§ 53.01 – What data must I have to use this template?

(a)    You must have a set of sample data collected recently from your service for the mode and type of service for which you would like to have a new template sampling plan. 

(1)   The sample data must be at the unit of sampling and measurement you have chosen for your sampling plan.  For example, if your sampling plan is in one-way bus trips, the sample data should show the UPT and PMT and other identification information for each one-way bus trip in your sample.

(2)   Follow the instructions in the template for details on data items and format required and how they should be entered into the template.

(b)   You should use your NTD sample if it is available and you have not made major changes to your service since the data were collected.

(c)    Otherwise, you may use data that have been collected for non-NTD purposes: 

(1)   One good source of non-NTD sample data for scheduled services is a comprehensive operational analysis (COA).  Typically, all trips or a large portion of all trips on your schedule are checked by ride-checkers or by APCs during a concentrated period.

(2)   Make sure that the non-NTD source of sample data has the required information for using this template.  For example, between-stop distances often are not collected during COA ride checks, and the resulting ride check data would not be useful if you do not have predetermined between-stop distances for your entire service.

(d)   The sample size of the sample data from any source must be 50 or more service units.

(e)    If you do not have a set of sample data recently collected from your service for the mode and type of service, then you should use one of the other types of sampling plans, such as ready-to-use sampling plans.

§ 53.03 – Must I identify and correct errors with my sample data before I enter them into the template?

(a)    Yes, you must identify and correct errors in your sample data for using the template.

(b)   If you are using sample data from previous report years, you should have identified and corrected such errors in developing your estimates then; make sure that you use the final version of your sample data. 

(c)    Refer to Subsection 67 for identifying and correcting errors in your sampling data during the course of an entire report year as you collect sample data.

(d)   If you are using sample data from the current report year, identifying and correcting errors in the sample data now for developing sampling plans saves you from doing it when you develop your estimates later for reporting purposes.

(e)    At a minimum, you should conduct the following checks to identifying potential errors:

(1)   Overall consistency – UPT and PMT should be closely related across all sample units.  For period-based sampling plans, check for errors if cell D26 in the PeriodCalculations worksheet is smaller than 0.5.  For interval-based sampling plans, check for errors if cell D20 in the IntervalCalculations worksheet is smaller than 0.5.

(2)   Consistency at the level of sampling units – UPT and PMT should both be zero or both be positive.  Check for errors if one is zero but the other is positive.

(3)   Comparing APTL with route length – Calculate the ratio of PMT over UPT and compare the result to the maximum length of your route.  This check is not applicable to demand response services.

(4)   If you consider the PPMT option, calculate the ratio of PMT over PPMT for each sample unit; the ratio should not be greater than 1.

(f)    It is in your interest to keep your sample data free of errors.  Experience indicates that sample data with errors tend to result in larger sample sizes than sample data free of errors. 

§ 53.05 – Do interval-based sampling plans have any special requirement for the sample data used in developing them?

(a)    Yes, the sample data must have 2 or more service units for every sample day.

(b)   If you sampled 3 one-way bus trips every 2nd day with an annual total sample of 549 trips during your last sampling year, your sample data would meet this requirement.  If you have identified and corrected errors in your sample data, you may use them for developing interval-based sampling plans with the template. 

§ 53.07 – What other data do I need to use this template?

(a)    If you want to consider the grouping option, you must have data:

(1)   on the size of each service group in the number of service units operated during the period for which the sample data were collected, and

(2)   on the expected size of each service group for future sampling years. 

(3)   Follow the instructions in the template on how you should enter the information on group size.

(b)   If you want to consider combining the grouping option with the APTL option, you must enter into the template whether you will have reliable 100% counts of UPT for each service group for estimating your annual PMT.  Follow the instructions in the template on where you should enter this information.  Your input affects several things:

(1)   The efficiency of your sampling plan.  In general your sampling plan is more efficient if you will have reliable 100% counts of UPT by service group.

(2)   How you should estimate your annual PMT.  For example, you will need to estimate your annual PMT separately for each service group if you have 100% counts of UPT by service group.  Refer to Subsection 83 for more guidance on estimation.  

(3)   The information you need to get from the template for determining whether you need to revise your template sampling plan.  Refer to Subsection 56 for more guidance on revision of sampling plans.

(c)    If you want to consider interval-based sampling plans, you must have information on the annual number of days of service and the daily number of trips.

§ 53.09 – What sample data should I use to develop template sampling plans for next year?

(a)    If you are not sampling this year, you should use the sample data from your most recent previous sampling year.

(b)   If you are sampling this year, you should use the sample data from this year.  It is critical that you process your sample data as they become available.  The objective is that you can apply the companion template of this Manual to the cumulative sample data any time during this year.  Refer to Section 60 for processing your sample data.

(1)   If you are familiar with the companion template of this Manual, you should wait until a couple of weeks before the end of this year to use the template.  It is fine if you miss a few service units in your annual NTD sample for this purpose.  Note that you will still need to use the full sample to estimate your annual service-consumed data for NTD reporting.

(2)   If you are not familiar with the companion template, you should start learning to use the companion template with your sample data from this year.  Once you become familiar with the template, you should still wait until a couple of weeks before the end of this year to use the template.

§ 53.11 – How does this template deal with fluctuations in sample data from one year to another for a given service?

(a)    The measured statistical variation in APTL or any relevant characteristic of your service fluctuates from one year to another. 

(b)   Such fluctuations can occur from minor changes in your service or changes in how people use your service even if you do not adjust your service.

(c)    Fluctuations can also occur due to errors in data collection as a result of sampling.

(d)   This template has built in a 25% margin of safety against such fluctuations.

(e)    This margin of safety is equivalent to increasing the minimum precision level from 10% to 8.94% at the same 95% confidence level.

§ 53.13 – Why do such annual fluctuations in sample data require a margin of safety?

(a)    FTA’s 95% confidence and 10% precision levels are imposed on your estimates of annual service-consumed data.

(b)   The template, on the other hand, can only impose these requirements on your template sampling plans, but not on the estimates you develop with data collected according to your template sampling plans.

(c)    Whether the estimates of annual service-consumed data from the sample data collected according to the template sampling plans meet FTA’s requirements depends on how the sample data fluctuate over time:

(1)   If the statistical variation is greater in the sample data used to develop your template sampling plan, your estimates will meet FTA’s requirements.

(2)   If the statistical variation is smaller in the sample data used to develop your template sampling plan, your estimates may not meet FTA’s requirements.  The margin of safety is built in to prevent such violations from happening.

§ 53.15 – How does this template ensure adequate sample size for developing template sampling plans in the future?

(a)    Statistical theory shows that a minimum sample size of 50 is required to get an adequately precise estimate of how much APTL or any other relevant characteristic of your service varies across your service. 

(b)   This template has built in a minimum sample size of 50 for all new template sampling plans.