USGS Home Page

U.S. Geological Survey Manual

Attachment 1

U.S. Geological Survey
Office of Financial Management
Statistical Sampling Plan

August 1999

Overview

Purpose

Guidelines Used

Contents of Plan

  1. Background information and definitions
  2. Statistical sampling process by document type
    1. Payment Documents
    2. SmartCard Statements
    3. Travel Vouchers
  3. Cost of audit
  4. Analysis and actions taken based on audit results
Attachments:
  1. Audit Checklist for Payments
  2. Audit Checklist for SmartCard Statements
  3. Audit Checklist for Travel Vouchers
  4. Statistical Sampling Process Flowchart for Payments (PDF)
  5. Statistical Sampling Process Flowchart for SmartCard Statements (PDF)
  6. Statistical Sampling Process Flowchart for Travel Vouchers (PDF)

I. Background Information and Definitions

EZ-Quant Software - Quantitative statistics software developed by the Defense Contract Audit Agency. Software is programmed to perform the statistical calculations needed to analyze current audit requirements.

Sampling Universe (Population) - Samples will be selected from all payment and travel documents (PV, P2, 4P, 4V, FP, TV, T8, 3L, 8L, 8V, 7P, 7V, 6V, 3V); and SmartCard statements.

Frequency of Audit - Quarterly. The frequency factor will be analyzed each time an audit is performed and adjusted accordingly.

Variable Sampling - Concerned with values; such as dollar errors.

Attribute Sampling - Concerned with error rates; such as the frequency of errors.

Type of Audit - Documents will be audited for mathematical errors (variable sampling), administrative errors and adherence to prompt payment (attribute sampling).

Sample Size - The EZ-Quant program ATTACC1 will be used to determine the sample size based on the universe size, the critical error rate, and the desired confidence level.

Basis for Sample Selection - Vouchers must be selected entirely at random, therefore no pre-sorting of vouchers into categories such as cost center, division, etc., will be performed.

Stratification - A method by which the population is divided into homogeneous groups (strata) and elements within each stratum are selected at random for sampling. (e.g. documents over a certain dollar threshold will be reviewed 100% or documents under a certain dollar threshold will not be reviewed at all)

Point Estimate - The projected value of the entire universe dollar errors based on the results of the sample.

Precision (Reliability) - Measures the accuracy of the point estimate by determining how close the point estimate is to the true dollar errors in the universe.

Confidence Interval - A range of values that has some designated probability of including the true dollar errors in the universe. (The point estimate plus/minus the precision)

Confidence Level - The degree of assurance or reliability in the estimate. The Department of the Interior requires a 95% confidence level (5% risk).

Critical Error Rate (CER) - The maximum error rate at a given confidence level that is considered acceptable. A 4% CER falls within the range to obtain a high degree of compliance.

Audit Checklists - (Attachments 1 - 3) Organized questions analyzed and answered by the reviewer for each document in the sample.

II. Statistical Sampling Process

Documents will be audited quarterly using both variable sampling and attribute sampling methods

A. Payment Documents (Attachment 4)

A file will be generated from the PVHT and PVLT tables in the Oracle database using Crystal Reports. This file will list all payment documents for a specific quarter based on the closed date from PVHT. The file will then be imported into a Microsoft Access database.

Initially all RDE documents over $25,000 will be chosen for 100% review. PV documents over $25,000 are reviewed by OFM before payment so these will be excluded. All documents with a zero dollar amount will be removed from the sample and not audited.

The number of RDE payments between $10 and $25,000 will be input into the EZ-Quant program ATTACC1 which will calculate the sample size for reviewing. All payments made by OFM (PV's) between $10 and $25,000 will also be included in this universe with the exception of grants, awards, and contract payments. These invoices are reviewed by OFM before they are paid

The random number function in Microsoft Access will randomly select these payments to be reviewed based on the desired sample size. A report will be generated from Access and list the payments randomly selected for review.

By consulting the UserID on the Access report, the reviewer will send an email to the person who input the document (and a cc: to the administrative cost center contact) and request copies of the supporting documentation (e.g. invoice, purchase order, receiving report), which will then be faxed or express mailed to OFM.

As the proper documentation is received in OFM, the reviewer will examine each voucher and complete the Audit Checklist for Payments (Attachment 1) on a form in Microsoft Access. Access will calculate the dollar errors for variable sampling and the frequency of errors for attribute sampling.

The dollar errors for variable sampling will be input into the EZ-Quant program SAMPL - Evaluate Physical Unit of Sample for Variables. The frequency of errors will be input into the EZ-Quant program ATTEVAL1 - Attribute Sample Evaluation.

SAMPL - The desired confidence level (95%), along with the reviewed and questioned dollar amounts will be input into the program. The program will then compute the actual point estimate (total dollar error amount in the universe), the actual confidence interval (the lower/upper limits of the total dollar errors in the universe) and the actual dollar precision.

ATTEVAL1 - The desired confidence level (95%), the desired precision limits (upper), total items in the universe, the sample size and the number of errors in the sample size will be input into the program. The program will calculate the actual critical error rate achieved (upper precision limit) from the audit results; this will be compared to the desired critical error rate. The universe is acceptable if the achieved error rate is equal to or less than the desired critical error rate. If the achieved error rate is higher than the desired rate, then analysis on the types of errors and the appropriate corrective actions will be performed.

The results of both the SAMPL and the ATTEVAL1 program will be analyzed to ensure that the sample results fell within the desired levels. If the results are not within desired levels, then it must be determined if the audit sample was adequate, if the pre-set levels and intervals were on target, and if further sampling is needed.

B. SmartCard Statements (Attachment 5)

The SmartCard statements will be a separate review but will be performed in a similar manner as the payment document audit. A file will be generated from the CARD table in the Oracle database using Crystal Reports. This file will list all cardholder statements for a specific quarter based on the purchase date on the CARD table. The purchase date should correspond to the cardholder's monthly statement. Statements which only have travel related purchases will be excluded from the sampling universe because these purchases will be reviewed in the travel audit. The file will then be imported into a Microsoft Access database.

All cardholder statements will be subject to audit if they are randomly selected for the sample. The number of cardholder statements for each quarter will be input into the EZ-Quant program ATTACC1 which will calculate the sample size.

The random number function in Microsoft Access will randomly select the cardholder statements to be reviewed based on the desired sample size. A report will be generated from Access and list the cardholder statements randomly selected for review.

The reviewer will send an email to the individual cardholder (and a cc: to the administrative cost center contact) and request copies of the supporting documentation, which will then be faxed or express mailed to OFM.

As the proper documentation is received in OFM, the reviewer will examine each statement and complete the Audit Checklist for SmartCard Statements (Attachment 2) on a form in Microsoft Access. Access will calculate the dollar errors for variable sampling and the frequency of errors for attribute sampling.

The analysis of the SmartCard sample using the EZ-Quant programs will be similar to the analysis of the payment process.

C. Travel Vouchers (Attachment 6)

A file will be generated from the spending journal in the Oracle database using Crystal Reports. This file will list all travel documents processed during the quarter. The file will be imported into a Microsoft Access database.

Initially, all TDL and TDT travel vouchers between $25 and $2,500 will be subject to audit if they are randomly selected for the sample. TDL and TDT travel vouchers less than $25 will be removed from the sample and not audited. There will be a 100% review of PCS travel vouchers, and all other travel vouchers greater than $2,500.

After these vouchers are stratified by dollar amount and travel type, the total number of travel vouchers in this universe will be input into the EZ-Quant program, ATTACC1 which will calculate the sample size.

The random number function in Microsoft Access will randomly select the travel vouchers to be reviewed based on the desired sample size. An Access report will list the travel vouchers randomly selected by Access for audit.

If the travel voucher is selected for review, the reviewer will send an email to the administrative cost center contact and request copies of the supporting documentation, which will then be faxed or express mailed to OFM.

As the travel voucher is received in OFM, the reviewer will examine the travel voucher and complete the Audit Checklist for Travel Vouchers (Attachment 3) on a form in the access database. Access will calculate the error results of the reviewed documents. (Dollar value of errors for variable sampling and the frequency of errors for attribute sampling )

The analysis of the travel voucher sample using the EZ-Quant programs will be similar to the analysis of the payment process.

III. Cost of Audit

Each process of the audit will be monitored periodically to determine the amount of time required for the completion of each process and calculated against the per hour cost of the individual performing the process. This total dollar amount will be documented and available as a tool for any future management needs.

Costs to be tracked:

IV. Analysis & Actions Based on Audit Results

V. Statistical Sampling Schedule and Timeline

Email notifications to the randomly selected participants in the Statistical Sampling Plan will be sent within 30 days after the end of the quarter. Emails will be directed as follows:

Payments--to the person who input the document (identified by their UserID) with a copy to the administrative cost center contact;

SmartCard Statements--to the individual cardholder with a copy to the administrative cost center contact;

Travel Vouchers--to the administrative cost center contact.

Statistical sampling analysis will be completed within 90 days after the end of the quarter. Management reports will be made available within 120 days after the end of the quarter to all managers and administrative staff throughout the USGS.


Attachment 1

Audit Checklist for Payments

  1. The voucher amount on PVHT did not match the invoice total.

  2. Unallowable sales taxes were included with the total amount paid.

  3. Unallowable charges such as service charges, insurance costs, brokerage fees, or handling charges were included with the total amount paid.

  4. Previous balances on the invoice were paid.

  5. The log date on the invoice did not match the log date in FFS.

  6. The invoice date on the invoice did not match the invoice date in FFS.

  7. The acceptance date on the invoice did not match the acceptance date in FFS.

  8. The discount terms in FFS did not match the discount terms on the purchase order or invoice.

  9. The quantities and unit prices on the invoice did not agree with the purchase order or receiving report.

  10. The invoice was a freight charge and over $100. (It should have been paid as a GBL in OFM)

  11. Equipment over $15,000 was charged to the incorrect object class. (Should be charged to 311.)

  12. The invoice and receiving report did not agree as to the description of goods or services.

  13. The vendor address in FFS did not match the "remit to" address on the invoice. (For check payments only, look at PVHT)

  14. The invoice number on the invoice did not match the invoice number in FFS.

  15. The invoice did not include one or more of the following : invoice number, purchase order or requisition number, invoice date, vendor name, accurate description of goods or services, shipping terms, payment terms, "remit to" address.

  16. The receiving report did not include one or more of the following: purchase order number, accurate description of goods or services purchased, accurate description of quantities received, acceptance date, signature of receiving official.

  17. One or more of the following items were not annotated on the invoice: log date, acceptance date, account number, object class, fiscal year, authorized signature, document control number.

  18. The check date is before the prompt payment date, and the number of days between the two dates is greater than 7.

  19. For invoices paid by OFM: The field office did not send the invoice to OFM within 5 days from the acceptance date stamped on the receiving report. (Provided the office received the invoice before the acceptance of goods)

  20. A modification was made to the purchase order and the amendment was not attached.

  21. The receiving report was not attached to the final payment of the purchase order.

  22. The payment did not reference the obligation. (Check DXRF)

  23. The document could not be located in Imaging.

  24. The RDE user did not send their documentation to OFM after repeated requests.


Attachment 2

Audit Checklist for SmartCard Statements

  1. The purchase listed on the statement is not an authorized use of funds.*

  2. The dollar amount on the sales receipt does not match the dollar amount on the statement.

  3. An individual purchase on the statement was more than $2,500 (unless cardholder is warranted CO).

  4. A single purchase was split to avoid the $2,500 single purchase limit.

  5. The signatures on the receipts are different. *

  6. A freight charge greater than $250 was listed as a purchase on the statement.

  7. Rental of land or buildings was billed centrally to the government on the statement. *

  8. The receipt does not include an adequate description of items purchased.

  9. The individual purchase was for a restricted commodity without proper authorization. *

  10. The purchase did not include proper documentation. For example, the charge slip, register receipt, packing list, or shipping documents was not included.

  11. The cardholder did not send their statement to OFM after repeated requests.

* Notes

1. Examples -- personal use items such as hair & nail care, children's apparel, centrally billed food/refreshment charges (unless documented), ATM (cash advance) charges not in connection with travel, bottled water (unless documented), memberships (except for libraries).

5. This implies the cardholder let someone else use their card without proper authorization. It is OK if another employee has signed the receipt (in his/her own name) as long as it includes a notation such as "telephone order by [cardholder name]" and the signature is annotated "received by."

7. This will be an error only if the rental period exceeds one year and/or the rent exceeds $2,500 for the year.

9. Restricted purchases include: printing, duplication, binding, furniture other than purchases from GSA or Unicor, reprographic equipment, radio equipment, and firearms or ammunition.

Examined by:

Date:


Attachment 3

Audit Checklist for Travel Vouchers

  1. The travel voucher did not have the receipt for the airline ticket, car rental, lodging, and/or other items over $75.00.

  2. The car rental dates claimed did not agree with the official travel dates.

  3. The lodging dates claimed did not agree with the official travel dates.

  4. The travel voucher was not signed by the traveler and/or approving official.

  5. The car rental was a luxury or sports car which was acquired at the full rental rate for that type of vehicle.

  6. The traveler had an outstanding advance which was not applied to the travel voucher.

  7. The advance was deducted from the travel voucher reimbursement, but did not update the TADV table in FFS.

  8. The traveler claimed long distance calls rather than using the FTS 2000 card.

  9. The traveler claimed direct reimbursement for airfare, car rental, lodging, and/or parking which were charged to the Government Nations Bank credit card. (Centrally billed items).

  10. The charges prorated for personal use were reimbursed directly to the traveler or paid by the Government Nations Bank credit card. (Verify USGS check register for possible payment by traveler.)

  11. The traveler claimed direct reimbursement for retail items which were charged to the Government Nations Bank credit card. (Centrally billed items).

  12. The per diem allowance was incorrect for the TDY location.

  13. The airline ticket was charged to a personal credit card resulting in non-government rate and reimbursement claimed.

  14. The traveler claimed full (instead of 3/4) per diem on the first and/or last day of travel.

  15. The car rental was not authorized on the TDT authorization.

  16. The voucher did not contain approval statement and signature for increased per diem claimed.

  17. The traveler did not reduce the meal allowance (M&IE) for meals provided.

  18. The airline ticket charge includes a destination in addition to the official travel destination. (Unofficial travel is not to be obtained at the government contract carrier fare or charged to the government charge card).

    And, if Temporary Duty Extended:

  19. The per diem was not reduced when period of travel exceeded 30 days (reduction is from first to last day).

  20. The travel voucher does not have a copy of the Temporary Duty Extended (TDE) authorization.

  21. The travel voucher was not taxed and reported to PayPers as taxable income.

    And, if Foreign Travel:

  1. The foreign travel voucher did not include an approved DI-1175.

  2. The receipts were annotated with the U.S. dollar rate.

  3. The traveler was not authorized to use a foreign carrier.

  4. The traveler was not authorized a stopover in CONUS.

  5. The traveler's rest stop was in excess of 24 hours in OCONUS.

  6. The traveler's flight time did not exceed 14 hours for entitlement of rest period (OCONUS).

  7. The per diem claimed for the rest period is incorrect.

  8. The amount claimed, based on the foreign exchange rate, is different than the amount posted to the government charge card.

  9. The per diem was not calculated when crossing the international dateline.


Return to Instructional Memoranda Page
Return to Survey Manual Home Page
U.S. Geological Survey, 807 National Center, Reston, VA 20192, USA
URL: http://www.usgs.gov/usgs-manual/im/statsample.html
Content Information Contact: jblickley@usgs.gov
Contact: OIS, Corporate Information Technology Branch
Last modification: 16-Sept-1999@13:08 (klm)