§ 10.01 – What is the purpose of this Sampling Manual?

(a)    It suggests procedures for obtaining annual data on unlinked passenger trips (UPT) and passenger miles traveled (PMT) for the National Transit Database (NTD) through random sampling according to the requirements in the NTD Reporting Manual.

(b)   In the event of a conflict between the requirements in the Reporting Manual and this Sampling Manual, the requirements of the Reporting Manual are definitive.

(c)    When it is possible to obtain annual UPT and PMT data according to the requirements in the Reporting Manual by using other methods not specified in this Sampling Manual, then you may do so.

§ 10.03 – What procedures does this Sampling Manual cover?

(a)    Development of sampling plans with two options:

(1)   Section 40 on ready-to-use sampling plans.

(2)   Section 50 on agency-developed template sampling plans with agency sample data.

(b)   Collection of sample data in Section 60.

(c)    Estimation of annual service-consumed data with two options:

(1)   Both UPT and PMT if you do not report 100% counts of UPT.

(2)   PMT if you report 100% counts of UPT.

(d)   Section 30 specifies several requirements for alternative sampling plans that template sampling plans must also meet. 

(e)    Figure 10.01 shows how these procedures and options relate to each other.

 Figure 10.01. Flow Chart of Procedures and Options

Flow Chart

§ 41.01 – Under what conditions may I use ready-to-use sampling plans?

(a)    Small Systems – If you operate no more than 30 vehicles operated in maximum service for all modes combined and if you choose to sample for determining your annual service-consumed data for any of the modes you operate. 

(b)   New Mode – If you will be sampling and reporting for the first time this current report year for a particular mode that you do not already operate.  For example, you would meet this condition if you will add light rail (LR) service this year, but you have not operated this service previously, or

(c)    New Type of Service – If you will be sampling and reporting this current report year for a particular type of service for the first time.  For example, you would meet this condition if you previously directly operated all of your bus (MB) service, but will contract out part or all of that service to a private entity for this year, or

(d)   Major Changes of Service – You have made major changes to your service since your last sampling year, or

(e)    No Sample Data – If you have reported your service to the NTD before through random sampling, but no longer have the original sample data, or

(f)    Sample Data Not Reliable – If you have reported your service to the NTD before through random sampling, but the sample data are found to be not reliable, or

(g)   Sample Data Not Enough – if you have reported your service to the NTD before through random sampling, but the sample size is smaller than 50 service units.

§ 41.03 – If I am using a ready-to-use sampling plan this year, may I use it again for my next report year?

(a)    You should not use it again if your next report year is your mandatory sampling year.  After you have collected the sample data from this year, you should develop a template sampling plan with that sample data for your next report year.

(b)   You may use it again for the next report year if that year is not your mandatory sampling year.

(c)    You may use it while you operate no more than 30 vehicles operated in maximum service for all modes combined.

§ 41.05 – For what modes are ready-to-use sampling plans available?

(a)    For non-scheduled services, they are available for demand response services (DR and DT) and commuter vanpool.  You should not use the ready-to-use sampling plans for commuter vanpool if your vanpool service does not serve commuters exclusively.

(b)   For scheduled services, they are available for bus services (bus (MB), commuter bus (CB), bus rapid transit (RB), and trolleybus (TB)), commuter rail (CR), and other rail modes. 

§ 41.07 – What sampling options are available?

(a)    The available sampling options vary in:

(1)   the unit of sampling and measurement,

(2)   sampling structure,

(3)   efficiency options, and

(4)   sampling frequency.

(b)   The particular unit of sampling and measurement used in these ready-to-use sampling plans varies by mode and whether the service is scheduled (Table 41.01).

(1)   For non-scheduled services, the unit is in vehicle days.

(2)   For bus services (MB, CB, RB, and TB), separate sampling plans are available with units in one-way trips and in round trips.

(3)   For commuter rail, the unit is in one-way car trips.

(4)   For other rail modes, separate sampling plans are available in units of one-way car trips and one-way train trips.

Table 41.01.  Options for Unit of Sampling and Measurement

Service

Mode

Units of Sampling and Measurement

Non-Scheduled

Demand Response (DR, DT)

Vehicle days

Commuter Vanpool

Vehicle days

Scheduled

Bus (MB, CB, RB, TB)

One-way trips, round trips

Commuter Rail (CR)

One-way car trips

Other Rail Modes

One-way car trips, one-way train trips

(c)    Two options are provided for sampling structure—period-based and interval-based.  Interval-based sampling plans are available for bus services only.

(d)   Three efficiency options are provided:

(1)   Base Option – you must estimate both UPT and PMT through random sampling.

(2)   APTL Option – you must report a 100% count of UPT, estimate the average passenger trip length (APTL) through random sampling, and obtain annual PMT by multiplying the 100% UPT with the estimated APTL.

(3)   Grouping Option – you must divide your bus routes into two groups by route length.  The grouping option is available for period-based sampling plans only.

(e)    Three frequency options are provided for period-based sampling plans—quarterly, monthly, or weekly.  You may choose whichever of these options is best suited for your agency.  One factor to consider is that the annual realized sample size may be larger at a lower sampling frequency due to rounding.  Another factor is that a lower sampling frequency means a larger annual number of acts for random sampling.

(f)    Up to six frequency options are provided for interval-based sampling plans—every day, every 2nd day, every 3rd day, every 4th day, every 5th day, and every 6th day.  Three sets of these sampling plans are available for 7-day weekly service, 6-day weekly service, and 5-day weekly service, respectively.  The sampling plans for 7-day weekly service and the base option are similar to those in Circular 2710.1A.

§ 43.01 – What period-based sampling plans are available for non-scheduled services?

 (a)    Table 43.01 shows the sampling plans available for demand response services (DR and DT) and commuter vanpool, respectively.

(b)   Sample size is stated in the number of vehicle days.

(c)    Separate sampling plans are available for the base option and the APTL option.

(d)   Sample size is shown for the entire year and for the relevant period for each frequency.

Table 43.01.  Period-Based Ready-to-Use Sampling Plans for Non-Scheduled Services

Sampling Frequency Sample Size

Demand Response

Commuter Vanpool

Reporting 100% UPT

(APTL Option)

Not Reporting 100% UPT

(Base Option)

Reporting 100% UPT

(APTL Option)

Not Reporting 100% UPT

(Base Option)

Quarterly Vehicle Days for a Quarter

13

22

31

45

Total Sample Size for Year

52

88

124

180

Monthly Vehicle Days for a Month

5

8

10

15

Total Sample Size for Year

60

96

120

180

Weekly Vehicle Days for a Week

1

2

2

4

Total Sample Size for Year

52

104

104

208

§ 43.03 – What period-based ready-to-use sampling plans are available for bus services?

(a)    Table 43.03 shows the available ready-to-use sampling plans for bus services, including bus (MB), commuter bus (CB), bus rapid transit (RB), and trolleybus (TB). 

Table 43.03.  Period-Based Ready-to-Use Sampling Plans for Bus Services

Sampling Frequency

Sample Size for Period and Year

One-Way Trips

Round Trips

Reporting 100% UPT

(APTL Option)

Not Reporting 100% UPT

(Base Option)

Reporting 100% UPT

(APTL Option)

Not Reporting 100% UPT

(Base Option)

With Route Grouping

Without Route Grouping

With Route Grouping

Without Route Grouping

(1)

(2)

(3)

(4)

(5)

(6)

Quarterly

Trips for a Quarter

52

78

138

39

59

103

Total Sample Size for Year

208

312

552

156

236

412

Monthly

Trips for a Month

18

27

46

13

20

35

Total Sample Size for Year

216

324

552

156

240

420

Weekly

Trips for a Week

4

6

11

3

5

8

Total Sample Size for Year

208

312

572

156

260

416

(b)    Sampling plans are provided separately for one-way trips and round trips. 

(c)   The number of one-way trips in a sampling plan based on round trips is about 50% greater than the number of one-way trips in a sampling plan based on one-way trips.  A sampling plan based on round trips requires a larger number of one-way trips because the pair of one-way trips making up a round trip are not selected randomly or independently.

(d)    If you choose the base option:

(1)   use column (3) to find the sample size in one-way trips, and

(2)   use column (6) to find the sample size in round trips.

(e)   If you choose the APTL option, you may choose one of two options:

(1)   With Route Grouping – Use column (1) to find the sample size in one-way trips and column (4) to find the sample size in round trips.  In using this option, you must divide your routes into two groups on the basis of route length and do sampling and estimation separately for each group.  For example, if you operate 10 routes, put the 5 routes with the shortest route distances in the group of short routes and the other 5 routes in the group of long routes. 

(2)   Without Route Grouping – If you prefer not to deal with grouping your routes, use column (2) to find the sample size in one-way trips and use column (5) to find the sample size in round trips.

§ 43.07 – What period-based sampling plans are available for commuter rail?

(a)    Table 43.05 shows the available ready-to-use sampling plans for commuter rail (CR). 

(b)   Sample size is in one-way car trips.

(c)    Separate sampling plans are available for the base option and the APTL option.

Table 43.05.  Period-Based Ready-to-Use Sampling Plans for Commuter Rail (CR)

Sampling Frequency

Sample Size for Period and Year

Reporting 100% UPT

(APTL Option)

Not Reporting 100% UPT

(Base Option)

Quarterly

One-Way Car Trips for a Quarter

13

80

Total Sample Size for Year

52

320

Monthly

One-Way Car Trips for a Month

5

27

Total Sample Size for Year

60

324

Weekly

One-Way Car Trips for a Week

1

7

Total Sample Size for Year

52

364

§ 43.09 What period-based sampling plans are available for other rail modes?

(a)    Table 43.07 shows the available ready-to-use sampling plans for other rail modes, including light rail (LR), heavy rail (HR), monorail and automated guideway (MG). 

(b)   Separate sampling plans are available for one-way train trips and one-way car trips.

(c)    Separate sampling plans are available for the base option and for the APTL option.

Table 43.07.  Period-Based Ready-to-Use Sampling Plans for Other Rail Modes

Sampling Frequency

Sample Size for Period and Year

One-Way Train Trips

One-Way Car Trips

Reporting 100% UPT

(APTL Option)

Not Reporting 100% UPT

(Base Option)

Reporting 100% UPT

(APTL Option)

Not Reporting 100% UPT

(Base Option)

Quarterly

Trips for a Quarter

13

45

13

72

Total Sample Size for Year

52

180

52

288

Monthly

Trips for a Month

5

15

5

24

Total Sample Size for Year

60

180

60

288

Weekly

Trips for a Week

1

4

1

6

Total Sample Size for Year

52

208

52

288

§ 43.11 – What interval-based ready-to-use sampling plans are available for bus services?

(a)    Table 43.09 shows the available ready-to-use sampling plans for bus services, including bus (MB), commuter bus (CB), bus rapid transit (RB), and trolleybus (TB). 

(b)   Separate sampling plans are available for the base option and the APTL option.

(c)    In determining whether using interval-based sampling plans, you should consider the tradeoff between at least two factors:

(1)   the advantage of using a sampling plan that you have been using for years, and

(2)   the potentially lost opportunity of a lower sample size.

(d)   You should first determine the number of weekly operating days for your next sampling year, and choose the set of sampling plans accordingly.  For example, if you will be operating 7 days a week, you should not consider any of the sets for the other weekly operating patterns.

(e)    You may choose any of the above plan numbers that correspond to the weekly operating pattern of your service.

Table 43.09.  Interval-Based Ready-to-Use Sampling Plans for Bus Services

Plan Number

Frequency of Sampling

Operating 7 Days a Week

Operating 6 Days a Week

Operating 5 Days a Week

Daily One-Way Bus Trips

Total Sample Size for Year

Daily One-Way Bus Trips

Total Sample Size for Year

Daily One-Way Bus Trips

Total Sample Size for Year

Base Option

1

Every Day

2

730

2

624

2

520

2

Every 2nd Day

3

549

3

468

4

520

3

Every 3rd Day

5

610

6

624

7

609

4

Every 4th Day

7

644

9

702

12

780

5

Every 5th Day

10

730

13

819

N/A

N/A

6

Every 6th Day

15

915

N/A

N/A

N/A

N/A

APTL Option

1

Every Day

1

365

1

312

1

260

2

Every 2nd Day

2

366

2

312

3

390

3

Every 3rd Day

3

366

4

416

5

435

4

Every 4th Day

4

368

6

468

8

520

5

Every 5th Day

6

438

8

504

 

 

6

Every 6th Day

10

610

 

 

 

 

§ 52.07 – What should I consider in choosing a sampling frequency?

(a)    Period-based sampling plans:

(1)   Your cycle of minor schedule changes during a year.  For example, quarterly sampling may be appropriate if you routinely adjust your schedule every three months. 

(2)   The scale of your operation and the method you will be using to select a sample.  For example, it may become difficult to select a sample for an entire quarter if the number of service units involved in a quarter is too large for a spreadsheet to handle.

(3)   The realized annual sample size.  The realized annual sample size can be larger than the initial annual sample size due to rounding in allocating the initial annual sample to each frequency level.  This is particularly true for weekly sampling when the initial annual sample size is relatively small.  For example, if the initial annual sample size is 55 and you choose weekly sampling, the weekly sample size would be 2 and the realized annual sample size would be 104. 

(b)   Interval-based sampling plans:

(1)   Your staffing needs for data collection.

(2)   The number of times you need to do sampling.  The number of times is larger with plans of higher frequency (e.g., every vs. every 2nd day). 

(3)   The annual total sample size.  The degree of rounding is higher with plans of higher frequency.

§ 61.01 – What are the basic elements of collecting sample data?

(a)    Selecting a sample at random according to your sampling plan (Subsection 63).

(b)   Collecting data from the random sample (Subsection 65).

(c)    Identifying and correcting any errors in the sample data (Subsection 67).

§ 61.03 – What criteria does this section cover to ensure that estimates of annual service-consumed data meet FTA’s 95% confidence and 10% precision levels?

(a)    Your sampling process covers your entire service.

(b)   You select your sample at random according to your chosen ready-to-use sampling plan or your template sampling plan. 

(c)    Your data-collection process is designed to avoid errors from happening.

(d)   Your data-collection is designed to identify and correct errors when they occur. 

§ 63.01 – What are the basic elements of selecting a sample at random?

(a)    Choosing a method with which you select a sample at random.

(b)   Developing lists of service units that you expect to operate:

(1)   A single list of all services if your sampling plan is not based on service grouping.

(2)   One list for each service group if your sampling plan is based on service grouping.

(c)    Selecting a sample at random from each list with the chosen method.

§ 63.03 – What method may I use for random sampling?

(a)    You may use the traditional method based on a table of random numbers.

(b)   You may use any other method for random sampling as long as it meets these two criteria:

(1)   sampling under the method is random. 

(2)   sampling under the method is without replacement.  Without replacement means that the method will not select the same service unit more than once.

§ 63.05 – What is a table of random numbers?

(a)    It is a list of integers whose frequency and sequence of appearance in the list have been determined entirely by chance. 

(b)   For convenience and simplicity in use, published tables of random numbers usually appear in the form of separate columns of five-digit numbers.  Both rows and columns may be consecutively numbered for easy reference.  Table 63.01 shows an example.

(c)    Appendix 98 is a comprehensive table of random numbers you may use for random sampling if you choose to use this method.

Table 63.01. Example of a Table of Random Numbers

Rows

(1)

(2)

(3)

(4)

(5)

1

10480

15011

01$536

02011

81647

2

22368

46573

25595

85393

30995

3

24130

48360

22527

97265

76393

4

42167

93093

06243

61680

07856

5

37570

39975

81837

16656

06121

§ 63.07 – What is the list of all service units that I expect to operate?

 

(a)    It is the amount of revenue service that you expect to operate. 

(b)   It is measured with the following characteristics:

(1)   It is in the unit of sampling and measurement of the sampling plan you have chosen.

(2)   It is for the duration corresponding to the sampling frequency that you have chosen for your sampling plan.  The duration would be a day for interval-based sampling but a week, a month, or a quarter for period-based sampling.

(3)   It is for each of the service groups you have defined if your sampling plan is based on service grouping.

(c)    For scheduled services, the list must include:

(1)   all service units that are listed on the schedule, and

(2)   all service units that are not on the schedule but are expected to be operated, such as trippers, shuttles, and other special operations.

(d)   It would be the number of one-way bus trips you expect to operate in a week by your MB express routes, for example, if:

(1)   your sampling plan is in terms of one-way bus trips,

(2)   you have chosen weekly sampling for your MB service,

(3)   your sampling plan involves service grouping, and

(4)   you use express routes as one of the groups.

§ 63.09 – How do I develop the list of all service units for using a table of random numbers?

(a)    Suppose that:

(1)   you have chosen weekly sampling for your vanpool service,

(2)   your sampling plan requires 2 vanpool-days per week,

(3)   you want to select a random sample for next week, and

(4)   you have 101 vanpools and each of them is expected to operate every day next week.

(b)   One way to develop the list would be to assign a serial number of four-digits to each vanpool-day for all 707 combinations of vanpools and service days.  The first digit would represent the day of week with 1 for Monday and 7 for Sunday.  The other three digits would represent vanpool numbers ranging from 001 through 101.

(c)    The vanpool numbers may differ from what you use for your internal purposes.  Make sure that you have a one-to-one list between the internal vanpool numbers and the new serial numbers if they differ.

(d)   The serial numbers on a given day must be consecutive without any gaps.

(e)    Write down just the first and the last for each day of week so that you will know the range of numbers.  For example, the range for Monday is 1001-1101.

(f)    You should summarize these serial numbers in a summary table as in Table 63.03.

Table 63.03.  Example of a Summary Table of Serial Numbers

Day of Week

Range of Serial Numbers

1 (Monday)

1001-1101

2

2001-2101

3

3001-3101

4

4001-4101

5

5001-5101

6

6001-6101

7

7001-7101

§ 63.11 – How do I use a table of random numbers?

 

(a)    Suppose that:

(1)   you have 101 vanpools that you expect to operate every day of next week,

(2)   you have numbered the vanpool days for all vanpools as shown in Table 63.03, and

(3)   your sampling plan for your vanpool service requires 2 vanpool days per week. 

(b)   You must work with a constant length of digits from a table of random numbers for sampling.  This constant length is given by the number of digits in the longest serial number you have assigned to members of the list of all service units. 

(1)   This constant length would be 4 for the example in Table 63.03 because all serial numbers are four-digits long. 

(2)   This constant length would be 5 if your serial numbers vary in length, ranging from one to five digits.

(c)    You must combine adjacent digits from the table of random numbers as needed to obtain a two-, three-, or four-digit number, or any other length number from the table. 

(1)   You may choose to work with rows or columns in combining adjacent digits.  You may start with any row if you choose to work with rows.  You may start with any column if you choose to work with columns.  You may also choose to work from right to left or from left to right if you choose to work with rows. 

(2)   If you choose to work with rows and to start with row 1, for example, the first 9 four-digit numbers from Table 63.01 would be:

 

1048, 0480, 4801, 8015, 0150, 1501, 5011, 0110, and 1101.

 

(d)   You must continue forming four-digit numbers until you find two four-digit numbers from the table of random numbers that are in the summary table of serial numbers (Table 63.03).  In the above example, these two numbers are 1048 and 1101.  That is, you should select vanpools 48 and 101 and collect sample data from them on Monday of next week.

(e)    Once you are done sampling for a week, you should mark where you stopped forming four-digit numbers in the table of random numbers.  For example, the symbol, $, has been inserted between digits 1 and 5 in row 1 and column 3 of Table 63.01.  You should start the above process for a later week after that mark.

(f)    Once you are done sampling for a year, you should mark where you stopped forming four-digit numbers in the table of random numbers if you plan to use this method of random sampling in the future.  

(1)   If you number your service units exactly as for a previous sampling year, you should start the above process for a new sampling year after the last mark of your previous sampling year.

(2)   If you number your service units with a different approach for a new sampling year, you may start the above process anywhere in the table of random numbers.

§ 63.13 – What are the pros and cons of using a table of random numbers?

 

(a)    Using a table of random numbers for random sampling has the following advantages:

(1)   It is applicable to all situations.

(2)   You do not need to list explicitly all service units.

(b)   It has the following disadvantages:

(1)   You must number your service units consecutively without gaps, at least within each subset of your service units.  Subsets are formed by route, type of service days, etc.

(2)   You may not be able to take advantage of the serial numbers you have already assigned to your service units for internal purposes.

(3)   It can be difficult to use if the serial numbers have a large number of digits.

§ 63.15 – What information should I keep from my sampling process?

 

(a)    You should have an auditable record of your sampling process.  That record should cover the following:

(1)   A written description of the sampling plan.

(2)   A written procedure for your method of selecting a sample at random.

(3)   The list of all service units from which you selected a sample at random for each act of sampling.  For example, you should have 52 such lists if you did weekly sampling for an entire year.

(4)   The random sample.

§ 65.01 – What method may I use to collect the sample data?

(a)    One common method involves one or more ride checkers observing and recording passenger activities while riding in a transit vehicle. 

(b)   An increasingly common method uses APCs to record passenger activities instead. 

(c)    You may use one of these or any other method to collect the data from each service unit of your random sample as long as your method meets these criteria:

(1)   it obtains data from direct measurement or direct observation without passenger intercept,

(2)   it provides data you must have to determine PMT for each service unit, and

(3)   it meets the requirements on measurement accuracy in the NTD Reporting Manual.

§ 65.03 – What approach may I use to determine PMT for each service unit of my sample?

(a)    You may use the load-based approach.  It determines PMT by multiplying the number of passengers onboard a transit vehicle between each pair of consecutive stops by the distance between these stops.

(b)   You may also use the distance-based approach.  It determines PMT by keeping track of the distance traveled by each passenger carried by a service unit of your random sample.

§ 65.05 – What data items must I collect to use the load-based approach for scheduled services?

(a)    You must collect the required data items separately for each one-way vehicle trip in your service unit. 

(1)   If your service unit is a round-trip bus trip, for example, it has two one-way vehicle trips. 

(2)   If a service unit is a one-way train trip with three passenger cars, for example, it has three one-way vehicle trips.

(b)   The following data items are required for each one-way vehicle trip:

(1)   the number of people who boarded at each stop,

(2)   the number of people who alighted at each stop,

(3)   the distance between any pair of consecutive stops at which boardings or alightings occurred,

(4)   the number of people onboard the vehicle between any pair of consecutive stops,

(5)   the number of people who stayed on from the previous one-way vehicle trip, and

(6)   the number of people who remained on the vehicle at the last stop.

§ 65.07 – What data must I collect to use the load-based approach for non-scheduled services?

(a)    For commuter vanpool, the following data items are required for each direction of commuting:

(1)   the number of people who boarded at each pick-up location,

(2)   the number of people who alighted at each drop-off location, and

(3)   the distance between any pair of consecutive stops at which pick-up or drop-off occurred.

(b)   For demand response (DR and DT), you must collect the required data items continuously during the entire vehicle day sampled:

(1)   the odometer reading at each pick-up location,

(2)   the number of people onboard between each pair of consecutive locations at which pick-up or drop-off occurred, and

(3)   the odometer reading at each drop-off location.

(c)    You should treat non-commuter vanpool as demand response for data collection. 

§ 65.09 – How should I determine between-stop distances for the load-based approach?

(a)    You should avoid using maps to estimate between-stop distances under all circumstances.

(b)   For services without designated stops, including demand response (DR and DT), jitney (JT), vanpool (VP), or fixed-route services without designated stops:

(1)   If available, you should always use the onboard odometer to determine the between-stop distances for each service unit in your random sample.  You should make sure that the vehicles for the sampled service unit have a working odometer onboard before they start for the sampled service unit.

(2)   If your vehicle does not have an onboard odometer, you should determine the distance by retracing the path and the stops by automobile.

(3)   You should record your odometer readings at least to one-tenth of a mile.

(c)    For ferryboat (FB), aerial tramway (TR) and all rail services, you may want to predetermine the between-station distances for all routes and directions.

(d)   For fixed-route services with designated stops, you should use one of two approaches:

(1)   predetermine the between-stop distances for all routes and directions, or

(2)   record the onboard odometer readings at individual stops.

(e)    You may use different methods to predetermine the between-stop distances:

(1)   use an up-to-date GIS of your network of routes and stops, or

(2)   record the odometer readings while you drive through all of your routes and related deviations and directions.

(f)    If you predetermine between-stop distances, you must keep them updated to reflect any changes in your services.

§ 65.11 – What additional data should I collect to identify each service unit of my sample?

 

(a)    You must record the date and the type of service days for all cases. 

(b)   For commuter rail (CR), heavy rail (HR), and light rail (LR), you must also record the weekday time period that you will be reporting to the NTD for weekdays.

(c)    If your sampling plan is based on the PPMT option, you must record route identification.

(d)   If your sampling plan is based on service grouping, you should also record information that is necessary to determine group membership of each service unit in the random sample.

§ 65.13 – When do I use the different approaches to determining PMT?

(a)    The load-based approach is applicable to all circumstances.  Under conditions of heavy loads or high boarding volumes, however, extra care must be taken to maintain the required level of measurement accuracy.

(b)   The distance-based approach is useful under several circumstances.  For examples:

(1)   All passengers board and alight at the same locations, such as inclined plane (IP), aerial tramway (TR), and most ferryboat (FB) operations.

(2)   Services with a small number of passengers who board and alight at a small number of stops, such as demand response (DR and DT).

(3)   Services with a small number of frequent passengers who board and alight at a small number of stops, such as commuter vanpool.

(4)   Any service with a ticketing system that keeps track of the origin and destination for every boarding with a known distance.

§ 65.15 – What instrument should I use to collect the data for the distance-based approach?

(a)    You may use any instrument that you have designed as long as you can use it to record the required data items correctly.

(b)   If you use human ride checkers, the instrument may be a piece of paper and a pencil or it may be a hand-held device.

(c)    Appendix 92 provides an example of both blank and filled-out paper instruments for the distance-based approach.

§ 65.17 – What instrument should I use to collect data for the load-based approach?

(a)    You may use any instrument that you have designed as long as you can record the required data items correctly.

(b)   If you use APCs, the instrument would be computer software and hardware that records the counts and other data items transmitted from the APCs.

(c)    If you use human ride checkers, the instrument may be a piece of paper and a pencil or it may be a hand-held device.

(d)   If you use a hand-held device, it is critical that the unit accepts counts of boardings and alightings that may not be equal for a one-way vehicle trip.

(e)    The instrument you design may take slightly different formats for different services.  Three examples of blank and filled-out instruments are shown as appendices:

(1)   demand response – Appendix 93.

(2)   commuter vanpool – Appendix 94.

(3)   fixed-route services – Appendix 95.

(f)    If you are going to use one of these example instruments, you should study it carefully before reading the following guidance.

§ 65.19 – What pre-survey procedures should I follow if I use a paper instrument and human ride checkers?

(a)    You should use an appropriate survey sheet for each service unit selected in the sample.  This survey sheet may be one of those from the appendices or one that you have designed, but it is important that you select a survey sheet that is designed for the service to be surveyed.

(b)   For non-scheduled services, including demand response (DR and DT) and vanpool (VP), a survey sheet should be used for each driver of the survey vehicle during an entire day.

(c)    For scheduled services, a survey sheet should be used for each one-way vehicle trip of your service unit that you have chosen for your sampling plan.

(d)   The ride checker(s) should carry several extra survey sheets in the event that extra sheets are needed.

(e)    You should learn about the likely load and boarding volumes for each service unit to be surveyed. 

(1)   You may need to use more than one ride checker under conditions of high loads and high boarding volumes. 

(2)   If more than one ride checker is used, the separate survey sheets should also be labeled with the door(s) that each ride checker is responsible for.

(f)    If you rely on the onboard odometer to determine the between-stop distances, make sure that it works properly.  If you cannot fix a malfunctioning odometer in time for a ride check, use a different vehicle with a working odometer. 

(g)   Before going into the field, the survey supervisor should fill in the data items that identify the service unit to be surveyed.  If your service unit has more than one one-way vehicle trip, identify each one-way vehicle trip separately.

(h)   If you rely on predetermined between-stop distances, the survey supervisor should also use the survey sheet(s) in Appendix 95 to:

(1)   Fill in the stop numbers in column (7) and stop descriptions in column (8) for all stops.

(2)   Cross out column (9) to avoid confusion in the field.

(i)     You should write the page numbers in the box in the lower right-hand corner of each survey sheet if you use more than one survey sheet for a given one-way vehicle trip.

§ 65.21 – What manual survey procedures should I follow for demand response (DR and DT) with the distance-based approach?

(a)    You may use the survey sheet for the distance-based approach in Appendix 92.

(b)   You must record individual trips by each pair of origin and destination.  The filled-out form in Appendix 92 illustrates three cases of pick-up and drop-off patterns:

(1)   The first pick-up illustrates the “one origin-many destinations” case.  The driver picked up 16 passengers at 1020 J Street.  He immediately distributed these passengers according to their three drop-off destinations under item (8). 

(2)   The second pick-up illustrates the “one origin-one destination” case.  At the next pick-up address, 506 10th Street, 17 passengers boarded and all were driven to the same destination.  Here, a single line records all the necessary information.

(3)   The last two pick-ups illustrate a “many origins-one destination” case.  Here, the driver records the pick-up addresses and repeats the destinations; that is, he handles each as an individual trip.

(c)    You must record the odometer readings for each pick-up and drop-off.

(1)   Item (7), “Pick-Up Odometer Reading,” should be recorded immediately upon picking up the passenger(s). 

(2)   Item (8), “Drop-Off Odometer Reading,” should be recorded immediately upon dropping off the passengers. 

(3)   You should also record the odometer readings to at least one-tenth of a mile when the odometer allows.

(d)   You should use additional pages if a service unit involves more stops than are given on a page.  You should write the page numbers in the box in the lower right-hand corner.

(e)    As an alternative to data recording by the driver, the driver can call in the information to the dispatcher if you have two-way radio communications on all vehicles.

(1)   When the driver arrives at a pick-up point, he can call in items (5)-(8).

(2)   When he arrives at each destination, he can call in item (9) and verify how many passengers were dropped off at that destination.

§ 65.23 – What manual survey procedures should I follow for demand response (DR and DT) with the load-based approach?

(a)    You may use the survey sheet for the load-based approach for demand response in Appendix 93.

(b)   For each pick-up or drop-off, you must record the following:

(1)   whether it is a pick-up or drop-off in column (5),

(2)   a description of the location in column (6),

(3)   the odometer reading in column (7), and

(4)   the leaving load in column (8).

(c)    The filled-out form in Appendix 93 shows the recorded information for the example shown in Appendix 92.  The leaving load at the last drop-off location should be 0.

(d)   You should record the odometer readings to at least one-tenth of a mile.

(e)    You should use additional pages if a service unit involves more stops than are given on a page.  You should write the page numbers in the box in the lower right-hand corner.

(f)    As an alternative to data recording by the driver, the driver can call in the information to the dispatcher if you have two-way radio communications on all vehicles.  Before leaving each pick-up or drop-off location, the driver can call in items (5)-(8).

§ 65.25 – What manual survey procedures should I follow for commuter vanpool with the load-based approach?

(a)    Suppose that you have chosen to sample on a monthly basis.

(b)   Before a new month starts, you should communicate with the driver of each sampled vanpool about the days on which he must collect sample data during the new month.

(c)    For each sampled vanpool day, the driver should fill in the identification data before he leaves home, including the date, the day of week, and the vanpool number. 

(d)   The driver is required to record the travel data, including the van odometer reading whenever any rider gets on or off the van along with the number of riders who get on the van and the number of riders who get off the van.  This recording is to be done separately for travel to work and travel from work.  The filled-out form in Appendix 94 shows the travel data for a vanpool of 6 riders with a round-trip-distance of 78 miles.  The driver picks up all 5 riders at a single location but drops them off at three different locations in the morning.  In the afternoon, the process reverses itself. 

(1)   Once the driver gets on the van in the morning, he should enter 1 in the ON column (6) and record the odometer reading at 29,366.0 in column (8). 

(2)   He then leaves for picking-up 5 fellow vanpoolers at another location.  Once they get on the van, he should enter 5 in (6), and record the odometer reading at 369.1 in (8) before departing that location.

(3)   At the first drop-off location, 3 vanpoolers get off.  Before leaving, the driver should enter 3 in the OFF column (7), and record the odometer reading at 395.3 in (8).

(4)   One vanpooler gets of the van at each of the next two drop-off locations.  Before leaving these locations, the driver should enter 1 in the OFF column, and record the odometer reading at 396.8 and 397.5, respectively. 

(5)   Finally, the driver arrives at his own destination.  Before he leaves the van, he should record the odometer reading again at 405.0.

(e)    The day after each sample day, you should communicate with each driver involved in collecting sample data to determine if he actually recorded the sample data.  Sample data may be not recorded for a variety of reasons. 

(1)   If the data were collected, the driver should send the filled-out survey sheet immediately,

(2)   If the driver forgot to collect the data, he should be asked to collect the data next day,

(3)   If the driver refused to collect the data or the vanpool has been terminated, a replacement vanpool day should be selected at random.

§ 65.27 – What instructions should I give my ride checkers if I do not use predetermined between-stop distances for fixed-route services?

(a)    Take enough copies of the survey sheet in Appendix 95 or an alternative sheet.

(b)   Use separate survey sheets for separate one-way vehicle trips if your service unit has more than one one-way vehicle trip.

(c)    Use additional survey sheets if needed for a given one-way trip.

(d)   Board the transit vehicle at the beginning point of the service unit and position yourself so that you can observe the doors for which you are responsible. 

(e)    Before the vehicle leaves the beginning point, record:

(1)   Stop #1 in column (7),

(2)   stop description of the beginning point in column (8),

(3)   odometer reading to at least one-tenth of a mile in column (9),

(4)   number of passengers who have stayed onboard from the last trip (13), and

(5)   number of passengers boarded in (10), including the passengers who have stayed onboard from the last trip in (e)(4).

(f)    When the vehicle leaves the beginning point, record the number of passengers onboard in column (12), including any passengers who have stayed onboard from the last trip.

(g)   Only at points where the vehicle stops during the service unit, record:

(1)   stop number in (7),

(2)   stop description in (8),

(3)   odometer reading (from the driver) in (9),

(4)   passengers boarded in (10), and

(5)   passengers alighted in (11).

(h)   Between stops, count the number of passengers on board (12).  You should record this number as the leaving load.  For example, between stops 2 and 3 you should record this number in the row for stop 2 rather than in the row for stop 3.

(i)     At the end point of each one-way trip, record the number of passengers who will stay on board to the next trip in column (14) and in column (11) as passengers alighted.

§ 65.29 – What steps should I take if I fail to collect the sample data from a particular unit?

(a)    You may fail to collect the required sample data from any particular unit in your sample:

(1)   The unit may be canceled for a variety of operational reasons.

(2)   The ride checker(s) may fail to show up for the unit.

(3)   The ride checker(s) may fail to collect the data.

(b)   You must replace the missed unit as soon as possible on the same type of service days.

(c)    For period-based sampling:

(1)   If the rest of the sampling period (a week, a month, or a quarter) has at least one service day of the same type as the missed service unit, select one service unit from the next service day of the same type.  For example, if the missed service unit is on the second Monday of a month and your sampling is monthly, you should get a replacement unit from the third Monday of the month.

(2)   If the rest of the sampling period does not have any service day of the same type left, select one service unit from the first service day of the same type during the next sampling period.  For example, if the missed service unit is on a Monday and your sampling is weekly, you should get a replacement unit from the next Monday. 

(d)   For interval-based sampling:

(1)   If the rest of the report year has at least one service day of the same type as the missed service unit, select one service unit from the next service day of the same type.

(2)   If the rest of the report year does not have any service day of the same type left, select one service unit from the next day of service.

(e)    The replacement unit must be selected at random in all cases.

§ 67.01 – What steps should I take after I have collected the sample data?

(a)    Design a format for recording your sample data.

(b)   Enter the raw data.

(c)    Process the entered data.

(d)   Identify errors in the entered data.

(e)    Identify sources of the data errors, if any.

(f)    Correct the data errors, if any.

(g)   What you should do within each step depends on your situation.  To be specific, the following example is used for the rest of this subsection:

(1)   you collect sample data with human ride checkers from one-way bus trips with designated stops,

(2)   you use the load-based approach to determining PMT,

(3)   you use predetermined between-stop distances, and

(4)   you have collected sample data from a route whose longest one-way trip is 4 miles.

§ 67.03 – What should I consider in designing the format for data recording?

(a)    The data items you have collected to determine PMT.  The exact data items depend on your service and the approach you have taken to determine PMT.

(b)   The data items that identify the service units in your sample.

(c)    A format that is easy for data analysis.

(d)   Suppose that you have designed a format in Table 67.01.  With this format, you enter the field data (7)-(14) as they appear on the field survey sheet shown in the filled-out form in Appendix 95.  Item (8) is not shown due to space limit.  The identification data items at the top of the survey sheet (1)-(6) are repeated for each stop of this one-way trip.

Table 67.01.  Example Format for Data Recording

Date

Day of Week

Time Period

Route No.

Trip No.

Direction

Stop Sequence

Distance to Next Stop

No. of Pass. Boarded

No. of Pass. Alighted

No. of Pass. On Board (Leaving Load)

No. of Pass. from Previous Trip (Stop 1 only)

No. of Pass. Continuing to Next Trip (Last Stop only)

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(9)

(10)

(11)

(12)

(13)

(14)

10/13/05

Thur

Midday

11

408

Outbound

1

0.3

20

0

20

2

 

10/13/05

Thur

Midday

11

408

Outbound

2

0.7

2

1

21

 

 

10/13/05

Thur

Midday

11

408

Outbound

3

0.6

0

2

19

 

 

10/13/05

Thur

Midday

11

408

Outbound

4

0.3

1

3

17

 

 

10/13/05

Thur

Midday

11

408

Outbound

5

0.5

1

10

8

 

 

10/13/05

Thur

Midday

11

408

Outbound

6

0.8

0

2

6

 

 

10/13/05

Thur

Midday

11

408

Outbound

7

0.2

0

1

5

 

 

10/13/05

Thur

Midday

11

408

Outbound

8

0.1

0

2

3

 

 

10/13/05

Thur

Midday

11

408

Outbound

9

0.1

0

2

1

 

 

10/13/05

Thur

Midday

11

408

Outbound

10

0.3

0

0

1

 

 

10/13/05

Thur

Midday

11

408

Outbound

11

0.1

0

0

1

 

 

10/13/05

Thur

Midday

11

408

Outbound

12

0.0

0

1

1

 

1

§ 67.05 – When should I enter my sample data?

(a)    You should have the sample data from each service unit entered immediately after your ride checker(s) has come back from the field.

(b)   This has a number of advantages:

(1)   reducing the chance of loss of information,

(2)   increasing the chance of correcting data errors, if any,

(3)   being ready to use the cumulative sample data any time during the sampling year, and

(4)   minimizing the rush at the end of the year.

§ 67.07 – How should I process the entered data?

(a)    You should process the data as follows immediately after you have entered them.  If your service unit has more than one one-way vehicle trip, you should process the data for individual one-way vehicle trips:

(1)   Sum the number of passengers boarded at individual stops in column (10) of Table 67.01 to get the total number of passengers boarded (UPT).  It is 24 in this case.

(2)   Sum the number of passengers alighted at individual stops in column (11) of Table 67.01 to get the total number of passengers alighted.  It is 24 in this case.

(3)   Calculate the load between every pair of two consecutive stops.  Table 67.03 shows an example and the formulas for calculating leaving loads and arriving loads.

(4)   Calculate PMT for each pair of consecutive stops by multiplying the calculated load with the between-stop distance.  Table 67.03 also shows the calculation of PMT with both leaving loads and arriving loads.

(5)   Calculate total PMT.  It is 47.8 miles in this case.

(6)   Divide total PMT by the total number of passengers boarded to get APTL.  It is 1.99 miles in this case.

(7)   Calculate vehicle trip length by summing up the predetermined between-stop distances in column (9).  It is 4.0 miles in this case.

(b)   If your sampling plan is based on the PPMT option, you should also process the data as follows immediately to calculate the ratio of PMT to PPMT for each service unit:

(1)   Calculate the average route length for each route.

(2)   Calculate PPMT by multiplying UPT by average route length.

(3)   Divide PMT by PPMT to get the ratio for each service unit.

(4)   Follow Subsection 85 on calculating average route length and PPMT.

Table 67.03.  Calculation of Loads and PMT

Load Type

Stop Sequence

Distance to Next Stop

No. of Pass. Boarded

No. of Pass. Alighted

No. of Pass. from Previous Trip

Calculated Load

PMT

Leaving Load

1

0.3

20

0

2

20

6.0

2

0.7

2

1

 

21

14.7

3

0.6

0

2

 

19

11.4

4

0.3

1

3

 

17

5.1

5

0.5

1

10

 

8

4.0

6

0.8

0

2

 

6

4.8

7

0.2

0

1

 

5

1.0

8

0.1

0

2

 

3

0.3

9

0.1

0

2

 

1

0.1

10

0.3

0

0

 

1

0.3

11

0.1

0

0

 

1

0.1

12

0.0

0

1

 

0

0.0

Total

4.0

24

24

2

N/A

47.8

Load for Stop 1 =

Current Boarding

 

Load for Other Stops =

Previous Load + Current Boarding – Current Alighting

Arriving Load

Stop Sequence

Distance from Previous Stop

No. of Pass. Boarded

No. of Pass. Alighted

No. of Pass. from Previous Trip

Calculated Load

PMT

1

0.0

20

0

2

0

0.0

2

0.3

2

1

 

20

6.0

3

0.7

0

2

 

21

14.7

4

0.6

1

3

 

19

11.4

5

0.3

1

10

 

17

5.1

6

0.5

0

2

 

8

4.0

7

0.8

0

1

 

6

4.8

8

0.2

0

2

 

5

1.0

9

0.1

0

2

 

3

0.3

10

0.1

0

0

 

1

0.1

11

0.3

0

0

 

1

0.3

12

0.1

0

1

 

1

0.1

Total

4.0

24

24

2

N/A

47.8

Load for Stop 1 =

0

 

 

 

 

Load for Other Stops =

Previous Load + Previous Boarding – Previous Alighting

§ 67.09 – How may I use the processed data to identify potential errors in the sample data?

(a)    Make sure that UPT and PMT are consistent for each service unit in the sample.  For example, they should both be zero or both be positive. 

(b)   Compare vehicle trip length with the longest actual length of the route.  Vehicle trip length must not exceed the longest route length.

(c)    Compare APTL with vehicle trip length and route length.  APTL must not exceed either.

(d)   Compare the total number of passengers boarded with the total number of passengers alighted.  They must be equal.

(e)    Examine the calculated load at the end point of the trip.  It must be zero for leaving loads.

(f)    If calculated, examine the PMT to PPMT ratio for each service unit.  It must not exceed 1.

§ 67.11 – How do I identify the sources of any errors?

(a)    If you have identified any error from the previous step, you should start with the calculations you did in processing the sample data to determine if an error is a calculation error or an error in the sample data.

(b)   Check if you have correctly calculated loads, UPT, PMT, and APTL.

(c)    Check if you have used between-stop distances and loads consistently:

(1)   If you use leaving loads, you must use the distance to the next stop.  The distance should be zero for the ending point of a one-way trip.

(2)   If you use arriving loads, you must use the distance from the previous stop.  The distance should be zero for the beginning point of a trip. 

(d)   Check the predetermined between-stop distances you entered.

(e)    Compare your calculated load with the observed load from the field.

§ 67.13 – How do I correct any errors?

(a)    Table 67.05 shows an example of sample data for a one-way trip where PMT is based on calculated leaving loads.  The first row of Table 67.07 shows the vehicle trip length, UPT, PMT, APTL, and the ratio of PMT/PPMT for this trip.

(b)   It has the following errors:

(1)   Vehicle trip length > route length.

(2)   APTL > route length.

(3)   Total boardings in column (10) < total alightings in column (11).

(4)   The load at the end stop < 0.

(c)    Error (b)(1) clearly indicates errors in the predetermined between-stop distances.  If you check what you have entered in column (9), you would notice the 7 miles from stop 3 to stop 4.  Correct the data-entry error by replacing 7 by 0.7.  After this change,

(1)   vehicle trip length matches the route length,

(2)   APTL becomes smaller than the route length, and

(3)   PMT is reduced to 34.7 with an APTL of 1.58 miles. 

(d)   To identify the source of error (3), compare the calculated loads with the observed loads.  You may notice that the calculated load is lower by 2 at the first stop.  It appears that the ride checker did not include the number of passengers from the previous trip in the number of boardings at the first stop.  With this correction,

(1)   Boardings become greater than alightings.

(2)   PMT is increased to 42.7 miles with an APTL of 1.78 miles.

Table 67.05.  Example of Correcting Data Errors

Stop Sequence

Distance to Next Stop

No. of Pass. Boarded

No. of Pass. Alighted

Observed Leaving Load

No. of Pass. from Previous Trip (Stop 1 only)

No. of Pass. Continuing to Next Trip (Last Stop only)

Calculated Leaving Load

PMT

(7)

(9)

(10)

(11)

(12)

(13)

(14)

(15)

(16)

1

0.0

18

0

20

2

 

18

0.0

2

0.3

2

1

21

 

 

19

5.7

3

7.0

0

2

19

 

 

17

119.0

4

0.6

1

3

17

 

 

15

9.0

5

0.3

1

9

8

 

 

7

2.1

6

0.5

0

2

6

 

 

5

2.5

7

0.8

0

1

5

 

 

4

3.2

8

0.2

0

2

3

 

 

2

0.4

9

0.1

0

2

1

 

 

0

0.0

10

0.1

0

0

1

 

 

0

0.0

11

0.3

0

0

1

 

 

0

0.0

12

0.1

0

1

1

 

1

-1

-0.1

Total

10.3

22

23

 

2

1

N/A

141.8

(e)    To identify why boardings are still larger than alightings, compare the recalculated loads with the observed loads again.  You may notice that the recalculated load starts deviating from the observed load at stop 5.  Since the calculated load is one passenger too big, increase the number of alighted passengers at stop 5 by 1 from 9 to 10.  With this correction,

(1)   Boardings equal alightings,

(2)   The leaving load at the end stop is 0, and

(3)   PMT is further reduced to 40.3 with an APTL of 1.68 miles.

(f)    You should also check the consistency between between-stop distances and the calculated loads.  Column (9) is based on the distance from the previous stop, but the calculated load is the leaving load.  Correcting this error leads to a final PMT of 47.8 with an APTL of 1.99 miles.

(g)   Table 67.07 also shows vehicle trip distance, UPT, PMT, APTL, and ratio PMT/PPMT for the cumulative corrections in (c), (d), (e), and (f).      

Table 67.07.  Impacts of Correcting Errors

Error Corrections

Vehicle Trip Length

PMT

UPT

APTL

PMT/PPMT

No corrections

10.3

141.8

22

6.45

0.63

§69.13 (c)

4.0

34.7

22

1.58

0.39

§69.13 (c) + (d)

4.0

42.7

24

1.78

0.44

§69.13 (c) + (d) + (e)

4.0

40.3

24

1.68

0.42

§69.13 (c) + (d) + (e) + (f)

4.0

47.8

24

1.99

0.50

§ 67.15 – What should I do if I fail to correct the data errors for a particular service unit?

(a)    You should treat the service unit as if it were missed for data collection.

(b)   You should follow the guidance in §65.29 for getting a replacement unit.

§ 67.17 – What steps should I take after I have identified and corrected errors in my data?

(a)    You should stack the corrected sample data from the field as shown in Table 67.01 in one or more worksheets.  These worksheets along with the survey sheets become an auditable record of your sample data.

(b)   You should enter the summary data at the level of your unit of sampling and measurement in a separate worksheet.  These summary data are ready for developing new sampling plans or for estimating service-consumed data for the NTD.

(c)    Table 67.09 shows how that summary worksheet looks with the column headings and the summary sample data for one service unit.

Table 67.09.  Example of Summary Sample Data

Date

Day of Week

Time Period

Route No.

Trip No.

Direction

Vehicle Trip Length

UPT

PMT

(1)

(2)

(3)

(4)

(5)

(6)

 

 

 

10/13/05

Thur

Midday

11

408

Outbound

4.0

24

47.8

§ 71.01 – What service-consumed data must I estimate without 100% UPT?

(a)    You must estimate both UPT and PMT.

(b)   The following table shows the data items you must estimate.

Table 71.01.  Service-Consumed Data Items without 100% UPT

If your mode is

you must estimate

demand response – taxi (DT)

·  annual total UPT and PMT

commuter rail (CR), heavy rail (HR), or light rail (CR)

·  annual total UPT and PMT

·  average daily UPT and PMT by type of service days

·  annual total UPT by weekday time period

any other mode

·  annual total UPT and PMT

·  average daily UPT and PMT by type of service days

§ 71.03 – What do I need to do in general to get estimates of these data items?

(a)    You must determine sample averages.

(b)   You must determine expansion factors.

(c)    You must combine the expansion factors and sample averages to get the corresponding estimates of service-consumed data.

§ 71.05 – What is a sample average?

(a)    A sample average is the sample total divided by the number of service units in the sample. 

(b)   It may be calculated for the entire annual sample, by the type of service day, or for specific weekday time periods. 

(c)    Sample averages are used to estimate service-consumed data when your sampling plan is based on the base option.

§ 71.09 – How is the guidance organized?

(a)    The guidance is separate in three subsections for three modal groups to reduce confusion over the different units of sampling and measurement that are typically used for these three modal groups:

(1)   Non-scheduled services, including demand response (DR and DT), vanpool (VP), jitney (JT), or público (PB) (Subsection 73).

(2)   Rail services, including heavy rail (HR), commuter rail (CR), light rail (LR), monorail and automated guideway (MG) (Subsection 75).

(3)   Bus services, including bus (MB), commuter bus (CB), bus rapid transit (RB), or trolley bus (TB) (Subsection 77).

(b)   Subsection 79 contains the guidance on combining expansion factors and sample averages to get estimates of service-consumed data.

§ 73.01 – What expansion factor should I use for non-scheduled service?

(a)    You should use your 100% count of vehicle days as the expansion factor.

(b)   For estimating annual total PMT and UPT:

(1)   Use annual total count of vehicle days if your sampling plan is not based on grouping.

(2)   Use annual total count of vehicle days by group if your sampling plan is based on grouping.

(c)    For estimating average daily PMT and UPT by type of service days, use annual total count of vehicle days by type of service days.

§ 73.03 – What sample average should I use for non-scheduled service?

(a)    You must use the ratio of sample total PMT over sample total vehicle days as the sample average to estimate PMT.

(b)   You must use the ratio of sample total UPT over sample total vehicle days as the sample average to estimate UPT.

(c)    For estimating annual total PMT and UPT:

(1)   Use the sample averages for the entire sample if you do not group your service.

(2)   Use the sample averages for each group if your sampling plan is based on grouping.

(d)   For estimating average daily PMT and UPT by type of service days, use the sample averages by type of service days.  If your sample happens not to include any vehicle-days for a particular day type (e.g. Saturday), use the sample average for the entire sample. 

§ 73.05 – How do I determine annual vehicle days actually operated for non-scheduled services?

(a)    You should use a spreadsheet or some other mechanism to record the daily number of passengers carried by individual vehicles in your fleet for an entire year. 

(b)   Figure 73.01 shows an example of such a spreadsheet with the following assumptions:

(1)   your fiscal year runs from July 1 through June 30,

(2)   you operate every day, and

(3)   your fleet has 100 vehicles. 

(c)    Once you have such a spreadsheet, you can easily determine the daily number of vehicles operated by type of service days in two steps:

(1)   For any one operating date, count the number of vehicles in the fleet that carried at least one passenger on that date.

(2)   Summarize the daily number of vehicles operated by type of service days.

(d)    You can also use such a spreadsheet to determine the number of vehicles operated for each group if your sampling plan is based on service grouping.

 

B

C

D

E

F

G

H

I

J

···

CZ

DA

1

Date

Month

Day of Week

Daily Passengers Carried by Vehicle No.

2

1

2

3

4

5

6

···

100

Total

3

7/1/09

July

Wed

 

 

 

 

 

 

 

 

 

4

7/2/09

July

Thu

 

 

 

 

 

 

 

 

 

5

7/3/09

July

Fri

 

 

 

 

 

 

 

 

 

6

7/4/09

July

Sat

 

 

 

 

 

 

 

 

 

···

···

···

···

 

 

 

 

 

 

 

 

 

367

6/30/10

June

Wed

 

 

 

 

 

 

 

 

 

368

Total

 

 

 

 

 

 

 

 

 

Figure 73.01.  Daily Passenger Count Worksheet for Non-Scheduled Services

§ 75.01 – What expansion factor should I use for rail services?

(a)    The unit of sampling and measurement you have chosen for your sampling plan determines the expansion factor you should use.

(b)   While you could have chosen any unit of sampling measurement for your sampling plan, it most likely is one of the following:

(1)   One-way car trips.

(2)   One-way train trips.

(3)   Round-trip car trips.

(4)   Round-trip train trips.

(c)    The following assumes that your sampling plan is based on one-way car trips.

(d)   For estimating annual total PMT and UPT:

(1)   Use annual total one-way car trips if your sampling plan is not based on grouping.

(2)   Use annual total one-way car trips by group if your sampling plan is based on grouping.

(e)    For estimating average daily PMT and UPT by type of service days, use annual total one-way car trips by type of service days.

(f)    For estimating annual total UPT for heavy rail (HR), commuter rail (CR), light rail (LR) by weekday time period, use annual total one-way car trips by weekday time period. 

§ 75.03 – What sample average should I use for rail services?

(a)    The unit of sampling and measurement you have chosen for your sampling plan determines the sample average you should use. 

(b)   While you could have chosen any unit of sampling measurement for your sampling plan, it most likely is one of the following:

(1)   One-way car trips.

(2)   One-way train trips.

(3)   Round-trip car trips.

(4)   Round-trip train trips.

(c)    Assuming that your unit of sampling and measurement is in one-way car trips, Table 75.01 shows the sample averages you should use.

(d)   If your sample happens not to include any one-way car trips for a particular weekday time period, use UPT per one-way car trip for weekdays to estimate the annual total UPT for this weekday time period. 

Table 75.01.  Sample Averages for Rail Services

If the service-consumed measure is

and if the measure is

and if your sampling plan is

and if your mode is

then you should calculate the following sample average

UPT

for annual total

not based on grouping

 

UPT/one-way car trip

based on grouping

 

UPT/one-way car trip by group

for daily average

 

 

UPT/one-way car trip by type of service days

by weekday time period

 

commuter rail (CR), heavy rail (HR), or light rail (LR)

UPT/one-way car trip by weekday time period

PMT

for annual total

not based on grouping

 

PMT/one-way car trip

based on grouping

 

PMT/one-way car trip by group

for daily average

 

 

PMT/one-way car trip by type of service days

§ 75.05 – What steps should I follow to calculate sample averages?

(a)    Aggregate the field sample data to the individual service units in your sample after you have identified and corrected any errors in the data. 

(b)   If the unit of sampling and measurement for your sampling plan is one-way car trips, sum these quantities over individual one-way car trips in the sample to get sample totals:

(1)   for the entire sample,

(2)   by type of service days,

(3)   by weekday time period if your service is commuter rail (CR), heavy rail (HR), or light rail (LR), and

(4)   by group if your sampling plan is based on service grouping.

(c)    Count the number of one-way car trips in the sample:

(1)   for the entire sample,

(2)   by type of service days,

(3)   by weekday time period if your service is commuter rail (CR), heavy rail (HR), or light rail (LR), and

(4)   by group if your sampling plan is based on service grouping.

(d)   Divide sample totals by the number of one-way car trips for the entire sample, by type of service days, by weekday time period if applicable, or by service group if applicable.

§ 75.07 – How should I determine annual services actually provided as expansion factors?

(a)    You should measure annual services actually provided in the unit of sampling and measurement you have chosen for your sampling plan.

(b)   For estimating annual totals of service-consumed data, you should follow these steps:

(1)   Start with your schedule.

(2)   When the schedule is changed by policy or by emergency conditions, the count from the schedule must be adjusted accordingly.

(3)   You must also include added services such as trippers and other special operations.

(4)   If your sampling plan is based on service grouping, you must also count annual services actually provided for each group.

(c)    For estimating service-consumed data by type of service days, you should also start your schedule, and exclude scheduled services on atypical days.

(d)   For estimating annual total UPT for commuter rail (CR), heavy rail (HR), or light rail (LR) by weekday time period, you should start with your weekday schedule, and exclude scheduled services on atypical weekdays.

(e)    Refer to the NTD Reporting Manual on determining what are considered atypical days.

§ 77.01 – What expansion factor should I use for bus services?

(a)    The unit of sampling and measurement you have chosen for your sampling plan determines the expansion factor you should use.

(b)   While you could have chosen any unit of sampling measurement for your sampling plan, it most likely is in one-way bus trips or in round-trip bus trips.

(c)    The following assumes that your sampling plan is based on one-way bus trips.

(d)   For estimating annual total PMT and UPT:

(1)   Use annual total one-way bus trips if your sampling plan is not based on grouping.

(2)   Use annual total one-way bus trips by group if your sampling plan is based on grouping.

(e)    For estimating average daily PMT and UPT by type of service days, use annual total one-way bus trips by type of service days.

§ 77.03 – What sample average should I use for bus services?

(a)    The unit of sampling and measurement you have chosen for your sampling plan determines the sample average you should use. 

(b)   While you could have chosen any unit of sampling measurement for your sampling plan, it most likely is in one-way bus trips or in round-trip bus trips.

(c)    Assuming that your unit of sampling and measurement is in one-way bus trips, Table 77.01 shows the sample averages you should use.

(d)   If your sample happens not to include any one-way bus trips for a particular day type (e.g., Saturday), use the sample average for the entire sample. 

Table 77.01.  Sample Averages for Bus Services

If the service-consumed measure is

and if the measure is

and if your sampling plan is

then you should calculate the following sample average

UPT

for annual total

not based on grouping

UPT/one-way car trip

based on grouping

UPT/one-way car trip by group

for daily average

 

UPT/one-way car trip by type of service days

PMT

for annual total

not based on grouping

PMT/one-way car trip

based on grouping

PMT/one-way car trip by group

for daily average

 

PMT/one-way car trip by type of service days

§ 77.05 – What steps should I follow to calculate sample averages?

(a)    Aggregate the field sample data to the individual service units in your sample after you have identified and corrected any errors in the data. 

(b)   If the unit of sampling and measurement for your sampling plan is one-way bus trips, sum these quantities over individual one-way bus trips in the sample to get sample totals:

(1)   for the entire sample,

(2)   by type of service days,

(3)   by group if your sampling plan is based on service grouping.

(c)    Count the number of one-way bus trips in the sample:

(1)   for the entire sample,

(2)   by type of service days,

(3)   by group if your sampling plan is based on service grouping.

(d)   Divide sample totals by the number of one-way bus trips for the entire sample, by type of service days, or by service group if applicable.

§ 77.07 – How should I determine annual services actually provided as expansion factors?

(a)    You should measure annual services actually provided in the unit of sampling and measurement you have chosen for your sampling plan.

(b)   For estimating annual totals of service-consumed data, you should follow these steps:

(1)   Start with your schedule.

(2)   When the schedule is changed by policy or by emergency conditions, the count from the schedule must be adjusted accordingly.

(3)   You must also include added services such as trippers and other special operations.

(4)   If your sampling plan is based on service grouping, you must also count annual services actually provided for each group.

(c)    For estimating service-consumed data by type of service days, you should also start with your schedule, and exclude scheduled services on atypical days.  Refer to the NTD Reporting Manual for determining what are considered atypical days.

§ 79.01 – How should I estimate the annual total of my service-consumed data?

(a)    If your sampling plan is not based on service grouping, you should multiply your sample average for the entire sample with the annual total of your corresponding expansion factor to get an estimate of the annual total.

(b)   If your sampling plan is based on service grouping, you should take a two-step approach:

(1)   Multiply your sample average with your corresponding expansion factor for each group, and

(2)   Sum the above products across all groups to get your annual total.

§ 79.03 – How should I estimate average daily of service-consumed data by type of service days?

(a)    If you operate your service only on weekdays, all you need to do is to divide your estimated annual total by the number of typical weekday days of service.

(b)   If you operate your service on Saturdays, Sundays, or both as well, you should take the following two steps:

(1)   Multiply your sample average(s) for each type of service days with the corresponding expansion factor to get the annual total for each type of service days.

(2)   Divide the annual total for each type of service days by the corresponding number of typical days for each type of service days.

§ 79.05 – How should I estimate annual total UPT for each weekday period for commuter rail, heavy rail, and light rail?

(a)    For each of the weekday periods, you should simply multiply the sample average UPT you have calculated for that period by the corresponding expansion factor to get an estimate of the annual total UPT for that weekday period.

§ 81.01 – What service-consumed data must I estimate with 100% UPT?

(a)    You must estimate PMT.

(b)   The following table shows the data items you must estimate.

Table 81.01.  Service-Consumed Data Items with 100% UPT

If your sampling plan is based on the

you must estimate

APTL option

·  annual total PMT

·  average daily PMT by type of service days

PPMT option

·  annual total PMT

·  average daily PMT by type of service days

§ 81.03 – What do I need to do in general to get estimates of these data items?

(a)    You must determine expansion factors.

(b)   You must determine sample ratios.

(c)    You must combine the expansion factors and sample ratios to get the corresponding estimates of service-consumed data.

§ 81.05 – What is a sample ratio?

(a)    A sample ratio is the ratio between the sample total of one measure of service-consumed and the sample total of another measure of service-consumed.  This Sampling Manual uses two sample ratios:

(1)   The ratio of sample total PMT over sample total UPT gives the sample APTL.

(2)   The ratio of sample total PMT over sample total PPMT gives the sample PMT/PPMT ratio.

(b)   They may be calculated for the entire annual sample or by the type of service days.

§ 81.07 – What is an expansion factor when I have 100% UPT?

(a)    A measure of actual services consumed during a given duration of time. 

(b)   It is used to convert sample ratios to totals in the given duration.  A 100% count of UPT consumed during an entire year is an example of an expansion factor in service consumed; when multiplied by sample APTL, it yields a measure of annual total PMT.

§ 83.01 – What expansion factor should I use for the APTL option?

(a)    You must use your 100% count of UPT as the expansion factor.

(b)   For estimating average daily PMT by type of service days, use your annual total 100% count of UPT by type of service days.

(c)    For estimating annual total PMT:

(1)   Use your annual total 100% count of UPT if your sampling plan is not based on grouping.

(2)   Use your annual total 100% count of UPT by service group if your sampling plan is based on grouping and you have reliable 100% counts of UPT by service group and have entered information into the template accordingly in developing your current template sampling plan.  Refer to cell M52 of the PeriodInput Worksheet in the template you used in developing your current period-based template sampling plan, and it should be 1.

(3)   Use your annual total 100% count of UPT if your sampling plan is based on grouping but you do not have reliable 100% counts of UPT by service group and you have entered information into the template accordingly in developing your current period-based template sampling plan.  Refer to cell M52 of the PeriodInput Worksheet in the template you used in developing your current period-based template sampling plan, and it should be 0.

§ 83.03 – What sample ratio should I use for the APTL option?

(a)    You must use the sample APTL as the sample ratio.

(b)   Use sample APTL by type of service days for estimating average daily PMT by type of service days.  If your sample happens not to include service unit for a particular day type (e.g., Saturday), use the sample APTL for the entire sample for this day type. 

(c)    For estimating annual total PMT:

(1)   Use sample APTL for the entire sample if you do not group your service.

(2)   Use sample APTL for each group if your sampling plan is based on grouping and you have reliable 100% counts of UPT by service group and have entered information into the template accordingly in developing your current period-based template sampling plan.  Refer to cell M52 of the PeriodInput Worksheet in the template you used in developing your current period-based template sampling plan, and it should be 1.

(3)   Use weighted sample APTL for the entire sample if your sampling plan is based on grouping but you do not have reliable 100% counts of UPT by service group and you have entered information into the template accordingly in developing your current period-based template sampling plan.  Refer to cell M52 of the PeriodInput Worksheet in the template you used in developing your current period-based template sampling plan, and it should be 0.

§ 83.05 – How should I determine the APTL from my sample?

(a)    You must determine the sample APTL for a given sample as the ratio of sample total PMT over sample total UPT for the following cases:

(1)   for the entire sample,

(2)   by type of service days, or

(3)   by service group if applicable.

(b)   You must not determine the sample APTL as the average of the APTL across individual service units in the sample.

(c)    To determine the weighted sample APTL for an entire sample as required by question §83.03(c)(3), you should follow these steps:

(1)   Determine each group’s size in the number of service units actual operated.

(2)   Compute each group’s share of the number of service units actually operated.  These shares must sum to 1.

(3)   Determine each group’s sample size in the number of service units in the sample.

(4)   Determine each group’s sample total UPT.

(5)   Determine each group’s sample total PMT.

(6)   Compute each group’s sample average for UPT.

(7)   Compute each group’s sample average for PMT.

(8)   Sum the product of each group’s share of the number of service units actually operated and its sample average UPT.

(9)   Sum the product of each group’s share of the number of service units actually operated and its sample average PMT.

(10)           Compute the ratio of the result from (9) to the result from (8).  This ratio gives the weighted sample APTL.

(11)           Table 83.01 shows these steps with an example.

Table 83.01.  Example of Estimating Weighted Sample APTL

Step

 Description

Service Groups

Short Routes

Medium Routes

Long Routes

1

Group size in service units

109,685

331,033

35,325

2

Group size in shares

0.2304

0.6954

0.0742

3

Sample size

116

386

47

4

Sample total UPT

1,157

8,181

1,592

5

Sample total PMT

3,989

42,966

7,003

6

Sample average UPT

10.0

21.2

33.9

7

Sample average PMT

34.4

111.3

149.0

8

Weighted sample average UPT

19.55

9

Weighted sample average PMT

96.38

10

Weighted sample APTL

4.93

§ 83.07 – How should I estimate annual total PMT for the APTL option?

(a)    If your sampling plan is not based on service grouping, you should multiply your sample APTL for the entire annual sample with your corresponding annual expansion factor (i.e., 100% count of annual UPT) to get an estimate of the annual total PMT.

(b)   You should take a two-step approach if your sampling plan is based on service grouping and you have reliable 100% counts of UPT by service group:

(1)   Multiply your sample APTL with your corresponding expansion factor for each group, and

(2)   Sum the above products across all groups to get your annual total PMT.

(c)    If your sampling plan is based on service grouping but you do not have reliable 100% counts of UPT by service group, you should multiple your weighted sample APTL as computed in question §83.05 with your 100% count of annual UPT.

(d)   If your sampling plan is based on service grouping, refer to cell M52 of the PeriodInput Worksheet in the template you used in developing your template sampling plan for what you have entered into cell M52 on whether you have reliable 100% count of UPT by service group.

§ 83.09 – How should I estimate average daily PMT by type of service days?

(a)    If you operate your service only on weekdays, all you need to do is to divide your estimated annual total PMT by the number of typical weekday days of service.

(b)   If you operate your service on Saturdays, Sundays, or both as well, you should take the following two steps:

(1)   Multiply your sample average APTL for each type of service days with the corresponding 100% count of UPT to get the annual total PMT for each type of service days.

(2)   Divide the estimated annual total PMT for each type of service days by the corresponding number of typical days for each type of service days.

§ 85.01 – What expansion factor should I use for the PPMT option?

(a)    You must use your 100% count of PPMT as the expansion factor.

(b)   For estimating annual total PMT:

(1)   Use annual total PPMT if your sampling plan is not based on grouping.

(2)   Use annual total PPMT by group if your sampling plan is based on grouping.

(c)    For estimating average daily PMT by type of service days, use annual total PPMT by type of service days.

§ 85.03 – How do I determine annual total PPMT?

(a)    Suppose that:

(1)   Your unit of sampling and measurement is one-way bus trips, and

(2)   You are going to estimate annual total PMT.

(b)   You should do the following for each route:

(1)   Determine the annual number of vehicle revenue miles for the report year.

(2)   Determine the annual number of vehicle revenue one-way trips for the report year.

(3)   Divide the annual number of vehicle revenue miles by the annual number of vehicle revenue one-way trips to get the average route length.

(4)   Multiply your route-level 100% counts of UPT with the calculated average route length to get route-level PPMT.

(c)    You should sum the calculated route-level PPMT across all routes to get your annual PPMT for all routes.

(d)   Table 85.01 illustrates how you may accomplish (b) and (c).

Table 85.01.  Calculating Annual Total PPMT for All Operating Routes

Route Number

Route Name

Annual Revenue Trips

Annual Revenue Miles

Average

Route Length

100% UPT

100% PPMT

90

Blue Line

3,869

9,975

2.58

22,866

58,952

50

Red Line

3,286

10,310

3.14

23,634

74,148

14

Prospect

1,643

10,690

6.51

24,506

159,446

12

Beechcrest

1,643

11,835

7.20

27,131

195,435

17

College

3,286

30,666

9.33

70,298

656,036

37

Park 100

1,325

22,733

17.16

52,112

894,068

8

Washington

3,392

61,077

18.01

140,012

2,521,072

19

Castleton

1,696

32,916

19.41

75,457

1,464,491

26

Keystone

1,378

28,505

20.69

65,344

1,351,666

10

10th St.

3,339

69,897

20.93

160,231

3,354,198

Total

10,729,514

(e)    If your sampling plan is based on service grouping, you must also determine PPMT for each group.  Suppose, for example, that you have grouped your short routes into one group and your longer routes into another.  Table 85.03 illustrates how you may use route-level information on annual revenue miles, annual revenue trips, and annual UPT to determine PPMT for each group.

Table 85.03.  Calculating Annual Total PPMT by Route Group

Route Group

Route Number

Annual Revenue Trips

Annual Revenue Miles

Average

Route Length

100% UPT

100% PPMT

Short Routes

90

3,869

9,975

2.58

22,866

58,952

50

3,286

10,310

3.14

23,634

74,148

14

1,643

10,690

6.51

24,506

159,446

12

1,643

11,835

7.20

27,131

195,435

17

3,286

30,666

9.33

70,298

656,036

Total

1,144,018

Long Routes

37

1,325

22,733

17.16

52,112

894,068

8

3,392

61,077

18.01

140,012

2,521,072

19

1,696

32,916

19.41

75,457

1,464,491

26

1,378

28,505

20.69

65,344

1,351,666

10

3,339

69,897

20.93

160,231

3,354,198

Total

9,585,496

§ 85.05 – How should I get the sample total of PPMT for each service unit?

(a)    You should follow the guidance for question § 85.03 to determine the average route length for each route.

(b)   Identify the route for each one-way trip in your service units in the sample.

(c)    Multiply the sample UPT by the average route length for each one-way trip to get sample PPMT for each one-way trip.

(d)   Sum the trip-level PPMT for all one-way trips in a service unit to get sample PPMT for each service unit. 

§ 85.07 – What sample ratio should I use for the PPMT option?

(a)    You must use the ratio of sample total PMT over sample total PPMT as the sample ratio.

(b)   For estimating annual total PMT:

(1)   Use sample PMT/PPMT ratio for the entire sample if you do not group your service.

(2)   Use sample PMT/PPMT for each group if your sampling plan is based on grouping.

(c)    For estimating average daily PMT by type of service days, use the sample PMT/PPMT ratio by type of service days.

§ 85.09 – How should I determine the PMT/PPMT ratio for a sample?

(a)    You must determine the sample PMT/PPMT ratio for a given sample as the ratio of sample total PMT over sample total PPMT.

(b)   You must not determine the sample ratio as the average of the PMT/PPMT ratio for individual service units in the sample.

§ 85.11 – How should I estimate annual total PMT for the PPMT option?

(a)    You should multiply your sample PMT/PPMT ratio for the entire annual sample with the annual total of your PPMT to get an estimate of the annual total PMT if your  sampling plan is not based on service grouping, or

(b)   You should take a two-step approach if your sampling plan is based on service grouping:

(1)   Multiply your sample PMT/PPMT ratio with your corresponding annual total PPMT for each group, and

(2)   Sum the above products across all groups to get your annual total PMT.

§ 85.13 – How should I estimate average daily PMT by type of service days?

(a)    If you operate your service only on weekdays, all you need to do is to divide your estimated annual total PMT by the number of typical weekday days of service.

(b)   If you operate your service on Saturdays, Sundays, or both as well, you should take the following two steps:

(1)   Multiply your sample PMT/PPMT ratio for each type of service days with the corresponding annual total PPMT to get the annual total PMT for each type of service days.

(2)   Divide the annual total PMT for each type of service days by the corresponding number of typical days for each type of service days.

§ 91.01 What do the sampling-related terms mean?

Alternative sampling plan.  A sampling plan that reflects the conditions of your service, and is independently developed and certified by a qualified statistician to meet FTA’s 95% confidence and 10% precision levels.  It is one of two forms of customized sampling plans.  The other form is template sampling plans.  It is equivalent to an alternative sampling technique as defined in the NTD Reporting Manual.

Confidence level.  The chance of an estimate of service-consumed data obtained through random sampling falling within a particular range of the true value. FTA requires a minimum level of 95% confidence for estimates of annual UPT and annual PMT reported to the NTD.  A particular confidence level is only meaningful when it is stated with a particular precision level.

Customized sampling plan.  A sampling plan that reflects the conditions of your service and meets FTA’s 95% confidence and 10% precision levels.  It is either a template sampling plan or an alternative sampling plan.  It differs from a ready-to-use sampling plan in that it takes account of the specific characteristics of your service.

Efficiency option.  A characteristic of a sampling plan that affects its sampling efficiency.

Initial annual sample size.  The annual necessary sample size of a period-based template sampling plan that is determined from the companion spreadsheet template before it is allocated to each quarter, month, or week.  This can differ from the realized annual sample size for a given set of sample data.

Mandatory revising year.  A report year for which you must consider whether you need to revise your template sampling plan. 

Margin of safety.  A percent increase in the statistical variation of your sample data in developing a template sampling plan.  For example, if the statistical variation of your sample is S and the margin of safety is 25%, you must use 1.25S as the statistical variation in developing your template sampling plan.  A margin of safety of 25% is used automatically for all period-based template sampling plans.  If you develop alternative sampling plans, you should also use this margin of safety.  The objective is to counter the potential fluctuations in the statistical variation in a sample from one year to another due to sampling and other reasons. 

Major change to a service.  Any change to your service that is likely to lead to major changes in how your customers use your service.  Examples of major changes include making transfers fare free; adding or cutting express routes; expanding or contracting your service by more than 25% in vehicle revenue miles; or restructuring your service affecting more than 25% of your service in vehicle revenue miles.

Necessary sample size.  The sample size that meets FTA’s minimum 95% confidence and 10% precision levels and uses a 25% margin of safety.

Precision index.  A number that reflects the level of precision that your current NTD sample achieves in the resulting annual total PMT.  It is used for you to determine whether you must revise your current template sampling plan after you have made major changes to your service since you started using the current template sampling plan.  Once you have entered your current sample data into the template for this Sampling Manual as if you are going to develop new template sampling plans, this precision index is made available in the PeriodPrecision Worksheet for period- based sampling and in the IntervalPrecision Worksheet for interval-based sampling. 

Precision level.  The degree of errors in an estimate of service-consumed data obtained through random sampling that is stated in percentage terms relative to the true value.  FTA requires a minimum of 10% precision for estimates of annual service-consumed data reported to the NTD.  A particular precision level is only meaningful if it is stated with a particular confidence level.

Random sampling.  Selection of one or more service units at random from a list of service units to be operated.

Ready-to-use sampling plan.  A sampling plan that has been developed specifically for this Sampling Manual with sample data from a variety of transit agencies.  It does not necessarily reflect the conditions of your service.  Ready-to-use sampling plans have limited applicability.

Realized annual sample size.  The annual necessary sample size of a period-based template sampling plan that is based on quarterly, monthly, or weekly sampling.  For example, if you choose weekly sampling and your template sampling plan requires 3 one-way trips per week, the realized annual sample size would be 156 one-way trips.

Sample size.  The number of service units that are sampled, and for which unlinked passenger trips and passenger miles traveled are measured.

Sampling efficiency.  The degree to which a sampling plan minimizes the necessary sample size for meeting FTA’s confidence and precision levels.  Sampling plans that take advantage of certain characteristics of your service can sometimes require a smaller necessary sample size.  A smaller necessary sample size reduces the time and cost of sampling, data collection, and data processing.

Sampling frequency.  The number of times per year that a sample is drawn; in this Manual, sampling frequency is quarterly, monthly, or weekly for period-based sampling, and is every day, every 2nd day, every 3rd day, every 4th day, every 5th day, and every 6th day for interval-based sampling.  For example, if your period-based sampling plan requires 10 service units per month, before the current month ends you must select at least 10 at random from the full list of all service units to be operated during the next month.

Sampling plan.  A plan for selecting service units at random, for collecting sample data, and for estimating annual service-consumed data that meets FTA’s 95% confidence and 10% precision levels.  Each sampling plan consists of four elements: a unit of sampling and measurement, a set of efficiency options, a sampling frequency, and a necessary sample size.

Sampling without replacement.  Selection of a sample of service units at random without the chance of a single service unit being selected more than once.

Sampling year.  Any report year for which you obtained annual UPT, annual PMT, or both through random sampling that meet FTA’s 95% confidence and 10% precision levels.  It can be a mandatory sampling year or an intermediate report year for which you choose to sample.

Service grouping.  One efficiency option for which you divide your service into two or more groups with the objectives of reducing within-group differences and increasing between-group differences.  For example, separating your bus routes into express routes and local routes is likely to reduce differences in average passenger trip length across one-way bus trips within each group.

Service unit.  An amount of revenue travel by a single transit vehicle, a set of transit vehicles, or a component of a transit vehicle.  For non-scheduled services, it is typically one vehicle day.  For scheduled bus services it is typically either a one-way bus run or else a round-trip bus run.  For rail services, it is either a one-way car run, a one-way train run, or a round-trip car run or a round-trip train run.

Statistical variation.  The degree of differences in a quantity across the full list of service units operated during a given duration of time, such as differences in PMT across all one-way trips of a bus service in a full report year.  A larger variation requires a greater sample size to meet given confidence and precision levels.

Table of random numbers.  A list of integers whose frequency and order of appearance in the list have been determined entirely by chance.  It is the basis of a commonly used method of random sampling.

Template sampling plan.  A sampling plan that is developed with the companion template of this Sampling Manual.

Unit of sampling and measurement.  A service unit you choose for your sampling plan.

§ 91.03 What do the data-collection terms mean?

Arriving load.  The number of passengers onboard a transit vehicle as it arrives at a stop.

Automatic passenger counter.  An automated means of counting passengers as they board or alight transit vehicles with treadle mats, infrared beams, or other devices placed by the doors of a transit vehicle.

Calculated load.  The number of passengers onboard a transit vehicle as it arrives at or leaves a stop that you calculate from data on boardings and alightings at individual stops.  It should equal the observed load.

Commuter vanpool.  A common form of vanpool service (VP) that comprises vans, small buses and other vehicles operating as a ridesharing arrangement, providing transportation to a group of workers commuting directly between their homes and their regular work sites within the same geographical area.  The vehicles would not be in revenue service during the working hours of the participating workers.

Distance-based approach.  A method to obtain PMT that keeps track of the distance traveled by every passenger.    

Leaving load.  The number of passengers onboard a transit vehicle as it leaves a stop.

Load-based approach.  A method to obtain PMT that is based on the boardings and alightings at individual stops and on the distance between consecutive stops.

NTD sample.  The sample of service units you select at random according to your sampling plan that meets FTA’s 95% confidence and 10% precision levels for reporting to the NTD.

Observed load.  The number of passengers onboard a transit vehicle as observed directly by a ride-checker while onboard that transit vehicle.

Ride check.  A method of collecting sample data with one or more persons observing and recording passenger boarding and alighting activities while riding in a transit vehicle. 

Stop.  Any spatial location at which a transit vehicle allows passengers to board or alight from the vehicle. 

Vehicle trip length.  The total distance traveled by a transit vehicle during a one-way trip for scheduled services.  For example, the cumulative distance traveled from the beginning point to the end point of a particular alignment of a route is the vehicle trip length for this trip.  The vehicle trip length may vary by direction and alignment for a given route.

§ 91.05 What do the estimation-related terms mean?

100% count.  A method of obtaining service-consumed data.  This term also often refers to the results of a 100% count.  For UPT, it involves counting passengers each time they board a transit vehicle in revenue service, such as through a registering farebox.  For PMT, it involves recording the distance traveled by all passengers.  A 100% count of PMT is typically only possible for systems that have only two stops, for rail systems that record entry and exit from the system, or for rail systems that rely upon destination-based tickets. 

Average passenger trip length (APTL).  The average distance traveled for an unlinked passenger trip. It is calculated as PMT divided by UPT.

Average route length.  The average length of a route actually traveled by vehicles in scheduled services.   It is calculated by dividing the annual vehicle revenue miles by the number of annual vehicle revenue one-way trips for that route. 

Expansion factor.  A measure of actual services operated or consumed during a given duration of time.  It is used to convert a sample average to the total of service-consumed data for that duration.  It varies with sampling plans.  The total number of one-way bus trips operated during an entire report year is an example of an expansion factor; when multiplied by the sample average PMT per one-way bus trip derived from annual NTD sample, it yields a measure of annual total PMT.

Passenger miles traveled (PMT).  The total distance traveled by all passengers during a given period.

Potential passenger miles traveled.  The maximum number of passenger miles that could have been traveled by all passengers along a given fixed route during a year (or some other time duration.)  It is calculated by multiplying a 100% count of UPT times the average route length of that route during that duration.

Sample average.  The sample total divided by the number of service units in the sample.  It may be calculated for the entire annual sample, or by the type of service day, or for specific weekday time periods.  For example, dividing the total PMT in an annual NTD sample by the total number of one-way bus trips in the annual NTD sample gives a sample average PMT.

Sample data.  The data collected from a sample of service units according to a sampling plan that meets FTA’s 95% confidence and 10% precision levels.

Sample ratio.  The ratio of the sample total for one measure of service-consumed over the sample total for another measure of service-consumed.  For example, the ratio of the sample data for PMT over the sample total for UPT gives the sample APTL.  It may be calculated for the entire annual sample, or by the type of service days, or for individual service group if your sampling plan is based on service grouping.

Sample total.  The sum total of all data across the service units in a random sample.  For example, if you are sampling for PMT, the PMT sample total is the sum of the PMT collected for each of the one-way bus trip in the sample.  It may be calculated for the entire annual sample, or by the type of service day, or for specific weekday time periods.

Service-consumed data.  Passenger miles traveled and unlinked passenger trips.

Type of service days.  Weekdays, Saturdays, or Sundays.  For scheduled services, service days in a report year are classified according to the schedule operated on that day.  If a weekday that is a holiday is served with a Sunday schedule, that weekday is considered to be a Sunday.  For non-scheduled services, service days are the actual days of a week regardless of whether they are a holiday or not.

Typical day.  For your scheduled services, it is a day on which you operate your normal, regular schedule and there are no anomalies such as extra service added for a special event or reduced service as a result of weather or interruption.  For your non-scheduled services, it is any day of operation.

Unlinked passenger trips (UPT).  The number of passengers who board transit vehicles in revenue service.  Passengers are counted each time they board a vehicle, no matter how many vehicles they use to travel from their origin to their destination.

Weekday time periods.  Weekday AM Peak, Weekday Midday, Weekday PM Peak, and Weekday Other.  The Reporting Manual instructs how you should define the start and end points of each period.

§ 91.07 What do the reporting terms mean?

Auditable record.  Documentation of information collected and processes used in collecting that information that demonstrates your compliance with NTD requirements. Such documentation may also help quality control within your agency when your NTD staff change over time.

First-time reporting.  Reporting of a particular service to the NTD for the first time.

§ 91.09 What abbreviations for general terms are used in this Manual?

APTL.  Average passenger trip length in miles

APC.  Automatic passenger counter

FTA.  Federal Transit Administration

NTD.  National Transit Database

PMT.  Passenger miles traveled

PPMT.  Potential passenger miles traveled

UPT.  Unlinked passenger trips

§ 91.11 What abbreviations for NTD modes are used in this Manual?

CB.  Commuter bus – Fixed-route bus systems that are primarily connecting outlying areas with a central city. Service typically uses over-the-road buses with service predominantly in one direction during peak periods, limited stops, and routes of extended length.

CC.  Cable car – A railway propelled by moving cables located beneath the street. While popular at the turn of the last century, the only surviving system is operated in San Francisco.

CR.  Commuter rail – Rail service operating on either old freight railways, or on tracks that are shared with freight railways, Amtrak, or both. The service is characterized by relatively long distances between stops, for service primarily connecting a central city with outlying suburbs and cities. The service may be either diesel or electric-powered and usually has grade-crossings with roadways.

DR.  Demand response – Shared-ride demand response service is scheduled in response to calls from passengers. Many transit systems operate demand response (DR) service to meet the requirements of the Americans with Disabilities Act (ADA).

DT.  Demand response – Taxi – A special form of the demand response mode operated through taxicab providers. The mode is always purchased transportation type of service.

HR.  Heavy rail – An electric railway that operates local service in exclusive right-of-way. The service is characterized by long trains of six to eight cars or more and by relatively short distances between stops for local service within a city and the immediate suburbs. The Nation’s traditional subway systems are classified as heavy rail.

JT.  Jitney – A transit mode comprising of owner-operated passenger cars or vans operating on fixed routes (sometimes with minor deviations) as demand warrants without fixed schedules or fixed stops.

LR.  Light rail – An electric railway that operates local service in mixed traffic with road vehicles, or has grade crossings with roadways. The service is characterized by short trains of one to four cars and by relatively short distances between stops for local service within a city and the immediate suburbs.

MB.  Bus – Fixed-route bus service is the most-prevalent mode in the country. MB service is powered by a motor and fuel contained within a vehicle. Deviated fixed-route service is also reported as MB.

MG.  Monorail and automated guideway – An electric railway that straddles a single guideway. It may have vehicle operators or may use computers to guide the vehicles.

PB.  Público – Publicos are jitney services operated in Puerto Rico. A transit mode comprising of passenger vans or small buses operating with fixed routes but no fixed schedules. Publicos (PB) are a privately owned and operated public transit service which is market oriented and unsubsidized, but regulated through a public service commission, state or local government. Publicos (PB) are operated under franchise agreements, fares are regulated by route and there are special insurance requirements. Vehicle capacity varies from eight to 24, and the vehicles may be owned or leased by the operator.

RB.  Bus rapid transit – Fixed-route bus systems that either (1) operate their routes predominantly on fixed-guideways (other than on highway high occupancy vehicle (HOV) or shoulder lanes, such as for commuter bus service) or (2) that operate routes of high-frequency service with the following elements: Substantial transit stations, traffic signal priority or preemption, low-floor vehicles or level-platform boarding, and separate branding of the service. High-frequency service is defined as 10-minute peak and 15-minute off-peak headways for at least 14 hours of service operations per day. This mode may include portions of service that are fixed-guideway and non-fixed-guideway.

SR.  Streetcar rail – Rail systems operating routes predominantly on streets in mixed-traffic. This service typically operates with single-car trains powered by overhead catenaries and with frequent stops.

TB.  Trolleybus – Fixed-route service using rubber tire buses powered by electric current from overhead wires using trolley poles. Service using rubber tire replica trolleys or historic trolleys, powered by an on-board motor are not included in this mode.

TR.  Arial tramway – A system of aerial cables with suspended vehicles.

VP.  Vanpool – A commuting service operating under prearranged schedules for previously formed groups of riders in vans.

YR.  Hybrid rail – Rail systems primarily operating routes on the National system of railroads, but not operating with the characteristics of commuter rail. This service typically operates light rail-type vehicles as diesel multiple-unit trains (DMU’s).