Customer Service at MVD Field Offices

              Customer Service at
MVD Field Offices
FINAL REPORT 544
June 2008
Prepared by:
Ian Tingen & David Lovis- McMahon
1050 South Stanley # 156
Tempe, AZ, 85281
Telephone: ( 480) 242- 7643
Email: itingen@ gmail. com, dlovism@ gmail. com
Prepared for:
Arizona Department of Transportation
206 South 17th Avenue
Phoenix, Arizona 85007
in cooperation with
U. S. Department of Transportation
Federal Highway Administration
The contents of the report reflect the views of the authors who are responsible for the facts and
the accuracy of the data presented herein. The contents do not necessarily reflect the official
views or policies of the Arizona Department of Transportation or the Federal Highway
Administration. This report does not constitute a standard, specification, or regulation. Trade or
manufacturers’ names which may appear herein are cited only because they are considered
essential to the objectives of the report. The U. S. Government and The State of Arizona do not
endorse products or manufacturers.
This report can also be found on our web site…
http:// www. dot. state. az. us/ ABOUT/ atrc/ Publications/ Publications. htm
Technical Report Documentation Page
1. Report No.
FHWA- AZ- 08- 544
2. Government Accession No.
3. Recipient’s Catalog No.
4. Title and Subtitle
5. Report Date
June 2008
Customer Service at MVD Field Offices 6. Performing Organization Code
7. Authors
Ian Tingen, David Lovis- McMahon
8. Performing Organization Report No.
9. Performing Organization Name and Address
Ian Tingen, David Lovis- McMahon
1050 S Stanley # 156, Tempe, AZ, 85281
10. Work Unit No.
11. Contract or Grant No.
SPR- PL- 1( 59) 544
12. Sponsoring Agency Name and Address
ARIZONA DEPARTMENT OF TRANSPORTATION
206 S. 17TH AVENUE
13. Type of Report & Period Covered
PHOENIX, ARIZONA 85007
Project Manager: John Semmens
14. Sponsoring Agency Code
15. Supplementary Notes
Prepared in cooperation with the U. S. Department of Transportation, Federal Highway Administration
16. Abstract
KEY FINDINGS
♦ Customer factors have little impact on wait times, if any
♦ The main issue at hand is the non- identifiability of MVD field office service representatives.
♦ Increasing staffing volume is unlikely to have any positive effect.
♦ These findings are generalizable across all MVD offices
KEY RECOMMENDATIONS
♦ An in- depth study of each of the highest volume offices is necessary to remediate the problem
♦ General remediation strategies will yield some results, but if cost is an issue, there should be targeted studies conducted
♦ Detailed data should be kept on CSR’s and transactions at each MVD office.
17. Key Words
Customer Service, Best Practices, Social Loafing
18. Distribution Statement
Document is available to the
U. S. public through the
National Technical Information
Service, Springfield, Virginia
22161
23. Registrant’s Seal
19. Security Classification
Unclassified
20. Security Classification
Unclassified
21. No. of Pages
34
22. Price
SI* ( MODERN METRIC) CONVERSION FACTORS
APPROXIMATE CONVERSIONS TO SI UNITS APPROXIMATE CONVERSIONS FROM SI UNITS
Symbol When You Know Multiply By To Find Symbol Symbol When You Know Multiply By To Find Symbol
LENGTH LENGTH
in inches 25.4 millimeters mm mm millimeters 0.039 inches in
ft feet 0.305 meters m m meters 3.28 feet ft
yd yards 0.914 meters m m meters 1.09 yards yd
mi miles 1.61 kilometers km km kilometers 0.621 miles mi
AREA AREA
in2 square inches 645.2 square millimeters mm2 mm2 Square millimeters 0.0016 square inches in2
ft2 square feet 0.093 square meters m2 m2 Square meters 10.764 square feet ft2
yd2 square yards 0.836 square meters m2 m2 Square meters 1.195 square yards yd2
ac acres 0.405 hectares ha ha hectares 2.47 acres ac
mi2 square miles 2.59 square kilometers km2 km2 Square kilometers 0.386 square miles mi2
VOLUME VOLUME
fl oz fluid ounces 29.57 milliliters mL mL milliliters 0.034 fluid ounces fl oz
gal gallons 3.785 liters L L liters 0.264 gallons gal
ft3 cubic feet 0.028 cubic meters m3 m3 Cubic meters 35.315 cubic feet ft3
yd3 cubic yards 0.765 cubic meters m3 m3 Cubic meters 1.308 cubic yards yd3
NOTE: Volumes greater than 1000L shall be shown in m3.
MASS MASS
oz ounces 28.35 grams g g grams 0.035 ounces oz
lb pounds 0.454 kilograms kg kg kilograms 2.205 pounds lb
T short tons ( 2000lb) 0.907 megagrams
( or “ metric ton”)
mg
( or “ t”)
mg megagrams
( or “ metric ton”)
1.102 short tons ( 2000lb) T
TEMPERATURE ( exact) TEMPERATURE ( exact)
º F Fahrenheit
temperature
5( F- 32)/ 9
or ( F- 32)/ 1.8
Celsius temperature º C º C Celsius temperature 1.8C + 32 Fahrenheit
temperature
º F
ILLUMINATION ILLUMINATION
fc foot candles 10.76 lux lx lx lux 0.0929 foot- candles fc
fl foot- Lamberts 3.426 candela/ m2 cd/ m2 cd/ m2 candela/ m2 0.2919 foot- Lamberts fl
FORCE AND PRESSURE OR STRESS FORCE AND PRESSURE OR STRESS
lbf poundforce 4.45 newtons N N newtons 0.225 poundforce lbf
lbf/ in2 poundforce per
square inch
6.89 kilopascals kPa kPa kilopascals 0.145 poundforce per
square inch
lbf/ in2
SI is the symbol for the International System of Units. Appropriate rounding should be made to comply with Section 4 of ASTM E380
TABLE OF CONTENTS
Executive Summary............................................................................ 1
Introduction......................................................................................... 3
Overview.......................................................................................... 3
Disclaimer........................................................................................ 3
Literature Review ............................................................................... 4
Background Information.................................................................. 4
Literature Review Introduction ....................................................... 4
Methods ........................................................................................... 4
Synopsis........................................................................................... 5
Findings ........................................................................................... 5
Data............................................................................................. 5
Analysis ...................................................................................... 5
Peer Information......................................................................... 6
Direction ..................................................................................... 6
Real Id......................................................................................... 7
Instrument and Methods.................................................................... 8
Study Background ...................................................................... 8
Instrument Design ...................................................................... 8
Research Methods .................................................................... 10
Data .................................................................................................... 11
Analysis .............................................................................................. 14
Conclusion and Implications ........................................................... 17
Appendix and Figure........................................................................ 19
List of Tables and Figures
Table 1. Correlations for all MVD locations................................. 12
Table 2. Correlations for Location 1 and Location 2 ................... 13
Figure 1. Data Collection Instrument ............................................ 30
GLOSSARY OF ACRONYMS
ATRC Arizona Transportation Research Center
ADOT Arizona Department of Transportation
CSA Customer Service Associate
CSR Customer Service Representative
DOT Department of Transportation
GAO Government Accountability Office
GIS Google Index Search
ID Identification
MVD Motor Vehicle Division
NAS National Academies of Science
SPR State Planning & Research
TAC technical advisory committee
TRB Transportation Research Board
TRIS Transportation Research Information Services
ACKNOWLEDGEMENTS
The authors would like to thank Project Manager John Semmens, the managers of the
Phoenix- area MVD field offices, and the personnel, and especially the interns: Anthony
Walter and Diana Artalejo. Without you, this research would have been unmanageable.
A gracious thanks to Diana Artalejo, lead intern. Without her, our project would have
been delayed ad infinitum.
A very special thanks to Gregory Neidert, whose guidance was invaluable.
1
Executive Summary
The Arizona Department of Transportation ( ADOT), through the Arizona Transportation
Research Center ( ATRC) has requested that ways to improve customer service and
reduce wait times at Motor Vehicle Division ( MVD) offices be researched. This research
has been conducted by a team of Arizona State University researchers, with the following
important points being noted:
Arizona is ahead of the curve in terms of documenting customer service methods.
After conducting a thorough review of transportation literature, journals, and polling of
other out- of- state sources, it is our conclusion that Arizona is a trendsetter in not only the
amount of data accessible to use to improve practices, but also in that it actually uses data
to try to remedy issues in customer service.
Arizona’s best practices are some of the best in the nation.
It is also the conclusion of our research that ADOT is very attentive to details in terms of
scientifically approaching its problems. Whereas many states use best guesses or other
such possibly inaccurate methods of management, ADOT seeks to use the best data,
processed by people who know how to use such data. There is a great dearth of any real
empirical practice anywhere else in the nation — ADOT should be confident of its
trendsetting practices.
Wait times at the MVD are greatly increased by a phenomenon known
as social loafing.
In the process of examining the issue, our team has concluded that the major contributing
factor to increased wait times is a phenomenon known as social loafing ( or the diffusion
of responsibility) — in lay terms, this is a weakening of group customer service
effectiveness caused by a lack of identifiability of individual efforts and a few other
factors. Loafing is common at all locations and sections of MVD offices studied – in
short, it did not matter where, when, or how the observations were made: diffusion of
responsibility appeared to permeate all offices equally. It is important to note that while
prevalent, this phenomenon is not intentional on the part of the customer service
representatives ( CSRs).
Customer factors play little, if any, part in increased wait times.
Considerations of customer factors weighed heavily in this research, indeed they were the
first group we turned to in looking for a possible answer to the issues. Via statistical
analysis, however, we concluded that the bevy of customer factors brought into each
MVD field office do not have any significant effect on wait times.
2
Remediation of social loafing factors should be implemented to reduce current
problems.
Given these last two points, it is the team’s recommendation that a few steps be taken to
remediate these “ loafing” situations. Among these are:
1) Creating more identifiability for Customer Service Representatives
2) Creating more open channels of goal communication
3) Further study of each MVD office to remediate situation- specific issues
3
Introduction
Overview
A major concern of Arizona MVD customers is the wait time experienced in the field
offices. This report seeks to at least partially explain the phenomenon.
This is not to say that the problem is one that is necessarily direct and straightforward.
The seemingly obvious methods of exploring and explaining were not as easily available
as one might hope. It should be known that Arizona’s methods and practices are quite
possibly the best in the nation; hence, comparison to other states very well may be taking
a step backwards. Still, this makes the only real comparison that is feasible that of the
Arizona MVD to itself; this presented unique challenges as well.
In trying to tease out the reasons behind long wait times, a bevy of factors immediately
jump to mind. Could this be an external issue – something related to what the customers
bring in to the field offices? Are wait time issues common to all offices, or just a select
few? What about the CSR’s? Is there anything about them that is contributing to wait
times?
Despite all these questions and challenges, this team was able to extract meaningful,
answers to the problems posed by this study.
Disclaimer
This report is completion of SPR Project 544. Ian Tingen and David Lovis- McMahon
assembled a research team from Arizona State University for this project. Any questions
should be directed to Ian Tingen ( itingen@ gmail. com) or David Lovis- McMahon
( dlovism@ gmail. com). All hereafter is original research conducted under the direction of
Tingen and Lovis- McMahon, with Dr. Gregory Neidert receiving a hearty thank you for
his consultation on this project.
4
Literature Review
Background Information
At the start of this project, ADOT provided the research team with customer service data
spanning from FY 2002 to FY 2006. These summary data were used as a basis on which
to compare other states’ data and approaches, and as a beginning point for rudimentary
analysis. These data were instrumental in assessing not only the effectiveness of ADOT’s
MVD offices and practices, but in establishing the relative position of these same things
to other states. A full literature review and analysis of such follow.
Literature Review Introduction
The literature review helps establish three important pieces of information:
1. Which states collect and report customer wait times.
2. What are the mean wait times at other states’ MVD facilities, and if possible, what are
the specific wait times for licensure and titling.
3. What improvement plans are states implementing to reduce customer wait times or in
what other ways are state MVDs improving customer satisfaction.
Methods
Three searches were conducted: a Google Index Search ( GIS) of all 50 states was run us-ing
a specialized Google search created by Washington State Department of Transportation
( DOT) ( http:// www. google. com/ coop/ cse? cx= 006511338351663161139% 3Acnk1qdck0dc),
a search of the Transportation Research Information Services ( TRIS) ( http:// ntlsearch. bts. gov
/ tris/ index. do), and a search of the Arizona State University digital catalogue
( www. asu. edu/ lib). The search terms used were:
o · customer service
o · customer wait times
o · best practices wait times
o · best practices customer service
o · driving licensure
o · licensure
o · titling
o · wait times
o · driver testing
All terms were also run through at least one of these additional modifiers: + improvement/ ing,
+ reducing, comparison between X and Y, + best practices. An additional search modifier was
introduced to remove all articles from dot. state. az, because the significantly larger body of
research originating from ADOT was heavily skewing searches. This point will be talked
about more in depth later.
5
Synopsis
1) Besides ADOT ( for customer wait times) only the Maryland Department of Transportation
( MDOT) and Oregon Department of Transportation ( ODOT) have publicly released figures of
wait times at their MVD facilities.
2) According to a 2004 report from MDOT, average wait time in its MVD offices reached a
lowest point of 34 minutes, where as ODOT’s average wait time was 13.6 in 2003 ( gleaned
from its 2005 report). These numbers were reached by the respective states’ MVD analogues
– with little definition as to what the times entailed.
Findings
Data
The FY 2002- 2006 reports provided by ADOT are of great interest to the project in that they
determine initial hypothetical causality on the two factors that this project is meant to improve
– customer service and wait times. Using data collected from the reports, three facts of interest
were found. First, the number of customer service representatives ( CSRs) or customer service
associates ( CSAs) on duty significantly negatively correlated1 with customer wait times ( r=-
.5892, p < .0443) – this is to say that the more CSR’s on duty, the less wait time customers had.
However, the number of CSRs / CSAs on duty significantly positively correlated with
customer time from counter to door ( r=. 934, p < .000) – this tells us that the more CSR’s on
duty, the longer the customer spends at the counter with a CSR. Third, the average wait time
from door to counter significantly negatively correlated with average wait time from counter
to door ( r=-. 765, p < .004); in other words, the more time it took a customer to see a CSR, the
less time it took the CSR to complete the transaction and vice versa. These numbers are
significant in that, taken as a whole, they point to the possibility of CSRs/ CSAs, or at least
some of the qualities of CSRs / CSAs, significantly impacting wait times in a negative
manner.
Analysis
As highlighted in the methods section ADOT had to be removed from the GIS search, because
of the paucity of research conducted by other states. This is significant in that it gives us a
relative view of how far ahead of the curve ADOT is in terms of research production,
management, publication, and use. The body of literature relevant to this topic is surprisingly
sparse in terms of substantive studies and relevant facts. Furthermore, this project is not the
only one to come to this conclusion. The 2003 Transportation Research Board report Research
for Customer- Driven Benchmarking of Maintenance Activities ( Booz Allen Hamilton 2003)
talks in- depth about the lack of research on the topic - and further states that much of what
does exist in terms of the body of literature is statistically invalid - that is, most states are using
a rather “ shot- in- the- dark” method of approximating success and failure. When the other
1 For a full description of positive and negative correlation, please see the appendix.
2 For a full description of r- values, please see the appendix.
3 For a full description of p- values, please see the appendix.
6
reports are taken in context of this scientifically significant study, it is clear that most states are
not using any degree of solid methodology in terms of creation, execution, and sustaining best
practices in MVD analogues. Arizona must avoid these pitfalls and continue in its tradition of
excellence.
Peer Information
Looking more broadly, it is clear that ADOT has little to look for in its analogues. In addition
to the TRB report mentioned above, a report from the Government Accountability Office
( GAO) adds a bit of supplemental evidence that the data collection and usage methods across
the nation and at the federal level leave much to be desired. Few standardized accountability
variables exist ( GAO 2006), and those that do are rarely used in any scientifically valid
manner. Taking this into account would be wise - with ADOT aware of its possible pitfalls, it
can use the data it has much more effectively, and fall squarely within the recommendations of
the GAO.
With that caveat in place, it would seem that ADOT has, for its size, much better service in
terms of wait times and customer satisfaction. The jurisdictions investigated conduct
significantly fewer types of transactions ( in terms of raw numbers). Additionally, MDOT and
ODOT are demographically much more homogenous than Arizona. Cautiously speaking, this
is good news for Arizona, in that initially it seems that Arizona is able to handle much more
diversity in transactions and people than its analogue organizations in other states. An in- depth
study of transaction types could possibly help flesh this out in a more helpful manner.
As mentioned before, Arizona is ahead of the curve in terms of available relevant research.
Though our queries were specific to this project, the amount of Arizona- produced literature
dwarfs other sources – for example, in the Google search, the ratio of Arizona- produced to
non- Arizona- produced articles approached 3: 1. Similar results were found in the other
searches conducted. In general, the quality of the ADOT reports was better as well.
Direction
A review of the relevant ADOT literature leaves us walking away with one very key point: we
must consider that the licensing and registration core functions are handled though the
agency’s “ legacy systems” which, in some cases, are over thirty years old. MVD has the
oldest legacy system of any major Arizona agency. Thirty years is well beyond the expected
life of the software applications. The relevant legacy system consists of seven different
systems that largely operate independently of one another. The combined systems have 554
screens, 733 transactions and 3,872 programs. MVD also collects in excess of $ 1.5 billion a
year in taxes and fees, making it the State’s second largest tax collection agency. This is
significant because thousands of governmental and authorized private entities access or update
the data within those systems - further complicating matters in the current system.
From the needs assessment we can see there are manifold changes that are about to alter the
primary functioning of the MVD. Moreover, the resulting reduction in complexity for the two
( license and registration) will drastically alter the way they are handled at MVD facilities.
7
Thus, the scope of this literature review is constrained by these impending system- wide
changes. The focus instead will be on:
o identifying practices that do not rely on the underlying technology
o identifying methods that capture benchmarking data that can be useful in continuous
improvement
o using this research to not only improve wait times and customer service, but to
possibly help get more efficiency out of the upcoming system overhaul
Real ID
This report would be remiss in its duty if it did not at least briefly touch on possibly one of the
most impactful things to hit ADOT — the Real ID Act. ( P. L. 109- 13) Though the ultimate fate
of Real ID rests outside the jurisdiction of ADOT, it is important to note what additional
customer impacts could come about because of it.
In short, Real ID will force states to create a uniform database able to synchronize with other
states and national databases. Costs for fully implementing the program for Arizona are
estimated at around $ 56 million to start – and that does not include upkeep ( Senate Bill 1152).
Real ID implementation would require all of Arizona’s estimated four million drivers to get a
new license that complies with federal regulations. This process would be handled in each of
the MVD field offices – creating a large need for processes and staff to be as standardized as
possible.
As of the time of this report’s publication, states have until December 31, 2009, to comply
with Real ID. Whether or not Real ID will be fully implemented is as yet unknown ( 22 states
have drafted non- compliance legislation in what promises to be a lengthy battle), but even
now it is certain that if it is implemented, ADOT must move quickly to minimize what will be,
at best, a strain on MVD field office resources.
8
Instrument and Methods
Study Background
As noted in the literature review, any comprehensive comparisons between ADOT practices at
MVD locations and other states’ analogues are nigh impossible. Additionally, upcoming
changes in the ADOT MVD system add to the complexity of the situation. Therefore, it was
necessary to develop methods and tools to assess things that MVD offices have control over
currently, and will continue to have control over throughout the changes. We have addressed
all of these needs in our methods and tools, and - explain them a bit more in depth here.
Initial interpretation of the data furnished by ADOT provided a solid starting place for
structuring the research. First, monthly trends in customer service data were clear, as
were possible explanations for said same fluctuations. These data also gave us some other
questions: what exactly went into “ wait times” and the process of actually being serviced
at an MVD field location? With these ideas in mind, initial trial observations were
conducted at four different MVD field offices. Offices to conduct pilot research were
chosen on a number of criteria. Population demographics were taken into account, and
those offices with medium to highly populated areas were selected. We also chose offices
which served diverse populations in terms of primary language and socioeconomic status
to capture the most variance possible. ( It should be noted that these were all offices that
had the Q- matic queue management system. Non- Q- matic offices are scarce throughout
the state and such offices are also located in low population areas; long wait times are not
likely to be an issue in such areas.) All of the offices chosen were in the Phoenix
metropolitan area. Observations were conducted at different times across the month.
These initial observations were used to craft a wait time measurement tool that allowed
assessment of wait times across a great number of factors: what partitions wait times had,
how customer traffic flow affected wait times, how staffing affected wait times, how the
actual structure of the field office affected wait times, and how customer traits affected
wait times. These factors helped create an instrument to track and analyze wait time. A
more in- depth description of each of these factors and their rationale follows.
Instrument Design
During initial observations, it became clear that assessing wait time would not necessarily
be as straightforward as it seemed. We observed different phases in wait time, something
that was not covered in any of the literature or data provided to us. As such, we examined
the process and came up with what seemed to be the natural process of customer service
at MVD locations, and the wait times associated with each. These intervals allowed us to
assess at which stage a customer waited the longest. ( For example, long wait times to get
a transaction ticket could be a result of a high volume of customers.) The identified
intervals were:
1. Initial customer arrival time to receipt of Q- Matic ticket
2. Time when a customer received a Q- matic transaction ticket to when customer
was called to a window.
3. Time when the customer was called to a window to customer’s arrival at window.
9
4. Time of the customer’s arrival to the window to customer leaving.
5. Overall wait times were calculated by adding up these different time intervals.
Our next question was that of customer flow. In an informal interview, an MVD
employee indicated that the busiest days were the 15th and last day of each month due to
registration renewals and other business needing to be transacted with these dates as a
deadline. These trends were also hinted at ( though not empirically tested) in the ADOT
yearly reports. Observational research indicated these postulates to be true; furthermore,
we were able to track trends across the month in terms of how much volume was
processed over each period of days. As a result, each 15 day interval within a month was
divided into high, medium, and low traffic periods. The 1st through 5th and 16th through
20th were the low traffic periods because they directly followed the two busiest days
which lead to a sequential decrease in customer volume. The 6th through 10th and 21st
through 24th were medium traffic periods, indicating a “ ramping up” in customer flow.
The 11th through the 15th and the 25th to the 30th/ 31st were observed to be high traffic
periods, in line with the previous informal observations.
An issue related to customer flow was that of staffing. During our observations, we
noticed that during different days, different amounts of staff were present to help
customers. As a result, the instrument was given a section in which to track number of
staff on duty, both as a raw score and as a percentage of total possible capacity. This
allowed us to look at rates both for individual offices and as a way to equalize all offices
observed.
Another factor considered in the creation of our instrument was that of the actual form
and physical location of the office. Some offices were noted as having a very open
structure, one that had monitors easy to see from waiting areas and were put together in
seemingly straightforward fashions. Others were noted as being somewhat less
convenient, with some waiting areas not having immediate or easy access to monitors.
We also took note of the location of each office; ostensibly different locations might have
different things going on in terms of wait time. Each of these factors was noted to see if
there was any effect of form on customer wait times.
The final and perhaps most obvious factor that we had to consider was that of the
customer. During initial observations, we observed many customer issues that could
potentially increase wait times. For example, we noticed that some customers came in
with many people, usually small children. Our instrument included an area to track how
many people came in with each customer, as we thought that the additional distractions
could make the customer oblivious to his number being called. We also tracked whether
or not a customer had other distractions while waiting – reading, listening to music,
sleeping, and cell phone usage were all monitored, as we thought that these things may
make a person less likely to see or hear when that his number was called, thus increasing
wait time. Demographic factors were also selected: gender, apparent ethnicity, and age
were taken into account to see if any effect of said same could be found.
1 0
Summarily, our developed instrument allowed analysis of five key factors potentially effecting
wait times:
1. Wait times at different points in the service process
2. Customer flow density
3. Number of staff present
4. Physical characteristics of location
5. Customer traits
Please see Figure 1 in the appendixes for an example of the data collection instrument.
Research Methods
After development of the instrument, three field observers conducted reliability training and
field observations at which all three observers were present. Inter- rater reliability analysis was
conducted across these sessions, and sufficient reliability ( α4 = .89) was reached to have each
observer conduct analysis on their own. Over a period of six months, 30 field observations
were conducted at five MVD field offices, resulting in ~ N5 = 300 data points for customer
wait times and transactions. These data were distributed roughly evenly across each office.
Data were then entered into a statistical analysis program, SPSS and appropriate statistical
tests were run.
4 For a full description of alpha levels, please see the appendix.
5 For a full description of N, please see the appendix.
1 1
Data
Correlations found among the variables of study are shown in Table 1. Several
significant associations between the variables were found after a bivariate correlation
analysis was conducted. First, the amount of customer traffic and number of minutes
from arrival until ticket dispensed were significantly and positively correlated ( r=. 368,
p<. 01). Also positively correlated with amount of customer traffic were the number of
total windows per office ( r=. 219, p<. 05). A few other correlations among the variables
were significant and noteworthy. As the number of open windows per office increased,
the number of minutes from arrival until a ticket was dispensed increased as well ( r=. 628,
p<. 01). In addition, the number of open windows per plant was significantly and
positively associated with the number of minutes between receipt of a ticket and reaching
the window ( r=. 465, p<. 01). Therefore, the number of open windows per office was
positively correlated with the number of minutes from initial arrival until reaching the
window ( r=. 548, p<. 01). Finally, as the number of total windows per plant increases, so
does the number of open windows per plant ( r=. 812, p<. 01).
A contrast analysis of Location 1 and Location 2 was also conducted. These locations are
the highest- volume locations, and a specific analysis of them was conducted to see if dif-ferential
effects were being noted as opposed to the total population. Correlations found
among the variables of study between Location 1 and Location 2 are shown in Table 2.
As was found in the correlation analysis of all the MVD locations, the amount of cus-tomer
traffic and the number of minutes from initial arrival until a ticket was dispensed
are positively correlated ( r=. 394, p<. 01). However, there are two significant correlations
that have especially important implications for customer service. As the amount of
customer traffic decreases between Location 1 and Location 2, the number of minutes
between a ticket being dispensed and the customer reaching the window increases ( r= -
.379, p<. 01). Therefore, as the amount of customer traffic decreases, the number of
minutes from initial arrival until reaching the window increases as well ( r= -. 268, p<. 05).
Also of statistical significance was the negative association between the amount of
customer traffic and the number of open windows per office ( r= -. 315, p<. 05). However,
this does not necessarily account for longer wait times, due to the positive correlations
found between number of open windows per office and segmentation of the dependent
variable. As the number of open windows per office increases, so does the number of
minutes from initial arrival until a ticket is dispensed ( r=. 611, p<. 01). The number of
open windows per office and the number of minutes from a ticket being dispensed until
the customer reaches the window is positively and significantly associated ( r=. 647,
p<. 01). In addition, as the number of open windows per office increases, so does the
number of minutes from initial arrival until arrival at the window ( r=. 713, p<. 01).
The analyses conducted with all of the MVD locations indicate that as the number of
customer service representatives increases, so does the customer wait time ( arrival to
ticket, ticket to window, and initial arrival to window). The hypothesis is further
supported when a contrast analysis of the highest trafficked locations ( Location 1 and
Location 2) is conducted. Not only do the significant positive correlations between
number of customer service representatives per plant ( operationalized by number of open
1 2
windows) and number of minutes from initial arrival until reaching the window still exist,
but an inverse relationship between amount of customer traffic and wait times co- exists.
Therefore, as shown by Table 1 and Table 2, periods of low traffic in these locations lead
to longer wait times than periods of high traffic.
Table 1.
Correlations for All MVD Locations
1 .368** -. 096 -. 008 .104 .219*
.000 .300 .928 .255 .016
122 121 119 119 121 121
.368** 1 .328** .510** .628** .397**
.000 .000 .000 .000 .000
121 121 119 119 120 120
-. 096 .328** 1 .979** .465** .228*
.300 .000 .000 .000 .013
119 119 119 119 118 118
-. 008 .510** .979** 1 .548** .283**
.928 .000 .000 .000 .002
119 119 119 119 118 118
.104 .628** .465** .548** 1 .812**
.255 .000 .000 .000 .000
121 120 118 118 121 121
.219* .397** .228* .283** .812** 1
.016 .000 .013 .002 .000
121 120 118 118 121 121
Pearson Correlation
Sig. ( 2- tailed)
N
Pearson Correlation
Sig. ( 2- tailed)
N
Pearson Correlation
Sig. ( 2- tailed)
N
Pearson Correlation
Sig. ( 2- tailed)
N
Pearson Correlation
Sig. ( 2- tailed)
N
Pearson Correlation
Sig. ( 2- tailed)
N
Traffic
Number of Minutes from
Arrival to Ticket
Dispensed
Number of Minutes from
Ticket to Window
Number of Minutes from
Arrival Time to Window
Number of Windows
Open Per Plant
Total Number of
Windows Per Plant
Traffic
Number of
Minutes
from Arrival
to Ticket
Dispensed
Number of
Minutes
from Ticket
to Window
Number of
Minutes from
Arrival Time to
Window
Number of
Windows
Open Per
Plant
Total Number
of Windows
Per Plant
**. Correlation is significant at the 0.01 level ( 2- tailed).
*. Correlation is significant at the 0.05 level ( 2- tailed).
1 3
Table 2.
Correlations for MVD Location 1 and Location 2
1 -. 072 .394** -. 379** -. 268* -. 315* .297*
.593 .002 .004 .044 .017 .025
57 57 57 57 57 57 57
-. 072 1 .093 .040 .055 -. 010 .006
.593 .492 .767 .687 .942 .967
57 57 57 57 57 57 57
.394** .093 1 .339** .507** .611** -. 355**
.002 .492 .010 .000 .000 .007
57 57 57 57 57 57 57
-. 379** .040 .339** 1 .982** .647** -. 573**
.004 .767 .010 .000 .000 .000
57 57 57 57 57 57 57
-. 268* .055 .507** .982** 1 .713** -. 592**
.044 .687 .000 .000 .000 .000
57 57 57 57 57 57 57
-. 315* -. 010 .611** .647** .713** 1 -. 795**
.017 .942 .000 .000 .000 .000
57 57 57 57 57 57 57
.297* .006 -. 355** -. 573** -. 592** -. 795** 1
.025 .967 .007 .000 .000 .000
57 57 57 57 57 57 57
Pearson Correlation
Sig. ( 2- tailed)
N
Pearson Correlation
Sig. ( 2- tailed)
N
Pearson Correlation
Sig. ( 2- tailed)
N
Pearson Correlation
Sig. ( 2- tailed)
N
Pearson Correlation
Sig. ( 2- tailed)
N
Pearson Correlation
Sig. ( 2- tailed)
N
Pearson Correlation
Sig. ( 2- tailed)
N
Traffic
Number of People with
Customer
Number of Minutes from
Arrival to Ticket
Dispensed
Number of Minutes from
Ticket to Window
Number of Minutes from
Arrival Time to Window
Number of Windows
Open Per Plant
Total Number of
Windows Per Plant
Traffic
Number of
People with
Customer
Number of
Minutes
from Arrival
to Ticket
Dispensed
Number of
Minutes
from Ticket
to Window
Number of
Minutes from
Arrival Time to
Window
Number of
Windows
Open Per
Plant
Total Number
of Windows
Per Plant
**. Correlation is significant at the 0.01 level ( 2- tailed).
*. Correlation is significant at the 0.05 level ( 2- tailed).
1 4
Analysis
The data from this study are quite telling. First, we must discuss what does NOT affect
customer wait times. Each of the following had non- significant relationships ( when
controlling all other variables) with customer wait times:
Customer Traits
‐ Customer demographic factors
‐ Customer distractors
‐ Number of people with customer
Customer Volume
‐ Number of customers
‐ Time of month ( low, medium, or high traffic)
MVD Field Office Factors
‐ Location
‐ Shape / Layout
Of these things, only customer volume had a negligibly positive relationship with wait time.
All other factors had no significant effect on wait time.
Controlling for all other factors, factors that did have a significant relationship with wait time
are:
• Number of Customer Service Representatives on duty
• Percentage of Customer Service Representatives on duty out of total possible
Customer Service Representatives
Each of these variables had a significant, positive relationship with customer wait times.
The relationships noted in the data are not necessarily the most intuitive, but can be explained
rather simply. As seen, customer variables have little, if anything, to do with wait time. This is
most likely an effect of the clarity of the Q- Matic system: when people’s numbers are called,
they are done so unambiguously and are responded to quickly. Customer demographics,
distractors, and the number of people with the customer were all non- issues.
Also a non- issue is customer flow. By a process called “ controlling,” we are able to
statistically remove the influence of all factors save our factor of interest. By doing this, we
can examine if different levels of our factor of interest are directly influencing wait times. In
the case of customer flow, there is no relationship. Though there is a slight upswing in wait
time with more customer flow, it is negligible and not the primary cause of long wait times.
The same is true of the time of month that is being examined.
Another non- issue is that of location and layout of the actual MVD office. Controlling for all
other factors shows that there is no effect of either location or layout. This is important to note
1 5
as the drastically different demographics of each location did not have as much of an effect as
expected.
Finally, we come to staffing. Both measures, those of raw number of staff and percentage of
staff working, were significantly, positively correlated with increased wait times. In short, this
means that the more staff available, the longer customers had to wait for service. These
statistics were derived in the same manner as those previous – removing the influence of all
other variables to see ONLY the effect of the current analysis. Further analyses of the staffing
level find that not only is it significantly positively related with wait times, but that these
effects are almost 40 times greater than chance alone.
In any study of this scope, it is important to take the utmost care to assess each individual
variable carefully, systematically, and rationally. This study has done just that; months of
research, observation and interviews culminated in identifying factors relevant to the research
question. Data were collected on these relevant factors and analyzed. The implications that
follow are directly pointed to by the data, and are relatively unambiguous.
First, it is important to note the universality of all trends discussed. The process of statistically
controlling is powerful and central to all analyses – it allows us to specifically examine each
part of the puzzle and see just what it contributes to the overall picture. Each finding does just
that: tells us what specifically it contributes to wait time. As stated earlier, no customer or
location factors play any significant role in long wait times. Further, there is little evidence to
suggest that these factors work in conjunction with any others to significantly increase wait
times.
We are, therefore, left with only one thing: numbers of customer service representatives. The
statistics are phenomenally large, and as such deserve an explanation. Fortunately, this same
sort of situation has been well documented in organizational psychology literature6 for years.
The numbers found at ADOT are a classic case of what is called “ social loafing.”
Social loafing, also known as the diffusion of responsibility, occurs in situations where
individual contributions to a group goal are not salient. This results in a less- active, more
relaxed work strategy on the part of the individual. Many different factors can contribute to
this occurrence, but they all result in the same general downward trend in productivity.
Specific to ADOT, this means longer customer wait times.
Put another way, diffusion of responsibility impacts ADOT’s main concern of this project –
customer satisfaction. It does this because of the negative impact it has on customer wait
6 For more reading:
• Latane, B., Williams, K., & Harkins, S., Many hands make light the work: The causes and
consequences of social loafing, Journal of Personality and Social Psychology, June 1979,
Vol. 37, 822- 832
• Karau, S. J. & Williams, K. D. ( 1993). Social loafing: A meta- analytic review and theoretical
integration. Journal of Personality and Social Psychology, 65, 681- 706.
• Jackson, J. M. & Harkins, S. G. ( 1985). Equity in effort: An explanation of the social loafing
effect. Journal of Personality and Social Psychology, 49, 1199- 1206.
1 6
times. What happens is this: as more people are expected at MVD branches, more personnel
are assigned. However, instead of the predicted increase in the ability of the branch to serve
customers in a timely fashion, the opposite happens. This puzzling result is because as more
CSR’s and CSA’s come on, the visibility of that individual’s contributions to the whole of
customer service is diminished.
An illustration of this point: if there are three people who are on duty, then it is quite obvious
to each individual what the others are doing. This makes people want to work faster, more
efficiently, and overall better because they are aware that everyone can see what they are
doing: it acts as a positive reinforcer. Now look at another situation, say, with 20 people on
duty. Because of the sheer number of people on duty, personnel are aware that they are not
being watched nearly as carefully. This can lead to a more lax approach to getting customers
processed, and leads to the downturn in time that we have seen.
1 7
Conclusion and Implications
As shown in this report, Arizona’s Department of Transportation has many things to be proud
of – a strong dependence on statistical evidence, one of the most thorough data collection
procedures in the nation, and a willingness to use those things to better processes. Before
continuing too far, we also must reiterate the fact that there is little to compare Arizona to
because of its reliance on good data and methodology. Few, if any, other states are even able
to come close to the soundness of these processes. Because of this, we recommend using the
best practices of this state against itself for benchmarking and measurement of success, all the
while keeping the message of the TRB report at the forefront of future directions. At this
point, Arizona has started down a great track. These next points are made while keeping this
tradition in mind.
First, the phenomenon of social loafing is the most prevalent finding of this study. After
carefully analyzing a bevy of factors ranging from customer traits to layout of the actual
offices, the conclusion is clear. This is not something that is unmanageable in any way; in fact
it is a rather common problem in workplaces, and one that can be remedied quite effectively,
given the proper investment in looking at the sources of downturn unique to each group of
workers. This problem is easy to generally identify, but specifics of each situation may vary
widely.
That is why the next steps to reducing this problem are crucial. In order to effectively combat
the diffusion of responsibility phenomenon, it is of the utmost importance to understand how it
works uniquely in each situation. Differences can include but are not limited to: visibility of
personnel’s individual contribution, visibility of group personnel efforts, ratio of personnel to
supervisors, ratio of personnel to customers, length of shift, office design, etc. The next phase
of this study should look at these and all other contributing factors in order to remedy the
problem.
The future implications for this study and ones to follow it are rather straightforward: ADOT
MVD field offices have an affliction common to many large- scale workplaces. Through
careful measurement and study, this problem can be at least partially nullified. We suggest a
thorough study of the most highly trafficked MVD offices to understand how diffusion of
responsibility affects each of them. From that study, proper incentive structures and
remediation strategies can be developed and implemented. This will benefit MVD in three
poignant ways. First, wait times will be reduced greatly, and customer satisfaction should
improve as a result. Second, MVD CSR’s will be encouraged in very organic ways to keep
their level of service high. Third, Arizona will be a trendsetter yet again in providing the very
best service available via a mix of social science and applied business practices.
General Remediation Strategies
In general, diffusion of responsibility can be remedied through increasing the
identifiability of the individual’s contribution to the group workload. In ideal situations,
constant feedback is the best way to increase this. If such a strategy is not practical, then
periodic updates of efficiency can help ameliorate the problem.
1 8
With especially large groups, another way to increase efficiency is to create smaller
groups that are part of the whole. For example, splitting 20 people in to four smaller
groups of 5 can help individuals realize their contribution to the whole. When used in
conjunction with the identifiablity suggestions from above, this can prove to be a very
useful tactic.
It should be noted that well- structured reward programs can also be incentives to improve
performance, given that the path from individual performance to group performance to
reward is made salient, and that the group is able to deal with freeloaders in a quick
manner.
It should be stated again, however, that these strategies are general. Targeted, specific
programs are always better at amelioration than ‘ blanket strategies’.
Recommendations
♦ An in- depth study of each of the highest volume offices is necessary to remedy the
problem.
♦ General remediation strategies will yield some results, but if cost is an issue, targeted
studies should be conducted
♦ Detailed data should be kept on CSR’s and transactions at each MVD office.
1 9
Appendix and Figure
Statistical Terms
Correlation: Correlation is a measure of how closely two given variables are related.
Correlation is measured either on a scale of - 1 to 0 ( negative correlation) or 0 to 1
( positive correlation). The closer ( in either direction, positive or negative) the number is
to 1, the more closely related the two variables are. Generally speaking, a correlation
from |. 00 to .30| is considered a light relationship, from |. 31 to .49| a moderate
relationship, and from |. 50| up is considered a strong relationship.
Negative Correlation: Negative correlation is a condition where the relationship between
two variables is inverse – as one variable goes down, the other goes up ( and vice versa).
Positive Correlation: Positive correlation is a condition where the relationship between
two variables is direct – as one variable goes up, so does the other; or as one variable
goes down so does the other.
R- values: R is the numeric representation of a correlation between two variables. Its
range is from - 1 to 1. A positive value indicates a positive correlation; a negative value
indicates a negative correlation.
Alpha Level ( α): Alpha is a measure of how consistently a group of observers describe
similar events. It can be a value anywhere between 0 and 1. Higher values indicate more
consistency. Generally, alpha values greater than .8 are sufficient for field research.
N: Simply put, N is the number of observed subjects or the number of discrete
observations in a study.
Article Review
ADOT/ MVD Strategic Plan -- FY 2001 ADOT/ MVD Strategic Plan -- FY 2003
ADOT/ MVD Strategic Plan -- FY 2002 ADOT/ MVD Strategic Plan -- FY 2006
Longitudinal Data Analysis of MVD Statistics for FY2002 to 2006:
The purpose of the longitudinal data analysis is to get a sense of how important statistics
like average customer total wait time have changed over the last five years. It is also
useful to see what current trends are and what those trends have to say about the future.
The following categories were used in this data analysis:
1. Number of MVD customers served in field offices ( in thousands)
2. Number of transactions ( thousands)
3. Average customer wait time ( door- to- counter) in field offices ( minutes)
4. Average transaction time ( counter- to- door) in field offices ( minutes)
5. Average customer total visit time ( door- to- door) in field offices ( minutes)
6. Average number of CSAs and CSRs ( for FY 2003 to 2006 only).
2 0
The below table summarizes the range of values as well as the average values over the
past five years:
Descriptive Statistics
Range Minimum Maximum Mean
Number of MVD customers served in field
offices ( in thousands)
153.3 289.7 443.0 380.055
Number of transactions ( thousands) 175.5 396.0 571.5 476.724
Average customer wait time ( door- to-counter)
in field offices ( minutes)
25.5 10.2 35.7 19.363
Average transaction time ( counter- to- door) in
field offices ( minutes)
1.3 7.7 9.0 8.338
Average customer total visit time ( door- to-door)
in field offices ( minutes)
24.7 19.2 43.9 27.702
Average number of CSAs and CSRs 224.0 664.0 888.0 761.229
The table shows that on average it takes approximately 19 minutes for a customer to be
seen by Customer Service Agent, and that it takes approximately 8 minutes for the
transaction to take place.
Descriptive Statistics
Range Minimum Maximum Mean
Number of Transactions per Customer .3735 1.0893 1.4629 1.257052
Average Wait Time divided by Number of
Transactions per Customer
19.01 7.99 27.00 15.3739
Average Transaction Time divided by
Number of Transactions per Customer
1.83 5.81 7.64 6.6514
While the average number of Transactions per Customer isn’t useful by itself, it is
necessary in order to calculate the wait time and transaction time for a single transaction.
Thus it takes approximately 15 minutes and 22 seconds in wait time for a single
transaction and an additional 6 minutes and 39 seconds in processing time for a single
transaction.
Beyond the descriptive statistics it is useful to get sense of how these different categories
have changed over time, and often the best way to do so is visually. The following
graphs illustrate how these different categories have changed from FY2002 to FY2006.
They also include Confidence Intervals ( also known as Error Bars). The Confidence
Intervals on these graphs are calculated to 95%, which allows us to visually determine
which years are significantly different from each other. If two years have overlapping
Confidence Intervals in any way, then those two years are not statistically different from
each other. Conversely if there is no overlap in Confidence Intervals we can be confident
in saying that any given two years are significantly different from each other.
2 1
Graphs:
While the graph clearly shows a downward trend in the number of Customers served in
field offices, it isn’t until FY2006 that we see levels significantly lower than FY2002.
2 2
The same holds true for the Number of Transactions, as with the Number of Customers
served, there is an observed downward trend that does not become significant until
FY2006.
2 3
Average customer wait time is far more interesting in that there is a dramatic indentation
of near 15 minute average wait times in FY 2003 and FY 2004, with a dramatic and
significant increase in wait times in FY 2005 and FY 2006.
2 4
As shown, FY 2004 had a dramatic and significant decrease in transaction times from
FY2002 & FY2003 levels, with a significant increase in time by FY 2006. It is worth
noting that FY 2006 is not significantly different from FY 2002 & 2003, so it would be
correct to say that FY 2006 is indistinguishable from pre FY 2004 levels.
2 5
The total average wait time reflects the trend found in average customer wait time,
because average customer wait time makes up a majority of total average wait time. Data
for FY 2003 & FY 2004 are significantly lower than those for FY 2005 & FY 2006.
2 6
What is striking about this graph is the significant decrease in the number of CSAs and
CSRs between FY 2003 and FY 2004, a trend that results in a significant difference
between FY2004 and FY2006.
Conclusion:
It is interesting to note that while the number of customers and the number of transactions
have dropped from FY2002 to FY2006, both the wait time and transaction time are not
significantly different in FY2006 than in FY2002. The next series of analysis will be
aimed at trying to discern why.
2 7
Correlations:
Moving beyond descriptive statistics and graphs, correlations are the next most common
tool for identifying relationships between two things. The following table summarizes
the correlations between the six different categories reported in the MVD Statistics.
Number of MVD customers
Number of transactions
Average customer wait time ( DtC)
Average transaction time ( CtD)
Average customer total visit time
( DtD)
Average number of CSAs and CSRs
Number of MVD customers
Number of transactions 0.853
Average customer wait time ( DtC) - 0.284
Average transaction time ( CtD)
Average customer total visit time ( DtD) - 0.281 0.999
Average number of CSAs and CSRs 0.474 - 0.734 0.516 - 0.712
( DtC: Door to Counter CtD: Counter to Door DtD: Door to Door
In the above table only statistically significant correlations are listed. A positive
correlation indicates a direct relationship where the increase or decrease in one variable
corresponds to an identical increase or decrease in the other. The strength of that
relationship is measured on a scale of 0 to + 1, where 0 indicates that there is no
relationship and + 1 indicates there is a perfect direct relationship. Conversely, a negative
correlation indicates an indirect relationship where the increase or decrease in one
variable is the inverse or opposite of the increase or decrease in another. Similarly the
strength of a negative correlation is measured on a scale of 0 to - 1, where 0 indicates that
there is no relationship and - 1 indicates there is a perfect indirect relationship.
For example, we expect and we find that the number of MVD customers has a significant
and highly positive correlation with the number of transactions (. 853), whereas, the
number of MVD customers has a significant but weak negative correlation with the
average customer wait time (- 0.284). Finally, the positive correlation between Average
customer wait time ( Door to Counter: DtC) and average customer total visit time ( Door to
Door: DtD) is excellent example of a near perfect correlation (. 999), and it illustrates how
insignificant a role average transaction time ( Counter to Door CtD) has in average
customer total visit time ( DtD).
2 8
Moving on to the analysis the weak negative correlation between number of MVD
customers and customer wait time lends support to our initial observation that despite
decreasing numbers of both customers and transactions there are increasing levels of
customer wait time (- 0.284). One plausible explanation is the strong negative correlation
between the average number of CSAs and CSRs and the average customer wait time
( DtC) (- 0.734). We know from the graphs that were discussed earlier that there has been
a significant decrease in the average number of CSAs and CSRs from 2003 to 2006. The
negative correlation implies that it is the decreasing average number of CSAs and CSRs
that is responsible for the apparent increase in wait times.
There is a way to test to see what effect the average number of CSAs and CSRs is having
on the correlations between other variables.
* Controlling for Average Number of CSAs and CSRs
Number of MVD customers
Number of transactions
Average customer wait time ( DtC)
Average transaction time ( CtD)
Average customer total visit time
( DtD)
Number of MVD customers
Number of transactions 0.883
Average customer wait time ( DtC)
Average transaction time ( CtD) - 0.493 - 0.470 0.398
Average customer total visit time ( DtD) 0.998 0.451
As the above table illustrates, the weak correlation between number of MVD customers
and average customer wait time ( DtC) disappears when the average number of CSAs and
CSRs is taken into account. Correspondingly, the correlation between number of MVD
customers and average customer total visit time ( DtD) also goes to zero.
Conclusion:
In summary, the descriptive statistics show:
1. Average customer wait time ( door- to- counter) in field offices ( minutes) is 19: 22
minutes.
2. Average transaction time ( counter- to- door) in field offices ( minutes) is 8: 20
minutes.
2 9
3. The shortest customer wait time ( door- to- counter) in field offices ( minutes)
occurred in 2003 with a record 14: 12 minutes.
4. The longest customer wait time ( door- to- counter) in field offices ( minutes)
occurred in 2006 with a record 27: 46 minutes.
5. The shortest transaction time ( counter- to- door) in field offices ( minutes) occurred
in 2004 with a record 7: 57 minutes.
6. The longest transaction time ( counter- to- door) in field offices ( minutes) occurred
in 2003 with a record 8: 39 minutes.
The data analysis shows that the decreasing number of CSAs and CSRs appears to be
mediating the effect that the number of MVD customers served has on average customer
wait time ( door- to- counter) in field offices.
Associated Files
ADOTDATA. sps
MVDtablerevisedII. xls
Bibliography
Arizona Department of Transportation. 2007. ADOT/ MVD Automated Business Systems
Planning for the Future. In press. Prepared for ATRC.
Arizona. Legislature. 2007. Real ID Act; implementation prohibited. SB1152. 48th Legislature.
1st Regular Session.
Booz Allen Hamilton. 2003. Research for Customer- Driven Benchmarking of
Maintenance Activities. NCHRP Web Document 58. Washington, D. C.: National
Cooperative Highway Research Program, Transportation Research Board.
Oregon Department of Transportation. 2006. Annual Performance Progress Report ( APPR)
for Fiscal Year 2005- 06. Salem, Oregon: the Department.
Real ID Act of 2005 ( P. L. 109- 13, Division B; United States Statutes at Large 119 ( 2005):
231)
US Government Accountability Office. ( GAO) 2006. Opportunities for Improving the
Oversight of DOT’s Research Programs and User Satisfaction with Transportation Statistics.
Washington, D. C.: the Office.
3 0
FIGURE 1 – Data Collection Instrument
Location Plant Configuration:
Q- Matic: Yes No
Kiosk: Yes No
Customer Transaction Arrival Number of Physical Descriptors Time ticket Window Time Time of Distractor( s) Other
Number Time people with Ethnicity/ Gender/ Hair/ Attire was Number called Time of Time new arrival
customer dispensed to window 2nd calling number called to window
1
2
3
4
5
6
7
8
9
10
11
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Distractors: C- children Possible
R- reading others or H- hispanic C- Caucasian As- Asian
P- phone descriptors:
Number of Windows: Number of Windows open:
B- blonde Br- brown hair R- redhead Bk- black hair
Motor Vehicle Division Study Behavioral Checklist
Date: M T W Th F S
Second calling
OL- other language S- Spanish
Time of Arrival:
Time of Departure:
AF- African AmericanO- other m- male/ f- female