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.
14 In the context of statewide forecasting, rural trip-making and long-distance intercity travel constitute important market segments. Information describing these markets and how they vary from state to state has historically been sparse, and many states do not have the resources to initiate original data collection to develop a set of model parameters. Yet these same states have a pressing need for confidence in reasonable transportation planning results for rural and long- distance travel. Furthermore, for the states where local data are available, there is little basis to assess how comparable their assumptions are with those from other states. This chapter will address this topic by first identifying differences in rural and urban travel in various states from existing surveys. A high-level analysis of 1995 ATS, 2001 NHTS, 2009 NHTS, and select statewide, super-regional, and tourist survey data is provided in this chapter to highlight how differences in rural and long-distance trip-making occur in different geographic regions and to identify any explanatory variables that could be used to adjust average values and reflect conditions in a particular state. The most recent NHTS contains over 20 separate add-on partners, some representing full states and some MPO planning areas (which may include rural areas within the MPO boundary). In conducting analysis, it is important that rural and long-distance data on transferable parameters be compared against urban short-distance data and typical model parameters. For example, according to the 2009 NHTS, short trips account for the vast majority of personal trips in the United Statesâthree-quarters of vehicle trips are less than 10 miles in length. However, these trips account for less than one-third (28.9 percent) of all vehicle miles traveled (VMT). Trips of more than 100 miles account for less than 1 percent of all vehicle trips but 15.5 percent of all household-based vehicle miles, as illustrated in Figure 2.1. With the potential impact on VMT, travel demand forecasts depend on knowing more about the current amount and nature of long-distance and rural travel in the United States. This chapter first assesses national data sources on personal long-distance and rural travel, with a focus on the 1995 ATS, 2001 NHTS, and 2009 NHTS. Long-distance travel includes air, intercity bus and rail, and personal vehicle as the primary modes. There is an acknowledged lack of sufficient data on long-distance trips, and too little understanding of how travelers make decisions regarding mode, what kinds of reasons people travel long-distance, and other basic characteristics of intercity, interstate, and long-distance travel. Information about surface modes (private vehicle and transit) are particularly important, since for distances less than 500 miles, surface transportation modes move the majority of people. Next, this chapter describes available statewide and regional household travel surveys that include a significant sample of rural and/or long-distance trips. This includes a discussion of the Ohio Statewide and Long-Distance Travel Surveys, conducted in 2002â2003, which contacted 16,529 households, of which 2,049 made long-distance trips. This is followed by discussion of C h a p t e r 2 Long-Distance and Rural Area Data Sources
Long-Distance and rural area Data Sources 15 the Michigan Travel Counts Study, conducted in 2004 and 2009, with both Michigan efforts including a retrospective component focused on trips of 100 miles or greater and sampling areas for rural travel. Following these main studies, the chapter includes details about other possible sources of rural or long-distance travel data, from recent and ongoing statewide and global posi- tioning system (GPS) surveys to other superregional travel surveys and tourism surveys. 2.1 National Travel Surveys This section provides an overview of the 1995 ATS, 2001 NHTS, and 2009 NHTS. Although the NHTS Add-On components are largely equivalent to statewide surveys discussed in the next section, these are still part of the national survey and use the same survey instrument and sam- pling plans. American Travel Survey (ATS) Overview: The ATS was a national survey of long-distance trips defined as 100 miles or more, one-way. Although over 15 years old, the 1995 ATS remains the primary source of information at the national, state, and metropolitan-area level about the amount and characteristics of long- distance travel flows between states and large metro areas. Sample Detail: Sample selection for the ATS was based on households that had participated in the Current Population Survey (CPS) (http://www.bls.gov/cps/). The sample was based on Pri- mary Sampling Units (PSU) (Lapham, undated), as defined below, and a selection of addresses within each PSU. The sample was distributed rather evenly across the states (a choice that gen- erated some discussion) to ensure representation from each state. The sample for each state was designed to include two or more PSUs. All the PSUs were in urbanized areas, so no rural households are represented in the dataset. The person trip file contains 116,176 individuals who reported 556,026 long-distance trips during the survey year. âA trip is defined as each time a person goes to a place at least 100 miles away from home and returns.â PSUs are small geographic areas carefully selected to represent larger geographic areas. The PSUs were grouped into two strata; self-representing areas and nonself-representing areas. Self- representing areas generally consist of a single PSU used to represent an entire metropolitan area. The remaining areas, called nonself-representing, were formed by combining PSUs that possess similar characteristics, such as geographic region, population density, population growth rate, Figure 2.1. Vehicle trips and VMT by trip length. Source: Authorâs analysis of NHTS 2009. 0 10 20 30 40 50 60 1/2 mile or less 1/2- 1 mile 1.01-10 miles 10.01-20 miles 20.01-30 miles 30.01-50 miles 50.01-75 miles 75.01-100 miles more than 100 miles Percent Of Trips Percent Of Miles
16 Long-Distance and rural travel transferable parameters for Statewide travel Forecasting Models and proportion of nonwhite population, as stated in the ATS overview document. A sample of nonself-representing PSUs was selected to represent all of the PSUs in the stratum. A total of 729 PSUs were sampledâ314 self-representing and 415 nonself-representing. Survey Conduct: The households sampled in each of the PSUs were contacted four times, once each quarter, to report long-distance travel by the household members. If for some reason the household was not contacted during a quarter, when contact was next made information about the missing quarter was obtained. People who moved out or into the sample household were retained through recall and imputation. Since the sample was based on addresses, if new people moved into the household, the household remained part of the sample, and retrospective data about long-distance travel was collected from the new household members and used in imputa- tion and weighting. The study approach included use of a survey package mailed out to the household with a post card reminder. The retired CPS households that had telephone numbers on record were inter- viewed via computer-assisted telephone interview (CATI) while the rest were interviewed using computer-assisted personal interview (CAPI) and in-person visits (approximately 55 percent CATI and 45 percent CAPI). Limitations of the ATS: The biggest concern with using the ATS is obviously the age of the data, now more than 15 years old. In the intervening decade and a half, major changes have occurred in economics and demographics, communication technology, and security precautions at air- ports, just to name a few. In addition, the limitations of the survey to trips 100 miles or more one-way might impact assessing the full continuum of travel through the travel demand forecasting process. In the NHTS data series, 30 percent of long-distance trips were in a midrange distance, between 50 and 100 miles one-way, and these trips are underreported in the daily estimates of travel. Trips of this distance are important to many corridor analyses, but would be missing from the ATS. The lack of rural households (HHs) could be another limitation, especially if it is found elsewhere that rural HHs make more long-distance trips. Uses of the ATS: The ATS was designed to be useful for multistate and corridor planning and research. The large sample size and representation from each of the states means that these data can be used to estimate state-to-state flows and even some flows between large metropolitan areas if the resulting margin of error (up to 20 percent) can be tolerated, a unique characteristic of the ATS. (Note that the margin of error can be calculated at the state-level based on existing reports, but data for recalculating new margins of error are not available.) The long-range trips captured in the ATS have a significant non-auto mode share, another unique characteristic, and separate detail about recurring trips such as long commutes and week- end trips to second homes. In addition, intermodal connections are captured, allowing analysis of access modes to airports, intercity rail, and intercity bus stations. In 1995, FHWA also conducted one of the national household surveys (then called National Personal Travel Survey), which also had a long-distance component (measuring trips of 75 miles or more taken within a 2-week period). After the release of the 1995 ATS and the 1995 NPTS, there was quite a bit of research to see if the surveys could be combined. A number of similarities and dissimilarities were noted between the two data sources. For instance, the ATS was conducted using a panel, where the same householdâs reports for each quarter were used. On the other hand, the 1995 NPTS asked randomly selected households to report long-distance trips for just the 2 weeks prior to the assigned travel day. The short recall period, it was found, is more likely to miss infrequent travelers and perhaps overcount frequent
Long-Distance and rural area Data Sources 17 travelers. ATS calculated trip distance based on trip origin (zip code centroid), trip destination, and mode used using a national network of highways and air routes. In 1995, the long-distance component of the NPTS calculated distance on the great circle distance of MSA centroid to MSA centroid, a less precise measure than one that uses network distance. With these and other relevant differences in mind, the two surveys were combined into a single sample in the 2001 NHTS. 2001 National Household Travel Survey Overview: The 2001 NHTS combined elements of the 1995 ATS and the 1995 NPTS into a single survey. The survey was designed to obtain trips to destinations 50 miles or more from home within the 4 weeks previous to the assigned travel day. The long-distance component was included in the national sample of 26,000 households, which included 60,000 people. In addition, the Add- On states of New York, Texas, and Wisconsin purchased a long-distance component along with the daily Add-On. Sample Detail: The NHTS sample was a list-assisted random digit dialing (RDD) telephone sample, which means it consisted of randomly selected telephone numbers matched against known lists of commercial and nonworking numbers to maximize the âhitâ rate of a working residential number. Numbers were only selected from working 100 blocks where at least two other numbers in the 100 block were assigned. This sample frame was updated quarterly to include newly assigned numbers on a regular basis. The drawbacks of a telephone sample include the lower response rates (40 percent in NHTS 2001 compared to over 80 percent in 1995 ATS) and the lack of representation of very low- income households that might not have telephones. On the plus side, the geographic representa- tion of the sample is much more widely spread, including households in all areas of the country weighted to Census division. Survey Conduct: The NHTS sampled telephone numbers were preassigned a day of the week as a travel day, with the travel period being the 4 weeks previous to the travel day, so the travel period covers all days of the year. The selected household was sent a precontact letter with an incentive and then called and recruited into the survey. A packet of information was sent to the household with a second incentive. In this packet was a memory jogger for trip reporting (travel diary) and a map with the household location geocoded (from the address) with a 50-mile circle drawn around it and a reminder to respondents to keep track of any trips to places outside the circle. The data were collected via CATI. Limitations of the 2001 NHTS: The goal of the long-distance component of the 2001 NHTS was to provide national-level statistics on basic parameters of long-distance travel, such as fre- quency of trips by mode and purpose. Analysis of the recall period of 28 days prior to the travel day shows lower-than-expected trip reporting for air and other trips not in private vehicles, when compared to the ATS. For example, the 1995 ATS had 18.02 percent of trips reported by air, while the NHTS had only 9.23 percent mode share for air (of trips 100 miles or more). In addition, because of the short recall period, the unweighted sample size for 2001 NHTS long-distance is much smaller than the 1995 ATSâthe 4-week travel period collected 45,165 long-distance trips compared to 550,000 in one year of ATS. Although these 4-week-period trips were expanded to represent 1 year of travel, the small sample limits the power of the data at anything less than a national overview. Uses of the 2001 NHTS: The 2001 NHTS long-distance data has a couple of unique components that might be of interest. First, the data can be used to estimate the amount and type of midrange trips (less than 100 miles in length), a travel component that can be important to corridor planning.
18 Long-Distance and rural travel transferable parameters for Statewide travel Forecasting Models Within this mileage range, where airâs mode share is negligible, information on the amount and type of travel is useful. Second, the 2001 NHTS, because of the lowered mileage threshold, included information about recurring trips, such as long commutes, weekend trips to second homes, and medical trips. This is a unique source of data on the type and frequency of recurring long-distance trips. Comparison of Select 1995 ATS and 2001 NHTS Travel Statistics One of the key strategies to assess the data sources is to benchmark the estimated travel parameters against each other. Table 2.1 shows some of the trip-based estimates of travel, such as mode, purpose, and trip length, from the 1995 ATS and the 2001 NHTS. In an effort to maxi- mize consistency of results, a mileage threshold of 100 miles was used for analyzing both surveys, even though the NHTS generally used a threshold of 50 miles or more to define long-distance Parameter Summary 1995 ATS More Than 100 Miles 2001 NHTS More Than 100 Milesa Percent of Trips by Mode Private Vehicle 78.51 87.13 Air 18.02 9.23 Other 3.47 3.64 Percent of Trips by Purpose Business and Bus/Pleasure 22.42 25.69 Visit Friends/Relatives 32.58 26.31 Leisure 30.53 26.21 Personal/Family or Medical 11.93 9.56 Other 2.54 12.22 Overall Mean Trip Length in Miles (One-Way All Modes)b 411.88 457.57 Mean Trip Length â Air 1,003.21 2,088.78c Mean Trip Length â Private Vehicle 276.53 301.54 Mean Trip Length â All Other 404.02 482.02 Mean Trip Length by Purpose in Miles (One-Way All Modes) Business and Bus/Pleasure 467.89 480.93 Visit Friends/Relatives 398.77 478.60 Leisure 406.70 516.44 Personal/Family or Medical 376.05 409.80 Other 316.03 276.28 Overall Travel Party Size (All Modes) 3.10 N/A Travel Party Size â Air 2.98 N/A Travel Party Size â Private Vehicle 2.42 N/A Travel Party Size â All Other 9.34 N/A Travel Party Size by Purpose Business and Bus/Pleasure 2.12 N/A Visit Friends/Relatives 2.81 N/A Leisure 3.93 N/A Personal/Family or Medical 2.91 N/A Other 6.34 N/A a NHTS 2001 includes trips of 50 miles and more. For this analysis only trips of 100 miles and longer one-way were included. b 1995 ATS âRound-Trip Distanceâ was divided in half to provide one-way estimates. c NHTS Trip Distance includes extreme values. Trip length was capped at the 99th percentile (5,252.18 miles). Table 2.1. Preliminary comparative statistics from ATS and NHTS.
Long-Distance and rural area Data Sources 19 trips. Analysis of these parameters from the 1995 ATS and 2001 NHTS datasets results in similar estimates at the national level. Primary differences appear in the mode split, where the propor- tion of air travel is very different, and the mean trip length for air, potentially impacted by 9/11. Travel party size was included in this table, although this statistic could only be calculated for the ATS. For the auto mode, this measurement would essentially represent an auto occupancy rate. Further analysisâat a smaller geography, for household-based estimates, or to estimate travel for different demographic groupsâmay reveal greater dissimilarities between the ATS and NHTS. The two surveys were also investigated for estimating trip generation rates. A key issue to con- sider in this regard is the timeframe for consideration. Since few households make over 100-mile trips on a daily basis, it was decided to look at such trips on an annual basis instead. Using this timeframe, the ATS indicates aggregate trip rates of 7.00 trips per household and 4.79 trips per person (18 years of age and older). More complete listings of potentially transferable parameters are provided in subsequent chapters of this report. 2009 National Household Travel Survey One of the important characteristics for rural travel parameters is transferability across geog- raphies. The 2009 NHTS is an interesting source of data for rural travel parameters because it includes rural samples from all over the nation. (The NHTS uses the Census definition of ârural,â that is any census block or block group outside of an urbanized area or urban cluster; on the 2009 NHTS the Census 2000 delineation is coded.) This makes it possible to compare household- based travel parameters between urban and rural households, which could start to inform the understanding about geographic differences in travel for the same demographic categories of households or persons. Urban household travel behavior can differ between different areas of the country (e.g., between households in the Northeast and households in the West). One can expect that rural household travel might likewise not be homogeneous, but would vary across locations as well as between socioeconomic classes. Table 2.2 shows the sample distribution of all rural households in the 2009 NHTS. A total of 43,583 households in rural areas completed the survey. Figure 2.2 shows a map of the Census divisions and their constituent states. The distribution is uneven across Census divisions because of the participation of the Add- Ons. The South Atlantic division has so many samples because nearly every state in the division included an Add-On (Florida, Georgia, South Carolina, North Carolina, and Virginia) and all had large samples that included rural areas. The other Census divisions are represented by only Item Rural Samplesa All Rural (National) 43,583 New England 1,560 Mid-Atlantic 5,721 East North Central 2,355 West North Central 2,684 South Atlantic 19,293 East South Central 1,570 West South Central 6,228 Mountain 1,727 Pacific 2,445 a Includes Add-On samples. Table 2.2. NHTS 2009 sample of rural households.
20 Long-Distance and rural travel transferable parameters for Statewide travel Forecasting Models one or two Add-Ons, such as West North Central, which includes the Add-Ons in South Dakota and Iowa, or East North Central, which includes the Add-Ons in Wisconsin and Indiana. Other divisions, such as Pacific, have only one Add-On, in this case the large sample in California. Table 2.3 and Figure 2.3 compare a few important travel estimates from the national sample for urban and rural households and the same estimates for each Census division. For the pur- poses of this analysis, rural trips were those that begin and/or end in rural areas. Figure 2.2. Census regions and divisions. Source: http://www.census.gov/geo/www/us_regdiv.pdf Person Trips per Person Average Vehicle Trip Length (Miles) VMT per Household VMT per Person Urban Rural Urban Rural Urban Rural Urban Rural All 3.8 3.6 8.0 12.0 43.5 72.1 17.6 27.4 New England 3.8 3.9 9.0 11.7 47.7 79.5 19.9 29.8 Mid-Atlantic 3.8 3.7 7.7 11.6 35.6 70.9 14.3 26.9 East North Central 4.0 3.6 7.7 11.8 43.2 75.9 18.3 28.6 West North Central 4.1 3.6 8.2 10.6 48.3 63.2 21.5 25.3 South Atlantic 3.7 3.6 8.3 12.6 44.4 72.0 18.5 27.8 East South Central 3.8 3.4 8.7 13.3 46.7 75.0 20.7 29.1 West South Central 3.8 3.7 8.2 12.3 47.0 72.6 18.6 26.3 Mountain 4.0 3.8 7.6 12.0 46.0 76.6 18.3 28.5 Pacific 3.8 3.7 7.4 10.6 42.1 64.6 15.6 24.1 Source: Authorâs analysis of 2009 NHTS, includes travel on weekends and holidays. Table 2.3. Travel parameters for urban and rural households by Census division, 2009 NHTS.
Long-Distance and rural area Data Sources 21 As Table 2.3 shows, the number of person trips per person does not vary much between urban and rural households. Overall, people in urban households make on average 3.8 trips per day, compared to people in rural households who make 3.6. These per capita estimates include people who reported travel and people who did not, by all modes and for all purposes, and across all days, including weekends and holidays. In developing final travel parameters for rural travel demand estimation and forecasting, weekday-only estimates will be preferable. However, estimates of vehicle travel, such as vehicle trip length and VMT per household, vary quite a bit more between urban and rural areas. The average vehicle trip length for urban house- holds is 8 miles compared to 12 for rural households (50 percent farther). As a result, overall vehicle miles of travel per capita in urban areas is 17.6 miles on average compared to 27.4 for people in rural areas (including people ages 5 and older). There are differences in these major estimates between regions of the country for both urban and rural households. For instance, VMT per person is lowest for households in the Pacific divi- sion rural areas (24.1 miles per day), and almost 25 percent higher per household for rural areas in New England (29.8 miles per day). This analysis is preliminary, and the apparent differences in travel demand across geography may disappear when normalized for household size, number of vehicles, and other traditional cross-classifications, or they may not. Although each state Add-On had a separate sampling scheme, they all included rural households, either as a distinct sample strata (such as in Florida) or as part of a population- proportionate sample (such as Texas and California). Figure 2.4 shows a map of the state Add-Ons to the 2009 NHTS. Table 2.4 shows the number of rural households sampled and weighted for each of the state Add-Ons. Remember that the sum of all rural households in the 2009 NHTS is 43,583 of which 39,739 are from Add-On states and the remainder are part of the national sample in non-Add-On states. Table 2.5 shows the same comparative travel characteristics (VMT per household, VMT per person, average vehicle trip length and person trips per person) for urban and rural households in the 15 Add-On states. As with the national and Census division comparisons, large differences were not observed across estimates of the person trips per household between rural and urban or between state Add-Onsâin fact, they all round to four, except for rural areas in Wisconsin. But vehicle travel estimates are very different. For example, the average vehicle trip length in rural areas in Georgia is 14 miles, the highest of all rural estimates of vehicle trip length, compared to its neighbors of Florida and South Carolina, with averages of 11.7 and 11.8, respectively. Note that all three of these states are in the same Census division of South Atlantic. Figure 2.3. VMT per person for urban and rural households by Census division. Source: Authorâs analysis of 2009 NHTS, includes travel on weekends and holidays. 0 5 10 15 20 25 30 35 All New England Mid Atlantic East North Central West North Central South Atlantic East South Central West South Central Mountain Pacific Urban Rural Vehicle Miles (VMT) per day
22 Long-Distance and rural travel transferable parameters for Statewide travel Forecasting Models Figure 2.4. 2009 NHTS state Add-Ons. California Texas Florida South Carolina New York Vermont North Carolina Virginia Tennessee Indiana WisconsinSouth Dakota Iowa Georgia Arizona State Sample Size (Rural Households) Weighted Households Weighted Persons Arizona 1,163 422,238 1,248,929 California 2,148 958,264 2,718,128 Florida 3,251 1,057,714 2,771,537 Georgia 3,569 1,136,647 3,016,386 Iowa 1,359 509,618 1,276,488 Indiana 1,133 674,247 1,796,446 North Carolina 4,664 1,478,175 3,742,939 New York 5,452 1,100,822 2,911,999 South Carolina 1,974 671,261 1,731,758 South Dakota 654 162,405 387,506 Tennessee 1,114 897,829 2,378,480 Texas 5,902 1,610,731 4,572,820 Virginia 5,551 838,240 2,147,874 Vermont 1,137 153,587 382,281 Wisconsin 668 753,627 1,884,870 Sum of All Add-Ons 39,739 12,425,406 32,968,440 Table 2.4. Sampled and weighted number of rural households in the 2009 NHTS Add-On states.
Long-Distance and rural area Data Sources 23 Analysis of the Florida 2009 NHTS Add-On surveys, conducted for the Florida DOT (Florida Department of Transportation and Cambridge Systematics, Inc., 2010), indicated a significant difference in average trip length when comparing mean values for urban and rural households. Preliminary analysis of the 2009 Florida Add-On found the average trip length for home-based work (HBW) trips to be 25.7 minutes in urban settings and 29.6 minutes for rural households. Average trip lengths provided elsewhere in this report from the 2009 NHTS national sample were only provided in miles traveled for all trip purposes. 2.2 Statewide Household Travel Surveys The Ohio and Michigan Departments of Transportation (DOTs) have made significant invest- ments in the documentation of demographic and travel behavior characteristics of those resid- ing in the state. This has included statewide household travel surveys with a sizeable sample of rural residents. The Ohio and Michigan DOTs also conducted long-distance household travel surveys, with some overlap in households that completed both surveys. Recent and ongoing statewide household surveys for California, Oregon, and Utah are described in a later section of this chapter due to their recent nature. Ohio Statewide Household Travel Survey The Ohio Statewide Household Travel Survey was conducted from 2001 to 2003. Sponsored by the Ohio DOT, the purpose of this survey was to document statewide and regional travel patterns in order to update travel demand models. (Unless otherwise noted, the details in this section are from NuStats, Technical Memorandum: 2002â2003 Ohio Statewide Household Travel Survey, prepared for the Ohio Department of Transportation, August 2004.) A total of Person Trips per Person Average Vehicle Trip Length (Miles) VMT per Household VMT per Person Urban Rural Urban Rural Urban Rural Urban Rural Arizona 3.9 3.8 8.5 11.0 49.2 78.4 19.2 26.5 California 3.8 3.5 7.6 11.7 44.0 71.0 15.7 25.0 Florida 3.6 3.7 7.7 11.7 41.0 68.6 17.1 26.2 Georgia 3.7 3.5 8.8 14.0 48.8 83.0 19.2 31.3 Iowa 4.1 3.7 6.5 9.7 34.8 57.2 16.1 22.9 Indiana 3.7 3.6 7.2 11.1 38.2 70.8 16.7 26.6 North Carolina 3.8 3.6 8.6 11.3 45.6 63.9 20.0 25.2 New York 3.7 3.5 6.3 10.5 24.0 60.3 9.4 22.8 South Carolina 3.9 3.6 8.2 11.8 47.1 69.3 19.8 26.9 South Dakota 4.1 3.8 7.1 10.8 43.6 58.2 19.2 24.4 Tennessee 3.7 3.5 9.1 11.7 49.2 70.3 22.1 26.5 Texas 3.8 3.5 8.5 12.1 48.9 74.1 18.8 26.1 Virginia 3.8 3.5 8.0 12.4 43.5 72.5 18.1 28.3 Vermont 3.9 3.6 7.3 10.4 38.2 59.3 17.8 23.8 Wisconsin 4.0 3.3 7.8 11.9 42.7 62.5 18.9 25.0 Average 3.83 3.58 7.81 11.47 42.59 67.96 17.87 25.83 Source: Authorâs analysis of NHTS 2009, includes travel on weekends and holidays. Table 2.5. Travel parameters for urban and rural households by Add-On state, 2009 NHTS.
24 Long-Distance and rural travel transferable parameters for Statewide travel Forecasting Models 16,112 households were surveyed, of which 2,530 households are located in rural Ohio, defined as outside an MPO area. The survey was conducted using state-of-the-practice methods, includ- ing recruiting households to participate, mailing travel diaries to them, and retrieving the travel details by phone or mail. All members of participating households reported travel for either a 24- or 48-hour period. Travel days covered all seven days of the week, and travel days were assigned throughout the fall and spring periods, except for weekdays immediately preceding or following a holiday. The survey documented the following data elements: ⢠Household DemographicsâHousehold size, household vehicles, dwelling type, home own- ership, home location, and income; ⢠Person DemographicsâAge, gender, relationship, licensed driver status, employment status, work location and other work-related details, student status, and school type and location; ⢠Travel Behavior CharacteristicsâActivity, travel mode, origin and destination of travel (geo- coded to x-y coordinates), trip purpose, travel party size and composition, and arrival and departure times. The following is a summary of relevant weighted travel statistics obtained overall and for the rural households, as documented in the final report for the survey: ⢠The overall average daily trip rates were 7.90 trips per household and 4.94 trips per person. For the rural households, these statistics were 7.78 trips per household and 4.79 trips per person. ⢠At the statewide level, trips averaged 18.4 minutes in length. Work trips averaged 20.9 minutes and shopping trips averaged 15.0 minutes. In rural areas, trips averaged 18.8 minutes summed for all purposes. Work trips averaged 21.0 minutes and those for shopping averaged 16.2 min- utes for rural areas. ⢠Ninety percent of all trips statewide were made by auto (driver or passenger). For the rural households, auto accounted for 89 percent of all trips. Figure 2.5 depicts the distribution of travel by modes statewide and in rural areas. Ohio Statewide Long-Distance Travel Survey In 2002, the Ohio DOT embarked on designing and conducting a long-distance travel survey as a supplement to its statewide household travel survey. (Unless otherwise noted, the details in Figure 2.5. Travel modes from Ohio Statewide Household Survey. Source: NuStats Technical Memo, Ohio Statewide Household Survey, Table 7.7. 68% 23% 1% 1% 5% 3% 1% 67% 23% 1% 0% 5% 4% 1% 0% 10% 20% 30% 40% 50% 60% 70% 80% Auto-D Auto-P Car/Vanpool Transit School Bus Walk Other Statewide Rural
Long-Distance and rural area Data Sources 25 this section are from NuStats, Technical Memorandum: 2002â2003 Ohio Long-Distance Travel Survey, prepared for the Ohio Department of Transportation, March 2005.) Over the 2-year study period, long-distance travel data were obtained from more than 8,000 households across the state, of which 6,113 were summarized in the report. For this study, a long-distance trip was defined as a nonwork trip of 40 miles or longer. (In Appendix E of the NuStats report, the CATI survey question LDQ1 asks about travel 50 miles or more away from home.) The long-distance survey asked respondents to list any trips over 40 miles, even though over 50 miles is the long- distance threshold defined in the model. The lower 40-mile threshold was used in the survey real- izing that respondents might leave out valid over 50 trips with a 50-mile survey cut-off. Reported trips between 40 and 50 miles were later excluded during model development. There were three approaches used in collecting the data: a 2-week retrospective survey admin- istered to households that were participating in the statewide household travel survey; a 2-week retrospective survey administered to randomly sampled households not participating in the state- wide household travel survey; and a 4-week prospective survey of nonhousehold travel survey house- holds that were prescreened and identified as having a probability of making a long-distance trip. Long-distance travel was documented for all household members, regardless of age for the specified time period. Trips were recorded regardless of the day of week (i.e., weekday or weekend). However, trips were not recorded during the summer or during a week where a holiday was observed. The survey documented the following data elements: ⢠Household DemographicsâHousehold size, household vehicles, dwelling type, home own- ership, home location, and income; ⢠Person DemographicsâAge, relationship, employment status, work location, worker occu- pation and industry, student status, and school type and location; and ⢠Travel Behavior CharacteristicsâNumber of noncommute long-distance trips, destination of travel, trip purpose, mode of travel, travel party size, and intermediate stops. The resulting Technical Memorandum summarized key statistics obtained from this survey for households completing the 2-week retrospective surveys (both those linked to the statewide household travel survey as well as those that were not). (The Ohio DOT has provided access to the full dataset for use in this analysis.) The overall trip rate was 6.44 for nonwork long- distance household trips per 2-week nonholiday and nonsummer period. More than one-third (37 percent) of the 2.5 million Ohio households reported some level of long-distance trips. As shown in Figure 2.6, households from Akron, Dayton, and Springfield were more likely to Figure 2.6. Geographic distribution of households reporting long-distance travel from Ohio Long-Distance Travel Survey over 2 weeks. Source: NuStats Technical Memo, Ohio Statewide Long-Distance Travel Survey, Table F.2. 6% 8% 15% 13% 15% 9% 10% 9% 10% 6% 0% 2% 4% 6% 8% 10% 12% 14% 16% Toledo (TMACOG) Lima (LACRPC) Dayton (MV)RPC Springfield (CCSTS) Akron (AMATS) Canton (SCATS) Mansfield (RCRPC) Steubenville (BHJTS) Youngstown (EDATS) Rural
26 Long-Distance and rural travel transferable parameters for Statewide travel Forecasting Models report long-distance travel as compared to those from other parts of the state. This is due to their proximity to a larger urbanized area. The majority of households (63 percent) reported no long-distance trip-making during the 2-week reference period, as indicated in Figure 2.7. The vast majority of long-distance trips (92 percent) were made by auto, either driver or passenger. As depicted in Figure 2.8, travel by air was a distant second at 5 percent of long-distance trips reported. An important feature of the Ohio survey is that the prescreened prospective sample was only used as a way to establish detailed characteristics of long-distance trips for micro-simulation while the random retrospective sampling established tour rates. This survey design allowed for obtaining adequate samples for micro-simulation even though the prospective sample is obvi- ously biased to households likely to make such trips. Michigan Statewide Household Travel Surveys In 2004, Michigan DOT undertook an extensive survey of 14,280 Michigan households (http://www.michigan.gov/documents/MDOT_TravCharTR_Final20060804_167340_7.pdf). Figure 2.7. Number of long-distance travel trips reported per household from Ohio Long-Distance Travel Survey over 2 weeks. Source: NuStats Technical Memo, Ohio Statewide Long-Distance Travel Survey, Table F.28. 63% 5% 12% 2% 9% 1% 3% 0% 3% 0% 3% 0% 10% 20% 30% 40% 50% 60% 70% 0 1 2 3 4 5 6 7 8 9 10+ Figure 2.8. Long-distance trip travel modes from Ohio Long-Distance Travel Survey. Source: NuStats Technical Memo, Ohio Statewide Long-Distance Travel Survey, Table 7.29. 52% 40% 1% 5% 1% Auto or van or truck driver Auto or van or truck passenger Bus (public transit) Commercial airplane Other, specify
Long-Distance and rural area Data Sources 27 The survey documented basic demographics and 48 hours of weekday travel information for all household members regardless of age. For sampling purposes, the state was divided into seven geographic areas, including three rural areas: southern lower peninsula rural (2,059 households), northern lower peninsula rural (2,073 households), and upper peninsula rural (2,027 house- holds). A second round of household surveys was conducted in 2009. The survey documented the following data elements (http://www.michigan.gov/documents/ MDOT_travelcounts_results_Appendices_20_to_24_142289_7.pdf.): ⢠Household DemographicsâHousehold size, household vehicles, household workers, home location, and income; ⢠Person DemographicsâGender, age, relationship, driverâs license status, transit pass status (and if one is held, what type and cost), educational attainment, student status, school type and location, employment status, industry, and work location; and ⢠Travel Behavior CharacteristicsâOrigin and destination of travel, activities at origin and destination, travel mode(s), travel party (if travel by auto), and parking cost. The following is a summary of relevant travel statistics obtained statewide and for the rural households, as readily available in the final report for the survey. ⢠The overall average weekday 48-hour person trip rates were 17.3 trips per household and 6.9 trips per person. In the southern lower peninsula rural area, the 48-hour statistics were 17.1 trips per household and 6.7 trips per person. In the northern lower peninsula rural area, the statistics were 15.5 trips per household and 6.3 trips per person. For households in the upper peninsula rural area, the average 48-hour trip rates were 16.0 trips per household and 6.7 trips per person. ⢠Average daily household trip rates were estimated at 8.70 trips per household, statewide. In contrast, the average daily trip rates for the three rural areas were 8.54 trips per household in the southern lower peninsula rural area, 7.75 trips per household in the northern lower peninsula rural area, and 8.00 trips per household in the upper peninsula rural area. Michigan Statewide Long-Distance Travel Surveys Both the 2004 and 2009 MDOT âTravel Countsâ statewide travel survey efforts included a long-distance travel survey component, in which long-distance trips were defined as any trip of more than 100 miles away from home that occurred in the 3 months prior to the survey. In terms of the data collection approach, the long-distance supplemental questions were printed directly into the travel log, with the instructions to âplease provide the following infor- mation for ALL trips you took within the last 3 months, greater than 100 miles (one-way) from homeâ (http://www.michigan.gov/documents/mdot/mdot_MTC_II_appendices_333723_7. pdf). By including the questions in the travel log, long-distance travel was requested for all household members, regardless of age for the specified time period. In the first effort, travel diary data were recorded from April 4 to June 10, 2004, then again from September through December 2004 (http://www.michigan.gov/documents/MDOT_travelcounts_results_Final_ Report_142283_7.pdf). In the second effort, travel was recorded September through December 2009, so long-distance data reflects late summer and fall of 2009. In both efforts, the survey documented the following data elements: ⢠Household Demographics (through the household travel survey component)âHousehold size, household vehicles, household workers, home location, and income; ⢠Person Demographics (through the household travel survey component)âGender, age, rela- tionship, driverâs license status, transit pass status (and if one is held, what type and cost), educational attainment, student status, school type and location, employment status, indus- try, and work location;
28 Long-Distance and rural travel transferable parameters for Statewide travel Forecasting Models ⢠Travel Behavior CharacteristicsâOrigin and destination of travel, activities at origin and destination, travel mode(s), travel party (if travel by auto), and parking cost; and ⢠Long-Distance Travel Behavior CharacteristicsâDestination (city and state or country if international), day of departure and day of return, primary reason for trip, primary mode to destination and transportation used at destination, and the number of times this specific trip was made during the 3-month period and the past 12 months. A total of 14,280 households from across the state were surveyed in the 2004 effort, with each member of the household completing a 24-hour travel diary. Both diaries included the long- distance 3-month retrospective log at the front of the diary. Table 2.6 summarizes the number of surveys, reported trips, and estimated long-distance trip rates resulting from each survey effort. The documentation of long-distance travel is more extensive in the 2004 survey compared to what is available for the 2009 survey. Figure 2.9 presents 2004 survey findings with respect to long-distance travel. In terms of trip purpose, the majority of long-distance trips were reported for pleasure (63 percent), followed by trips for personal business (18 percent) and business pur- poses (15 percent). The majority of long-distance trips were made by private auto (87 percent), as depicted in Figure 2.10. Air travel accounted for 10 percent of the trips, while bus and train accounted for very small proportions of the long-distance travel. (Note that for business trips, airplane travel accounts for 23 percent of all trips.) 2004 Survey 2009 Survey Number of Households 14,818a 1,975 Number of Long-Distance Trips 37,338 4,567 Household Trip Rate (3-Month Period) 2.18 2.31 Household Annual Trip Rate 7.34 6.25 Source: Abt SRBI Comparison Report, Michigan Travel Counts Survey, Table 5.16. a The difference in number of households in the 2004 survey reflects initial results (14,818) versus results after additional data cleaning (14,280). Table 2.6. Michigan travel counts long-distance trip rates. Figure 2.9. Michigan travel counts, long-distance trip purpose. Source: Michigan DOT Travel Characteristics Technical Report, Figure 45. 14.8% 62.9% 17.7% 2.8% 1.8% Business Pleasure Personal Business School/Church Other
Long-Distance and rural area Data Sources 29 Other pertinent findings from the 2004 Michigan long-distance surveys include ⢠The majority of long-distance trips (60 percent) are to destinations within the state, and ⢠In rural areas, most long-distance trips are for personal business. According to the Michigan DOT, this âindicates that the services available from trade centers within 100 miles of these locations lack amenities availableâ in the larger cities. Although trip distances were not provided in the summary data, the Michigan DOT did include a map summarizing the proportion of long-distance trips to different areas of the United States. This map, see Figure 2.11, shows that most trips are to surrounding areas in state and in neighboring states, with the exception of Florida attracting a large number of Michigan âsnowbirdsâ and tourists. 2.3 Supplemental Sources of Rural and Long-Distance Data In addition to the surveys discussed above, there are other survey sources that could be useful in developing the rural and long-distance transferable parameters. These sources are considered secondary because they are (1) recently completed and without complete documentation, (2) currently being fielded and thus unavailable, or (3) require data mining to obtain useful results for this study. These include ⢠Front Range Travel Counts Long-Distance Survey (Colorado), ⢠California Statewide Travel Survey, ⢠Oregon Household and Activity Survey, and ⢠Tourism surveys. For these secondary travel surveys, the travel survey team either recently completed data collection (so the data would be available to inform this study but a report with results has not yet been issued), collected relevant data as a naturally occurring event during the travel diary period but did not analyze it separately (so the data would be available to inform this study but will require some mining efforts), or knows currently travel surveys are being conducted that document the needed data (so the data would be available at some point to inform future studies). Similarly, tourism surveys are usually conducted for economic development purposes, but contain some O/D detail. They are summarized here because the data may be useful in estimating the relative attractiveness of certain areas over others for long-distance travel in particular. Figure 2.10. Michigan travel counts, long-distance travel mode. Source: Michigan DOT Travel Characteristics Technical Report, Figure 46. 86.6% 10.2% 1.1% 0.4% 1.7% Private vehicle Airplane Bus Train Other
30 Long-Distance and rural travel transferable parameters for Statewide travel Forecasting Models Front Range Travel Counts Long-Distance Survey The Front Range Travel Counts Survey (FRTC) is a comprehensive study of the demographic and travel behavior characteristics of Colorado Front Range residents across four MPO regions: North Front Range (Fort Collins), Denver, Colorado Springs, and Pueblo, as depicted in Figure 2.12. The objective of the survey effort is to document demographic and travel behavior characteristics across the Front Range in order to support travel demand modeling and answer travel-related questions. As part of the FRTC, a household travel survey was conducted in all four MPO regions. In addition, surveys were conducted to obtain data regarding commercial travel, travel into and out of the region, and residential travel that was more than 50 miles in distance. Data collection for the effort was conducted from 2009 to 2011. At the time of this report, the project was in the post-collection data processing stage. As indicated in Figure 2.12, the study area includes metropolitan Denver, but also the smaller urban and rural communities along the I-25 corridor. A total of 12,415 regional households com- pleted the 24-hour diary study. When analyzed, the geographic diversity will help to document differences in travel for rural versus nonrural households. In addition, within the household travel survey design, respondents were instructed to report all trips within the study area. Of more than 115,000 trips in the cleaning stage, approximately 889 trips are preliminarily flagged as being 50 miles or longer one-way. Of these 889 long-distance trips, one-fourth reflect a work commute. Given the expectation that long-distance travel would be a relatively rare event, a long-distance survey was added as a supplement to the household travel survey in order to document residen- tial travel more than 50 miles in distance. This supplemental survey was conducted in two stages. First, for all households where data collection had been completed and the household indicated an interest in participating in follow-up studies, the household was mailed a long-distance log and asked to record all travel 50 miles or longer for noncommute purposes made by the household Figure 2.11. Michigan travel counts, long-distance trip destinations. Source: Michigan DOT Travel Characteristics Technical Report, Figure 44.
Long-Distance and rural area Data Sources 31 members in the 2 weeks prior to a reference date (logic being that the commute trips would be captured in the 24-hour travel logs while all other long-distance trips were less regular and thus might not be captured in the 24-hour travel logs). Second, for households in the data collection stage of the survey, the long-distance log was provided in the same packet as the 24-hour travel diary and households were asked to record all travel 50 miles or longer that was not a work com- mute trip made by household members in the 2 weeks prior to the travel date. For this second group of households, if no long-distance travel was reported, the household was asked for the date and location of their last trip that met the definition of a long-distance trip. One-fourth (25 percent) of all households reported a long-distance trip, with a total of 8,680 trips in post-collection processing. This suggests that the unweighted trip rate is approximately 2.77 long- distance trips per household per 2-week period. These data have not been processed for summary statistics yet, weighted, or expanded. Comparisons of long-distance trip-making among different household travel surveys in this report are preliminary and reflect different thresholds for identify- ing long-distance travel. California Statewide Household Travel Survey The California DOT (Caltrans) is currently conducting a statewide household travel survey (http://www.scag.ca.gov/modeling/mtf/presentations/012611/mtf012611chts.pdf). The pro- jected sample size is 60,000 households, with most households completing a 24-hour weekday diary but some completing a 48-hour diary to capture weekend travel. The design includes a purposeful oversample of rural areas and a supplemental long-distance survey. When completed in 2012, this will provide a rich source of information for both rural and long-distance research. The last statewide household travel survey in California was conducted in 2000/2001 (http:// www.dot.ca.gov/hq/tsip/tab/documents/travelsurveys/2000_Household_Survey.pdf). A total of Figure 2.12. Front Range study area. Source: NuStats, 2011.
32 Long-Distance and rural travel transferable parameters for Statewide travel Forecasting Models 17,040 households participated in the survey, documenting demographic and travel behavior characteristics in a 24-hour diary study, conducted using state-of-the-practice methods. Data elements included household size, income, vehicle ownership, employment status of each house- hold member, and housing unit type as well as trip times, mode, activity at location, origin and destination, and vehicle occupancy. As with the Front Range effort, rural areas were surveyed and long-distance travel was obtained as it naturally occurred during the travel period. Thus, with data mining, this older survey may provide insights to aid in the estimation of rural and long-distance travel parameters. Oregon Household Activity Survey The Oregon Household Activity Survey is a statewide effort to document demographic and travel behavior characteristics of its residents (http://www.oregon.gov/ODOT/TD/TPAU/Survey.shtml). Launched in 2009, the survey will provide insights gleaned from over 17,000 households that are providing 24-hour travel details for all household members. In terms of documenting rural travel, the sampling plan was designed to cover the entire state, especially the rural areas outside the met- ropolitan areas in the state as shown in Figure 2.13. In addition, this effort instructs respondents to report all travel within the State of Oregon, so intercity travel can be captured. The Oregon DOT is currently considering a long-distance supplemental survey. Recent Global Positioning System (GPS) Surveys Several recent GPS-based household surveys have been analyzed for results relevant to rural and long-distance travel. Similarities in designs across these studies influence some of the results (e.g., weekend versus weekday proportions). In addition, the studies purposefully assigned the GPS units to long-distance commuters. The GPS-based survey statistics compiled for this analysis include the following, along with the breakdown of long-distance, rural, and urban trips for each survey: Figure 2.13. Map of Oregon. Source: http://sites.nustats.com/otas/
Long-Distance and rural area Data Sources 33 ⢠Atlanta Household Travel Surveys (2011): â 580 long-distance trips (1%); â 16,932 rural trips (26%); and â 48,098 urban trips (73%). ⢠Denver Household Travel Surveys (2010): â 395 long-distance trips (1%); â 9,836 rural trips (24%); and â 31,377 urban trips (75%). ⢠Massachusetts Household Travel Surveys (2010): â 176 long-distance trips (1%); â 3,349 rural trips (24%); and â 10,325 urban trips (75%). ⢠Chicago Household Travel Surveys (2008): â 102 long-distance trips (1%); â 1,572 rural trips (10%); and â 14,032 urban trips (89%). Additional information on these GPS surveys can be obtained at the following web addresses: ⢠Atlanta Regional Commission/Atlanta Household and Activity Travel Survey/2011 (http:// www.atlantaregional.com/transportation/travel-demand-model/household-travel-survey). ⢠Denver Regional Council of Governments/Front Range Travel Survey/2010 (http://www. drcog.org/index.cfm?page=RegionalTravelBehaviorInventory). ⢠Massachusetts Department of Transportation/Massachusetts Statewide Household Travel Survey/2010 (http://sites.nustats.com/travelsurvey/faq). ⢠Chicago Metropolitan Agency for Planning/Travel Tracker household travel survey for Chicago/ 2008 (http://www.cmap.illinois.gov/travel-tracker-survey). Tourism Surveys Surveys conducted to track tourism visits are another source of data regarding long-distance travel. These surveys are conducted at state, regional, and location-specific levels, conducted pri- marily to inform economic development efforts and fine-tune marketing campaigns. From a travel behavior point of view, these surveys do not document the key variables necessary to inform the development of parameters or for use in a model. However, they provide qualitative details that can suggest attraction rates for existing tourist areas, thereby improving the fit of a long-distance model. Statewide Tourism Surveys Tourism can play a large role in some statesâ economic budgets. Some states conduct surveys of visitors to improve touristâs needs. Examples of the data available from state surveys that might help with forecasting long-distance travel are shown in Table 2.7. The transferability of such surveys, usually collected in high tourism states, should probably be limited to other states with a large tourist component to their economy. Regional Tourism Surveys Some regions conduct surveys of visitors to learn what motivates tourists to visit their attrac- tions. Regional survey findings vary by location. Specific locations obtain different survey results. Each regionâs results will determine the type of variables needed to predict future visitor num- bers. Results from two regional surveys are shown in Table 2.8. National and State Park Surveys Many national and state parks collect visitor information from patrons using surveys. Their surveys are collected routinely with an abundance of data. A majority of the information collected varies by park location. Four park survey results are shown in Table 2.9.
Characteristic State of Oregon State of Hawaii State of Florida Data Collection Methods Online survey Combined surveys and statistics from other sources Combined surveys and statistics from other sources, including international Universe (or Sample Size) 512 (sample) 6,517,054 visitors 80,879,000 visitors Data Collection Duration April 29 to May 17, 2005 2009 2009 Origins California (14%); Oregon (14%); Washington (6%); Florida (4%); Arizona (4%); Pennsylvania (4%) Continental U.S., Japan, Canada, Europe, Other; U.S. West â 2,718,818 (by air); U.S. East â 1,561,468 (by air); 96,606 (by cruise ship) Canada 2,644,000 (3%); Domestic 71,246,000 (88%); Overseas 6,989,000 (9%) Group Size N/A Average 2.1 Average 2.3a Age N/A N/A 46a Travel Mode N/A 6,517,054 visitors arrived by air; 96,606 visitors by cruise ship 41,509,000 (51.3%) air; 39,370,000 (48.7%) non-air a Domestic trips only (auto and air). Table 2.7. Statewide tourism surveys. Characteristic Yuma Area, Arizona Cape Cod, Massachusetts Data Collection Methods Paper Surveys Online Survey Sample Size 1,316 N/A Data Collection Duration January 2010 to December 2010 2006 Origins One-fourth from Arizona (26.9%), (25.2%) came from California. After California, the other top 10 states for domestic visitors were Washington, Texas, Oregon, Michigan, Illinois, Florida, Nevada, Maryland and Virginia. Within Arizona, Greater Phoenix area accounted for 48.1%, 25% were from Pima County. Among foreign visitors, Canadians accounted for 70.3%, United Kingdom provided 11.9% of visitors. Average: Massachusetts (42%), Connecticut (16%), and New York (10%) Group Size Average 2.6; only 9.2% had children under 18 Spring 2006 â 49% traveled as a couple; summer 2006 â 31% traveled as a family; fall 2006 â 63% traveled as a couple Age Average 51 years old Average 31-45 years old Travel Mode Three-fourths of visitors to the Yuma area (75.5%) traveled in an automobile â 57.6% in a private car and 17.9% in a rental car; in addition, 13.2% traveled in a RV/camper, 7.4% by airplane, and 2.3% by motorcycle N/A Stay Duration Day visitors spent an average of 3.6 hours in the Yuma area while overnight visitors spent an average of 2.7 nights Spring 34% stayed 1-2 days and 17% stayed 3-4 days; summer 25% were here on a day trip while 30% stayed 3-4 days; fall 33% were here on a day trip and 34% stayed 1-2 days Table 2.8. Regional tourism surveys.
Long-Distance and rural area Data Sources 35 (continued on next page) Table 2.9. National and state park surveys. Characteristic Grand Canyon National Park Pennsylvania â Ohiopyle (SW), Pymatuning (NW), Greenwood Furnace (SC), Parker Dam (NC), Ridley Creek (SE), and Lackawanna (NE) Wyoming â Bear River, Boysen, Buffalo Bill, Curt Gowdy, Edness Kimball Wilkins, Glendo, Guemsey, Hawk Springs, Hot Springs, Keyhole, Medicine Lodge, Seminole, Sinks Canyon Chickamauga and Chattanooga National Military Park, Fort Oglethorpe, Georgia Data Collection Methods Intercept surveys at the park 25- to 35-minute interviews at the park Interviews â every 5th vehicle at park entrances Postcard intercept surveys and license tag O/D Sample Size 4,451 surveys returned (57% rate of return) 1,139 (adults over 18) 3,914 597 Data Collection Duration September 2003 to August 2004 summer 2008 (Memorial Day through Labor Day) 2009 July 2003 Who Traveled California (12.2%), Arizona (8.9%), Texas (4.8%), Florida (3.4%), and New York (3.2%). Foreign origins (17%), United Kingdom (3.8%), Canada (3.5%), Japan (2.1%), Germany (1.9%), and Netherlands (1.2%) N/A Half from out-of-state Park trips (17%) and cut-through trips (83%) Group Size Average 3.4; only 30% had children under 18 Average 5; majority visited in groups of 1-3 Average 3.14 Average 2.3 Age Average 48.5 Mostly between 36-50 Average 36.3 N/A Education 85.2% some college Advanced education N/A Repeat Visitors 58.6% first time; 41.4% repeat 64% previous visitors N/A Travel Mode Private vehicles (59.7%) combined with rental vehicles (37.4%) were used by 97.1% of respondents. Allowing for multiple responses, commercial airlines (16.4%) and RVs (7.8%, private and rental). N/A 75% travel with travel trailer, motor home, or other camping unit Auto (96.7%)
36 Long-Distance and rural travel transferable parameters for Statewide travel Forecasting Models Table 2.9. (Continued). Characteristic Grand Canyon National Park Pennsylvania â Ohiopyle (SW), Pymatuning (NW), Greenwood Furnace (SC), Parker Dam (NC), Ridley Creek (SE), and Lackawanna (NE) Wyoming â Bear River, Boysen, Buffalo Bill, Curt Gowdy, Edness Kimball Wilkins, Glendo, Guemsey, Hawk Springs, Hot Springs, Keyhole, Medicine Lodge, Seminole, Sinks Canyon Chickamauga and Chattanooga National Military Park, Fort Oglethorpe, Georgia Origins Traveling largely in personal vehicles, visitors averaged 792 miles of driving in Arizona, most often using Interstate 40 as a travel corridor (60.1%). Day users, passive recreationists (picnickers, beach/ pool users), lower- income, and minority visitors were more likely to say that the proximity of the study park to their home was excellent and were more likely to visit only that state park (and visit more frequently than other groups). Half from out-of- state, 40% of boaters were from Colorado Not reported Stay Duration N/A Day visitors Overnight visitors (average 45.3 hours) N/A