) (3); – peak hour of generator – determine the highest morn- ing or afternoon hour of trip generation from trip gen- eration counts at the survey site. Survey from 1⁄2 hour before the beginning of the peak until 1⁄2 hour after the end of that peak hour; – midday – collect from 1 hour after the A.M. street peak hour to 1 hour before the P.M. street peak hour unless a shorter period has been established with the review agency for the resulting analysis; and – daily – Collect survey data during the active part of the 24-hour period (e.g., when businesses are open; between about 6 A.M. and 10 P.M. for typical residential). • Specify the preferred day of the week (weekday, Saturday, or Sunday), based on the period analyses are to cover. If a weekday, select a typical day of the week for the land uses to be surveyed. • Specify the preferred season of the year (holiday shopping, summer, school-in-session), based on the period analyses are to cover. Step 5B: Identify Buildings or Uses at Which to Collect Internal Capture Data Identify the specific buildings at which to collect internal capture data. This will include all buildings and occupants or a representative sample of each. Specifics will depend on resources available, the site size, the number of land uses to be surveyed, and agreements with the agency that will need to accept the survey results. Generally, for a single time period, it is desirable to have at least 50 usable interviews per land use (30 minimum). Generally sample sizes of less than 30 are avoided to ensure the sample results benefit from the central limit theorem that says the sampling distribution of the means will approach that of a normal distribution even if the popu- lation being sampled is not normally distributed (4). C-4

C-5 Data Specific Information Desired Comments Name Record the common name for the overall mixed-use development site Development Type Specify whether the site is contained within a single-block, multiple blocks or a district Site Maturity Record the year the site opened. If opened in stages, also specify the date of the latest significant building opening. Primary Tenant(s) Determine the primary tenant (i.e., the tenant that serves as the primary driving force behind the overall site being developed as a mixed-use site); some sites may have more than one major (anchor) tenant. Other Land Uses within Site List the other land uses within the site. Use standard nomenclature. ITE trip generation land use classifications are preferred. Overall Characteristics of Site Building/Area Names and Addresses If the overall site is subdivided into sectors with different names or building addresses, identify them. Site Plan Obtain a site diagram, sketch, plan, or aerial photo of the site, preferably to scale. The diagram should show: • overall site layout with building footprints, • building entrances and pedestrian pathways, • access points from street system, and • parking supply. Site Area, Size, and Density Record total site acreage. Record number of development units for each building or area (gross square footage, number of dwelling units); at a minimum, collect dwelling units listed for each ITE trip generation land use category. Identify the developed portions by phase for developments to be expanded (if applicable). Locations and Types of Access Document the overall site access plan for motorists (including delivery and service vehicles), pedestrians (including transit patrons), and bicyclists, including: • location of each access point, • type of traffic control at or serving each access point (i.e., signalized or unsignalized), and • transit stops and station entrances along with existing or planned transit service. Internal Circulation Facilities Locate the internal roadways and driveways used by motorists. Locate the pathways for pedestrians (and describe whether pathways are enclosed, covered, or open-air). Locate the pathways or lanes designated for bicyclists, if any. Physical Characteristics of Site Location and Quantity of Parking Document the location of single-use or shared parking facilities. • Record the quantity of spaces in each facility. • Document the type of parking facility (e.g., surface, garage). • Assess how much of the development truly shares parking. • Record the daily/hourly cost for parking. Building Size Quantify the building size in development units such as office building square footage (GSF), amount of leased retail space (GLA), number of restaurant or theater seats, or number of residential units. Also obtain the number of stories. Primary Land Use Identify the primary land use within the building as being either retail, restaurant, office, residential, hotel, entertainment, or other. If more than 5 percent of the building square footage is occupied by a secondary use, treat it as a separate land use so internal capture can be quantified. For both the primary and secondary land uses in a building site, classify them in accordance with ITE Trip Generation Land Use codes (1). List the ground floor uses separately since counts may be needed for each. Characteristics of Individual Buildings within Mixed- Use Site (This information is needed for each individual building or area.) Space Allocated to Individual Land Uses Quantify the space allocated to primary and secondary land uses (any exceeding 5 percent of the building). Since it may be desired to estimate trip generation for specific land uses, it is suggested that the land uses be disaggregated by the following land use categories (which are more detailed than the seven general land use categories listed above): • For retail, subdivide into: • Convenience (e.g., grocery, drug store, bank, dry cleaner) • Full service • Discount Other/specialty Other Table C-2. Descriptive data for MXD sites. (continued on next page)

C-6 Data Specific Information Desired Comments Characteristics of Individual Buildings within Mixed- Use Site (This information is needed for each individual building or area.) Space Allocated to Individual Land Uses • For restaurant, subdivide into: • Fast-food • Sit-down with no bar • Family • Quality • Sit-down with bar • Family • Quality • For office, subdivide into: • Boutique • General • Medical (nearly all space is doctor offices and medical related uses that serve patients) • For residential, subdivide into: • Single-family detached • Townhouse • Condominium • Rental apartments • For hotel, subdivide into: • High price • Mid-price with meeting facilities • Mid-price with no meeting facilities • Low-price • For entertainment, subdivide into: • Cinema • Other • For other, specify the use Building Occupancy Quantify the building occupancy (e.g., occupied office, retail, and apartments, not just leased). In a multi-tenant building, contact the property manager, leasing agent, or owner to obtain occupied space data. Building “Primary Access Point” or “Center of Gravity” if multiple access points are available Determine the main access point. If multiple access points exist, designate the “center of gravity” (or “access point”) for the building. One characteristic of a mixed-use site that has a significant effect on internal capture is the proximity of its complimentary uses. To measure this proximity, the trip end points must be defined at a certain level of precision. For some buildings (for example, a multi-story office building), the center of gravity seems obvious (in this example, the building lobby). However, for multi-tenant retail buildings, the definition of center of gravity is much less clear. For the purposes of internal capture data collection and data analysis, the following convention for determining a building center of gravity is used: • for an enclosed retail mall with more than one anchor store, use inside entrances for anchor stores. It is important to use the location of the mall-side, not outside, entrance; • for an open-air community or neighborhood shopping center or for an enclosed mall with a single anchor store, use the location of the main entrance for primary tenant. The primary tenant could be a grocery store, any other big box or a discount store; • for an office building, use the office lobby; • for a hotel, use its lobby or registration desk; • for a restaurant, use its main customer entrance; • for a residential site, use its approximate center of gravity of the ground floor dwelling unit entrances; and • for an entertainment facility, use its main lobby. Another possibility is to disaggregate all data to individual building entrances. In that case, no center of gravity needs to be determined. In any case judgment will often need to be used. Three examples of centers of gravity include: • midway between two entrances on the same building face if both have similar levels of inbound and outbound volumes; • center of block face with numerous entrances; and • center of block for a land use covering an entire block with entrances on each side, each with similar volumes. Table C-2. (Continued).

C-7 Data Specific Information Desired Comments Building Proximity Measure proximity of the building to each other building in the mixed- use site. Measure as walking distance along pedestrian facilities between building “centers of gravity” (as defined previously). The desired level of precision for each of the above measurements is 10 percent of the approximate total distance or 100 ft, whichever is less. Connectivity between Buildings (Not currently part of recommended procedure, but a consideration in evaluating internal connectivity) Rate the connectivity between the building and each other building in the mixed-use site, using the following scale: • fully-integrated uses – the pedestrian connection between uses is direct and internal to the development, and does not require crossing a parking facility at-grade; • outside sidewalks with at-grade, priority street crossings – pedestrians use street sidewalks. Any street crossings (whether midblock or at intersections) assign priority to pedestrians; and • informal – the pedestrian connection requires walking through parking aisles or along streets without sidewalks. Parking Supply Rate the parking supply within 600 ft of the building entrance for building tenants and visitors, in particular its convenience. Rate as either ample or limited (based on availability of parking at the ITE Parking Generation report rates) (2). Report total parking spaces and rates if any. Indicate number of spaces reserved for each land use and any time restrictions. Location within Urban Area Classify location of the overall site as either rural, suburban, urban, midtown/activity center, urban core, or special district Setting/Context of Site within Surrounding Region1 External Competition for Individual Components of Site (Not currently part of recommended procedure, but may influence internal capture) Consider the degree to which off-site land uses will compete with those on-site and assess if that will affect how representative the candidate site will be to the survey. A development with extreme off-site competition may have fewer internal trips than one with almost no competition. Selected data collection sites should be representative of typical conditions or of a similar proposed development to be analyzed. 1 Quantification of the site setting and context measures is facilitated if a GIS linkage is provided for the mixed-use site. Table C-2. (Continued). Step 5C: Identify Intercept Locations at Study Sites Identify all means/routes of entering or exiting the building (or significant use within the building) whether to make an internal or external trip. Identify the entrances/exits that can be used to make a trip internal to the mixed-use site, whether by foot, bike, or vehicle. At each of these latter entrances/exits, select an interview location. It is not necessary to interview at external site access points where only external trips from specific single-use buildings can be made, but these must be counted (person trips by mode) instead. This is because all trips directly between on-site build- ings and the external transportation system are (1) external, (2) can be added to trip interview data from that building, and (3) can be counted as person trips by mode. Usually this con- dition only occurs when a garage has access directly to an exter- nal street. Pedestrian access does not assure that the person is actually going external unless it is a direct connection to a tran- sit station or an off-site garage. Table C-3 provides guidance on where to conduct surveys. Survey Site Location Survey and Count Requirements Office building connected to a retail building by walkways at several levels in a fully- integrated mixed-use site; office building has elevator/stairs to parking garage Survey at either end of each walkway connecting the office and retail uses. Count (1) each walkway connecting office and retail and (2) people entering/leaving the office building via the garage or any other entrance. Stand-alone office building situated near or adjacent to a retail shopping center; parking provided on surface and below-grade (accessed via elevator or stairs in office building); pathway to retail leads to/from building lobby Survey everyone who passes through lobby or who uses garage (because a person could drive to the adjacent retail site and thus would be considered an internal trip). Count at lobby and garage entrances. Regional mall with nearby office and residential uses Survey at the mall entrances Count each mall entrance separately (including any outside entrances for anchor stores). Table C-3. Survey and count requirements for several sample locations.

Step 5D: Identify Count Locations at Study Sites The data collection program must include a count of all people (not simply vehicles) entering or exiting the building at which interviews are being conducted. Therefore, appro- priate count locations must be identified. These will usually be doors to the property being surveyed (count people enter- ing and existing), garage access points (count vehicles and occupants); there may be other access points. The count should keep track of entering and exiting peo- ple separately. The counts will be used for two purposes: 1. person trip generation count for establishment being sur- veyed and 2. for computing an expansion factor to be applied to the interview data. Separate data are necessary for survey factoring and for determining an overall internal capture rates for the surveyed site. Table C-3 provides guidance on the extent of a count program for sample mixed-use sites. The survey should include interviews at as many establish- ments as possible while obtaining the desired number of inter- views per land use during each survey period. Interviewers should be deployed to representative establishments within each land use. Under the best scenario, interviews will be con- ducted at each establishment. If that is not possible, conduct interviews at a representative cross-section within each land use. When using the sampling approach, deploy interviewers to the busiest locations in each land use. If interviewers are assigned to low volume access points, they will not complete many interviews. This may be partially offset by having inter- viewers intercept people at multiple adjacent establishments. A competent interviewer (actively approaches people to get interviews, responses are complete and accurately recorded) located at a moderately active entrance should be able to com- plete interviews with at least 10 people per hour. However, activity levels will vary and typically result in a range of 5 to 20 completed interviews per hour. An average interviewer should be able to obtain completed interviews from one out of every three to four persons approached. Recognize that some inter- view candidates will decline to participate or have been inter- viewed previously and not want to participate again. Step 5E: Determine Staffing Requirements For mixed-use sites, it is desired to conduct 50 or more interviews per land use per survey time period. This may not be possible for land uses that are small or are relatively in- active during the survey time period (e.g., weekday morning retail). One way an interview sample can be expanded is by conducting interviews during the same time periods over multiple days. The survey supervisor should determine how many survey- ors are needed, based on the survey location requirements described in Step 5D and on the minimum sample require- ments described previously. If there is a steady stream of pedes- trians at a survey location, a rate of 20 complete interviews per hour is a reasonable expectation for each surveyor. For less active locations, estimate 5 to 10 complete interviews per hour for well-trained interviewers who are experienced at approach- ing strangers. When estimating manpower requirements, it is important to assess the pedestrian traffic flow to be intercepted. Step 5F: Develop Survey Instrument and Other Data Collection Forms Interviews of persons are typically conducted as they leave a single land use or building within the site. Each interview can obtain information on both the trips to and from the sur- veyed building and to and from the overall mixed-use site. Figure C-1 provides a sample list of interview questions. The questions are written for exit interviews at building or garage access points (i.e., interviews of people as they leave a loca- tion). If the interview is to be conducted as people enter the location, the form shown in Figure C-2 should be used. In general, interviews should be conducted in both directions. However, if that is impossible, complete interviews con- ducted in one direction can yield usable data since informa- tion is asked in each interview for one outbound and one inbound trip. If the survey will be conducted at the cordon driveway or other type of location, the supervisor may need to revise the questions to capture the last (for exit interviews) or first (for inbound interviews) on-site stop. Other modifications may be needed for special locations or applications. The survey super- visor should make sure that the questionnaires to be used fit the conditions as well as collect the desired data. In general, use of questionnaires such as those shown in Figures C-1 and C-2 will be adaptable to nearly any standard survey and can be automated if desired. Each item is needed for a complete analy- sis or for checking responses. However, some survey sites may need supplemental questions to firmly and clearly establish the characteristics of the trips being reported. The field survey form should include a space for the inter- viewer to record the date, the name of the development, the interviewer’s location within the site, the time each interview begins, as well as the interviewer’s name. It is important that every single item be filled out completely and accurately for each interview. Omissions can make an interview unusable. Inaccurate entries, guesses, or incomplete entries will also invalidate an interview, wasting both time and money. Interviews will be completed for a sample of all persons exiting establishments or the site. Factoring will be used to expand the survey data to represent the universe of trips C-8

As persons DEPART intercept as they leave a specific entrance Interviewer name:__________________________ Building: ___________________ Date: ____________ Start Time: ________ A.M. P.M. 1 Hello. May I please have a moment of your time to ask you a couple of questions for a Mockingbird Station survey? Where are you headed now? How are you going to get there? Where did you come from immediately before you came to [name place being exited] How did you travel from there? What time did you arrive here on that trip? How did you initially travel to (name the study site) today? 1. Office 2. Retail 3. Restaurant 4. Residential 5. Medical office 6. Cinema 7. Hotel/motel 8. Other (specify) 1. Office 2. Retail 3. Restaurant 4. Residential 5. Medical office 6. Cinema 7. Hotel/motel 8. Other (specify) If not as driver, did you have an auto available for your trip here? Building Entrance Time 1. Within (name study site) 2. Outside (name study site) Specify business/building 1. Auto driver 2. Auto passenger 3. Walk 4. Rail 5. Bus 6. Bicycle 1. Within (name study site) 2. Outside (name study site) Specify business/building 1. Auto driver 2. Auto passenger 3. Walk 4. Rail 5. Bus 6. Bicycle 0. I live here 1. Auto driver 2. Auto passenger 3. Bus 4. Rail 5. Walk 6. Bicycle 1. Yes 2. No am pm am pm am pm am pm am pm am pm am pm am pm am pm am pm am pm 2 3 4 5 6 7 8 9 10 11 12 13 14 Figure C-1. Sample exit interview questionnaire.

Intercept persons as they look like they will ENTER a specific entrance Building entrance: _______________ Interviewer name:__________________________ Date: ____________ 1 2 Are you headed into (name of establishment where you are interviewing) Where are you coming from? How did you travel to get here? Before you were at (prior place) where were you before then? (Immediately prior to last place) About what time did you arrive there? How did you travel to get there? How did you initially travel to (name survey site) today? 9. Office 10. Retail 11. Restaurant 12. Residential 13. Medical office 14. Cinema 15. hotel/motel 16. Other (specify) 9. Office 10. Retail 11. Restaurant 12. Residential 13. Medical office 14. Cinema 15. Hotel/motel 16. Other (specify) If not as driver, did you have an auto available for your trip here? Time 3. Yes 4. No (If “no,” terminate interview) 1. Within (name survey site) 2. Outside (name survey site) Specify business/building 7. Auto driver 8. Auto passenger 9. Walk 10. Rail 11. Bus 12. Bicycle 3. Within (name survey site) 4. Outside (name survey site) Specify business/building 1. Auto driver 2. Auto passenger 3. Walk 4. Rail 5. Bus 6. Bicycle 7. I live here 8. Auto driver 9. Auto passenger 10. Bus 11. Rail 12. Walk 13. Bicycle 3. Yes 4. No am pm am pm am pm am pm am pm am pm am pm am pm am pm am pm am pm am pm am pm 3 4 5 6 7 8 9 10 11 12 13 14 15 Figure C-2. Sample inbound interview questionnaire.

represented in the survey. Counts of all persons exiting the survey locations (or all locations) will be needed to develop the expansion factors. This expansion process will need to be developed as part of the survey design so the proper counts can be made. Figure C-3 shows a manual count form that can be used to count people exiting (or entering) each door of each establishment where interviews are to be conducted or that the interviews are to represent. This form or an automated equiv- alent can be modified to meet specific survey site needs. Cordon counts may also be needed for factoring and/or checking total external trips. These counts should be direc- tional and by travel mode. Vehicle occupancies should be counted since the recommended estimation methodology (and therefore survey methodology) is for person trips. Counts should cover all access points. Figure C-4 shows a manual cordon count form that can be used for this type of survey. This form can be automated or modified as needed for specific survey conditions. Step 5G: Recruit and Train Field Personnel After recruiting the survey field personnel, the survey super- visor should conduct a training exercise. Some personnel will need to conduct door counts—the counts of people entering and existing establishments to be surveyed. Some personnel will conduct interviews. Generally the most outgoing and assertive staff will make the best interviewers. Retiring person- alities should not be deployed as interviewers but may make good counters. The interviewers should be made familiar with the survey instrument through practice of intercept interviews. The same is true for counters. All survey personnel should be provided with maps showing each location where counts and/or inter- views are to be performed. The survey supervisor should include on each map the overall MXD site with names of buildings, tenants, and areas to which interviewees might refer as well as the specific location and movements the counter or interviewer is to handle. Field surveys are not trivial. They require thorough prepa- ration and training as well as good supervision. Most surveys of this type will require one supervisor for each 10 to 15 inter- viewers and counters. Specifics of the survey site, including size and distribution of survey personnel, may increase or decrease the number of supervisors needed. Step 5H: Conduct Field Data Collection Inbound and Outbound Door Counts. As noted previ- ously, total person counts are needed at each location where intercept surveys are to be conducted. If several adjacent C-11 Figure C-3. Sample door count form. Location: __________________ Counter: _____________________ Date: ____________ Hour Starting ____:00 am pm 1 2 Business/PlaceMinutes after hour Direction 3 3 3 3 3 3 3 3 :00 to :15 In Out :15 to :30 In Out :30 to :45 In Out :45 to :00 In Out 4 5 6 6 6 6 6 6 6 6

establishments are to be surveyed, one counter may be able to count multiple doors concurrently. This will depend on sight lines and placement of the counter. A counter should only be assigned those movements to count that can easily be seen while looking in one direction. Requiring a counter to look in multiple directions will result in missed persons entering or exiting doors being counted. As mentioned previously, every establishment door where interviews are conducted must have entering and exiting peo- ple counted. Counts should be made by 15-minute periods beginning on the hour or half hour when the survey begins. Counts should be made for the complete survey period. The survey supervisor should have extra personnel to pro- vide short breaks for the counters to use restrooms. It is sug- gested that breaks be permitted every two hours. With cell phones now in common use, they can be used by survey per- sonnel to request restroom breaks, if needed before scheduled breaks. Survey personnel should be cautioned to stay hydrated, especially on hot days, but not to drink so much that frequent trips to restrooms are needed. Counters should be trained in what they are to do. Train- ing should be completed prior to the survey. Training often requires at least four hours and often more. It can be beneficial to begin the first day’s survey an hour early to make sure the survey personnel are comfortable with their job before the sur- vey period actually starts. On the first survey day, the super- visor should walk each counter to the assigned survey location. The supervisor should make clear what doors and movements are to be counted and where on the form each movement should be recorded (form for each counter should be set up in advance). The supervisor should ask each counter if he or she has any questions to make sure instructions are clear. After the survey begins, the supervisor should circulate among the counters to check to see that counts are being made and recorded correctly. Common problems are line of sight obstructions (unanticipated or resulting because counter moved), inattention, recording counts in the wrong column, not keeping track of time, talking to another counter, and socializing with passersby. Interviews. The survey supervisor should carefully recruit and select interviewers. The ideal interviewer is outgoing, assertive, willing to approach and talk to strangers, sounds pro- fessional, and understands the purpose and procedure for the interviews. The survey supervisor will need to train all survey personnel, but spend more time with the interviewers. It is rec- ommended that each interviewer perform a few practice inter- views under supervision prior to beginning actual surveys. On the first survey day, the supervisor should walk each interviewer to the assigned interview location. The supervisor C-12 Figure C-4. Sample cordon count form. Location: __________________ Counter: _____________________ Date: ____________ Hour Starting ____:00 am pm 1 2 Personal Vehicles Motorcycles Delivery/Service Trucks Occupants Riders Occupants Minutes after hour Direction 1 2 3 4+ 1 2 1 2+ Walk Bike :00 to :15 In Out :15 to :30 In Out :30 to :45 In Out :45 to :00 In Out 3 4 5 6 7 8 9 10 11 12 13 14

should make clear what doors and movements for which inter- views are to be conducted and make clear where the inbound and outbound trips are to be recorded. If appropriate, the supervisor should also discuss the strategy for approaching people to interview. The supervisor should ask each counter if he or she has any questions to make sure instructions are clear. Practice or test interviews are recommended. It may be bene- ficial to begin interviews an hour early the first shift worked by each interviewer to make sure the interviewer is comfortable and approaching and interviewing people correctly. After the survey begins, the supervisor should circulate among the interviewers to check to see that candidate respon- dents are being approached professionally and that interviews are being conducted and recorded correctly. Common prob- lems include: • shyness in approaching people to interview, • not asking questions correctly or leading respondents by guessing answers for them, • incomplete recording of responses, • not asking all questions, • not keeping track of time, • talking to another survey staff member, and • socializing with passersby. The selection of a representative and sufficient sample of workers, shoppers, visitors, and residents at the survey site is critical to the success of the survey. Therefore, the survey supervisor should closely monitor the real-time progress of the intercept surveys to make adjustments as necessary to achieve the representative and sufficient sample, keeping in mind the stated objectives for data collection effort. This may require redeployment of interviewers to different locations that have more activity or making other changes that will increase the number of usable interviews for each land use. Cordon Counts. One counter should be assigned respon- sibility for each cordon count location. Since the counter must be able to count not only vehicles, but also vehicle occupants as well as pedestrians and bicyclists, the counter will need to be close to where the cordon crossing is located. Ideally the counter can be immediately adjacent to the driveway, street, garage entrance, or other cordon location. In some cases, two adjacent cordon locations will be so close together that a single counter can count both with accuracy. In either case, each form should be set up specifically for the location(s) to be counted. A counter should only be assigned those movements to count that can easily be seen while looking in one direction. Requiring a counter to look in multiple directions will result in missed persons and vehicles crossing the cordon line. Counts should be made by 15-minute periods beginning on the hour or half hour when the survey starts. Counts should be made for the complete survey period. The survey supervisor should have extra personnel to pro- vide short breaks for the counters to use restrooms. It is sug- gested that breaks be permitted every two hours. Cell phones can be used by survey personnel to request restroom breaks, if needed before scheduled breaks. Survey personnel should be cautioned to stay hydrated, especially on hot days, but not to drink so much that frequent trips to restrooms are needed. Counters should be trained in what they are to do. Training should be completed prior to the survey. On the first survey day, the supervisor should walk each counter to the assigned survey location. The supervisor should make clear what move- ments are to be counted and where on the form each move- ment should be recorded (form for each counter should be set up in advance). The supervisor should ask each counter if he or she has any questions to make sure instructions are clear. After the survey begins, the supervisor should circulate among the counters to check to see that counts are being made and recorded correctly. Common problems are line of sight obstructions (unanticipated or resulting because counter moved), inattention, recording counts in the wrong column, not keeping track of time, and socializing with passersby. Use of Electronic Recording Devices. A number of elec- tronic survey recording devices, including laptop computers, are now available. They can be successfully used for these counts and interviews, if they are set up in formats that are easily used. Formats that do not allow counters or interview- ers both ease of use and logical positioning of response only invite confusion and errors. For example, use of an electronic intersection turning movement count board for a door count where several doors are to be counted by one person will probably not present a logical input format and lead to errors. Step 5I: Supervise in Field Survey supervisors should have a survey check procedure developed as part of the QA/QC procedure suggested at the beginning of this chapter. This procedure should be in place prior to training. The check procedure should include assign- ments of supervisors to check each counter and interviewer and how to perform the check. Supervisors should observe interviewers at work and suggest refinements in their approach and conduct of interviews. Spot checks of interview records should be made early in the first interview period to make sure the responses are both logical and complete. Supervisors should understand that errors in procedure usually continue until corrected. If not corrected, interviews for an entire day could be lost as unusable. The same is true for counts. It is important for the supervisors to keep circulating among those being supervised. Even though the counts or interviews C-13

are being performed correctly, other supervisory needs may arise. Common needs include complaints from business or landlords wanting survey personnel to relocate or stop their survey, unexpected movements that are being missed by the survey, too much activity for one person to cover, “no” activ- ity to count or interview, business opened or closed unexpect- edly, and survey staffer unable to perform as needed. Step 5J: Check Data after Each Period The survey supervisors should perform a check of the counts and interviews immediately after each survey period. The check should be included in the QA/QC plan, but should generally include at least the following: • Counts: – count covers full period; – inbound and outbound balances are logical; – variations by 15-minute period are logical; – modal splits are within the expected ranges; – vehicle occupancies are in expected ranges; – for cordon counts, it is desirable to total the counts to see if they appear reasonable, particularly the balances between inbound and outbound; – for door counts, for each land use, compare inbound and outbound totals to make sure the balance appears logical; and – if discrepancies are found, determine if corrections can be made, and if not, schedule a recount(s) as needed. • Interviews: – times of interviews are recorded; – are responses within range of permitted choices (i.e., are codes consistent with choices available)? – are write-in responses complete and understandable? – destination for outbound trip is logical and mode fits origin-destination pair; – origin of inbound trip is logical for reported time of trip (i.e., was it really the immediately prior trip?); is time reported for that trip logical for immediately prior trip? – check response to whether a vehicle was available for trip; is it logical for reported mode of trip? – is mode of access to site logical given mode reported for these trips? and – where discrepancies or errors appear to exist, review forms with interviewer (call as soon as possible while memory still clearest) to determine if corrections can be made or if interviews must be discarded. If necessary, repeat interviews where prior interviews had to be dis- carded. After the survey has been completed in the field, the super- visor should complete the checking of all counts and inter- views. Those that are unusable should be deleted. Erroneous counts should have been repeated. Small percentages of unusable interviews should be deleted. Large numbers should have been repeated. Step 6: Process Internal Capture Data For each survey site (establishment), the analyst should determine the number of usable interviews. Under normal cir- cumstances, 50 or more usable interviews should be available for each land use (100 desirable, 30 minimum). In some cases, this will not be possible because the land use will not be active (e.g., retail closed during A.M. peak hour) or because the quan- tity of development in a land use category will be small. That number can be compared to the total door counts for the same period. The sampling percentages can be calculated by divid- ing the number of usable interviews by the number of people counted in the same direction (inbound or outbound). The same can be performed for each land use by aggregating all establishments within specific land uses. Since the interviews represent a sample, the next step is to compute an expansion factor to expand the sample to repre- sent the total for that universe. This can be accomplished in at least two ways: • by land use (normal approach): – separate each interview record into individual trip records; there will be one or two usable trips in each interview record depending on how many occurred dur- ing the survey period; – aggregate by land use numbers of inbound and out- bound trips (aggregate to the interview end of the trip) reported during the survey period from those inter- views; this includes both trips reported in the interview if they were during the designated survey period (TL for each direction); – aggregate door counts to the land use level (CL for each direction); – determine number of development units (e.g., gross square feet) covered by interviews and the number of development units for which no interviews were con- ducted (in cases where only a portion of establishments within a given land use were interviewed); calculate a sample percentage for each land use (S); – the expansion factor (FL) for reported trips for each land use and each direction will be: FL = (CL/TL)/S; and – apply directional land use expansion factor FL to each trip record; and • by establishment: – separate each interview record into individual trip records; there will be one or two usable trips in each C-14

interview record depending on how many took place during the survey period; – aggregate by establishment the numbers of inbound and outbound trips (aggregate to the interview end of the trip) reported during the survey period from those inter- views; this includes both trips reported in the interview if they were during the designated survey period (TE for each direction); – aggregate by establishment the door counts (CE for each direction); – for each establishment surveyed, compute the expan- sion factor to apply to trips to and from that establish- ment; it will be the establishment’s directional door count divided by the establishments usable trips in the same direction (CE/TE); – determine number of development units (e.g., gross square feet) covered by interviews and the number of development units for which no interviews were con- ducted (in case where only a portion of establishments within a given land use were interviewed); calculate a sample percentage for each land use (SE); – Apply directional establishment expansion factor FL to each trip record for each establishment (Ei), then sum to aggregate trips to the land use level, or – FEi = (CEi/TEi)/SE; and – those expansion factors are then applied to trip records for each surveyed establishment; the sum equals the total for that land use. After the expansion factors are applied at either the land use or establishment levels, a summary of internal capture can be created. This should be performed for each end of a trip and in the inbound and outbound directions; that is: • Land Use A – outbound trips to internal destinations at each other land use, plus outbound trips to external desti- nations; and • Land Use A – inbound trips from internal origins at each other land use, plus inbound trips from external origins. Tables C-4 and C-5 show a format for this summary. Using the trip records and expansion factors from the survey, sum the expanded trips in origin-destination format. This should be a straight forward process to begin from the origin end of trips and sum to produce a table similar to Table C-4. This provides a distribution for all trips departing a given land use (the exam- ple shown is referred to as Land Use 3). Some trips will end in the same land use, although at another establishment. Some trips will travel to other internal land uses. Some will leave the surveyed development and travel to an external destination. All trips must travel to either an internal or external destination. For Land Use 3, those outbound trips will total 100 percent. Hence, each of the entries in the Land Use 3 row can be con- verted to percentages. For example, if there are 100 outbound trips from Land Use 3 and 8 trips travel to Land Use 2, then 8 percent travel to Land Use 2. Since this is internal, 8 percent were internally captured by Land Use 2 (see Table C-5). ITE has a large trip generation database built from counts of external traffic (vehicle trips) from single-use developments (or at least single classifications). ITE trip generation data excludes internal trips. For the Table C-4 data to match the ITE definition, internal trips must be deleted. Table C-5 shows C-15 To InternalFrom Land Use 1 Land Use 2 Land Use 3 Etc. External Total Land use 1 Number or % Number or % Number or % Number or % Number or % 100% Land use 2 100% Land use 3 4 (4%) 8 (8%) 20 (20%) 0 (0%) 68 (68%) 100 (100%) Etc. 100% External 100% To InternalFrom Land Use 1 Land Use 2 Land Use 3 Etc. External Total Land use 1 Land use 2 100% Land use 3 4 (5%) 8 (10%) 0 (0%) 68 (85%) 80 (100%) Etc. 100% External 100% Table C-4. Sample summary format—outbound trips. Table C-5. Sample summary format—outbound trips (ITE definition).

how that is accomplished. Movements between establish- ments within the same land use are not considered; they are deleted from the trip table. Table C-5 shows the hypothetical results with the trips internal to Land Use 3 deleted. The inter- nal trips to other land uses remain. The total trips external to Land Use 3 remain the same as do the external trips, which are the trips of most interest in transportation impact studies. Trips also travel into the surveyed development and its land uses. A similar summary of inbound trips can be created as Table C-6 shows. These numbers and percentages may be dif- ferent than the numbers in Table C-4. Logic supports such a finding. For example, in a MXD with retail, restaurants, and office, the restaurants will send few P.M. street peak hour (e.g., 5–6 P.M.) trips to office uses because few, if any, office workers will travel to their office at that time. However, restaurants may receive a significant percentage of their 5–6 P.M. trips from internal office uses (people going for an early dinner, drinks, or hors d’oeuvres). Hence, it would be logical to expect different directional percentages between office and restaurant during the P.M. street peak hour. References 1. Trip Generation, 7th Edition, Vol. 3: User’s Guide, Institute of Trans- portation Engineers, Washington, D.C., 2003, p. iv–viii. 2. Parking Generation, 3rd edition, Institute of Transportation Engi- neers, Washington, D.C., 2004. 3. Trip Generation, Volume 1 of 3: User’s Guide, Institute of Transporta- tion Engineers, Washington, D.C., 2008, p. 7. 4. Fundamental Research Statistics for the Behavioral Sciences, John T. Roscoe; Holt, Rinehart and Winston, Inc., 1969. C-16 To InternalFrom Land Use 1 Land Use 2 Land Use 3 Etc. External Land use 1 Number or % Land use 2 Number or % Land use 3 Number or % Etc. Number or % External Number or % Total 100% Table C-6. Sample summary format—inbound trips.