Model Driver Screening and Evaluation Program
Volume II: Maryland Pilot Older Driver Study


Chapter 3: 
Pilot Study Data Collection

The measures obtained during the Maryland Pilot Older Driver Study included low-cost "first-tier" indicators of functional status; selected computer-based tests offering potentially greater specificity and/or more objective and standardizable measurement capabilities; and questionnaire responses describing (self-reported) driving habits and health problems of study participants. The materials and procedures involved in collecting the functional performance and questionnaire data for the identified study samples are described below. Key driving abilities associated with each included measure are also noted.

Functional Screening Measures and Screening Procedures

The pilot study was designed to examine the validity and evaluate the administrative feasibility of measuring functional status as a driver screening and evaluation program activity. Thus, multiple criteria were applied in defining a battery of candidate measures of perceptual-cognitive and physical functions important for safe driving. These included brevity; low cost; and the ability to be administered by either professionals or volunteers, with limited training, in diverse settings; plus an expectation that the selected procedures would be the most valid indicators of gross changes in functional status for key driving abilities, based on a synthesis of prior research.

The goal of assembling a battery of gross indicators of functional ability deserves emphasis. There was a very explicit recognition that the selected measures could lack the sensitivity and specificity claimed by more sophisticated tests administered by specialists under controlled conditions. But the need in this research to identify the most scientifically valid and practical tools for program administrators was paramount; these considerations jointly served as the defining attributes of the gross impairments screening (GRIMPS) battery applied in the Maryland effort.

The GRIMPS battery included five perceptual-cognitive and four physical abilities measures, as described in the following pages. As an adjunct to the GRIMPS battery, one component of the Useful Field of View test protocol was included in the pilot study, under the sponsorship of the National Institute on Aging.1 To conform to the time-of-testing limitations associated in this research with "first-tier" screening, the full protocol (see page 21)--which can require 20 minutes or more to complete--was abbreviated to provide the type of functional assessment deemed most valuable in the context of this study.2 Briefly, the Useful Field of View Subtest 2 was used; this subtest, instead of examining differences in the size of drivers' area of visual attention, measured the peripheral target duration required for correct detection by a driver, at a single angle of eccentricity. This procedure is described in more detail below.

Perceptual-Cognitive Measures

Five tests were chosen to measure perceptual-cognitive abilities: the Motor-Free Visual Perception Test (Visual Closure subtest); Delayed Recall; a Scan Chart test to detect visual neglect; the Trail-Making test (Part B); and a PC-based variant of the Trails B procedure using a dynamic traffic scene instead of a blank background.

First, the Visual Closure subtest of the Motor-Free Visual Perception Test (MVPT/VC), was used to detect poor visual pattern perception and as a measure of the ability to visualize missing information. The MVPT is a multiple-choice test that measures a person's ability to visualize incomplete figures when only fragments are presented (Colarusso and Hammil, 1972).

This ability is important to the driving task, insofar as drivers must recognize a sign or other traffic control device that is only partly visible, or quickly perceive the safety threat represented by a vehicle or pedestrian that is partially obstructed (e.g., by a building or parked car) at the side of the road, and may be about to move into the driver's path.

The MVPT stimulus booklet for the Visual Closure practice item and Visual Closure stimulus items 22 through 32 are required to administer this test. To begin, the test administrator shows the examinee an example containing a practice figure and four alternative figure fragments (see figure 2). He/she points to the four alternative figures, saying, "If you finished drawing these figures, which one would look just like the one above? Please point to the correct alternative." After the examinee responds, the examiner points to the correct alternative, then to the stimulus figure, saying, "Yes (No), if we added these lines, this one would look just like this." Then the examiner proceeds to the actual test items stating "Now I'd like you to do the same thing for the figures I'm about to show you."

Figure 2.
Practice Item for MVPT/VC Showing Target Stimulus 
and Four Response Alternatives

Figure 2.
Practice Item for MVPT/VC Showing Target Stimulus and Four Response Alternatives

The MVPT/VC test includes eleven test items, each showing a target figure above four alternatives as in the example in figure 2. The only response required from the examinee is that he or she point to whichever one of four alternatives is correct. The examinee is not allowed to trace any figures. The test administrator encourages the examinee to look at all four alternatives before making a final decision. This is not a timed test, and the examinee must be given a reasonable amount of time (about 15 or 20 seconds) to answer each test item. No confirmation or feedback was given for the examinee's responses. The test administrator scores the responses by marking the appropriate space on an accompanying scoring sheet, then presents the next item, until all eleven responses have been obtained. At the conclusion of the test, the test administrator records the total number of incorrect responses given by the examinee. This procedure was designed to be completed by most examinees within 3 minutes.

The Delayed Recall test, from the Mini-Mental Status Examination (Folstein, Folstein, and McHugh, 1975), was used in GRIMPS as a measure of working memory. Working memory is important to safe driving because it allows a driver to recognize and remember the meaning of traffic control devices and roadway features; to remember and apply rules of the road and safe driving practices; and to perform navigational tasks that require the sequential recall of route-following instructions while actively searching for navigational cues and meeting moment-to-moment demands for hazard detection and vehicle control.

In performing this brief (~30 sec) test, the examinee was required to repeat back as many of a probe set of three words presented earlier in the Cued Recall test as possible, when requested by the test administrator. As per MMSE procedures, the Cued-Recall test was used to confirm understanding of the 3-word probe set that the examinee was required to remember for later recall. The probe set ('bed', 'apple', 'shoe,' or one of three equivalent alternative word sets) was initially presented by the test administrator, who stated, "I'm going to say 3 short words now as a memory test. Please repeat them back to me." If the examinee could not repeat all 3 words, they were presented again, up to a maximum of 6 times. After successful repetition by the examinee, the test administrator said, "I will ask you again later to remember these same 3 words and say them to me." Delayed Recall was measured, approximately 10 minutes later after a number of intervening procedures in the test battery had been completed, by asking the examinee to recall same memory (probe) set repeated earlier. Performance was scored as the number of items recalled correctly.

The Scan Chart test developed for this project was based on the Brain Injury Visual Assessment Battery for Adults (BiVABA) ScanBoard (Warren, 1990), following the recommendation by an Occupational Therapist and Certified Driving Rehabilitation Specialist (OT/CDRS) active in the MRC that a screening procedure was needed to rule out neglect of one side of the visual field (hemianopia) while driving. This is a deficit associated with recovering stroke/cerebral vascular accidents (CVA's). The research literature also documents driving impairment resulting from age-related effects combined with the effects of visual field losses in older patients with CVA's (Szlyk, Brigell and Seiple, 1993).

Figure 3. 
Scan Chart Developed for GRIMPS

 Figure 3. Scan Chart Developed for GRIMPS

The Scan Chart developed for GRIMPS measured 140 cm by 22 cm (55 in by 8.5 in), and contained 10 common symbols arranged in 2 rows of 5 columns each (see figure 3). With the chart held at eye level, one arm length in front of the examinee by the test administrator, the examinee must identify each shape without turning his or her head (i.e., scanning is accomplished by eye movement only). Specifically, the test administrator stated, "Without moving your head, scan the chart and identify each shape you see. Please name all the shapes you see in any order that you wish." The examinee's verbal report indicates a normal scanning pattern vs. hemi-neglect. A normal scan pattern of a cognitively-intact individual may be any of three: (1) rectilinear (left to right/top to bottom); (2) clockwise; or (3) counterclockwise. Drivers with impaired scanning capabilities demonstrate disorganized, random and/or abbreviated or truncated strategies. Those with hemi-neglect often show an asymmetrical pattern, initiating visual search from the right side rather than the left and confining all search efforts to the right side. Also, whereas intact persons do not overlook or repeat a stimulus on this test, those with impaired scanning abilities may commit both of these errors. The test administrator scored performance as normal, versus erratic (haphazard pattern) or neglect (two or more shapes ignored) on this procedure. Administration time for the Scan Chart test was gauged at under one minute.

The Trail-making test, Part B, was used to measure participants' abilities to perform a directed visual search and to divide attention effectively (Reitan, 1958). This is a continuous demand when navigating a route in the information-rich, visually complex driving environments common to cities and suburbs. As per clinical applications of this procedure, Part A was administered first; as described below, this afforded examinees practice on the aspect of the test devoted to directed search for an ordered sequence of test stimuli, before introducing the divided attention aspect of the test.

Specifically, in the Trail-making test the examinee uses a pencil to connect a sequence of numbers (integers) or a mix of numbers and letters, printed on a blank piece of paper, in ascending order as quickly as possible. Part A of the test contains only numbers; Part B contains numbers and letters that must be connected in an alternating fashion. Performance is timed. If an error is made during the test, it is pointed out by the test administrator, who instructs the examinee to continue with the test from the last correct connection. The clock does not stop during error correction. In both Part A and Part B, the time-to-complete all items is the examinee's score, measured to the nearest second. Only Part B was scored for GRIMPS. Longer times connote poorer performance on this test; the maximum value scored for this procedure was 6 minutes, at which point the test was discontinued.

As applied in GRIMPS, Part A was abbreviated to contain only the numbers 1-8 instead of 1-25 as used in clinical applications. This shortened the test administration time while still providing examinees some understanding of what they would be expected to do in the Trails B procedure. First, examinees received the following instructions for the (abbreviated) Trails A procedure: "Now I will give you paper and pencil. On the paper are the numbers 1 through 8, scattered across the page. Starting with 1, use the pencil to draw a line to connect each number to the next higher number. I will time how fast you can do this. Ready? Go." After this was completed, the examiner placed a practice version of Trails B containing only four numbers and four letters in front of the examinee and said, "Now you will do the same thing, only this time with numbers and letters, like you see in this example. This time, start with 1, then draw a line to A, then continue the line to 2, then to B, then 3-C, 4-D, alternating back and forth between numbers and letters." This practice was not timed. After pointing out any errors and insuring that the examinee understood the test requirement, the test administrator said: "On the other side of this sheet of paper the numbers 1 through 13 and the letters A through L are mixed up in the same way. Starting with the number 1, draw a continuous line that alternates between numbers and letters, until you finish with the number 13. I will time how fast you can do this." The test sheet was then turned over, and the test administrator said, "Ready? Go," while directing the examinee to place his/her pencil at the starting point (number 1). The Trails B test sheet is shown in figure 4. The planned duration to provide instructions and to conduct the abbreviated Trails A procedure, the Trails B practice exercise and the actual Trails B screening measure, together was 5 to 6 minutes.

Figure 4. 
Test Sheet Used to Administer the Trail-Making Test,
Part B (cf.Reitan, 1958)

Figure 4. Test Sheet Used to Administer the Trail-Making Test, Part B (cf.Reitan,

A variant of the Trails B procedure was developed for GRIMPS to further increase the divided attention demands of the test. This "Dynamic Trails" procedure used a PC to present a similar mix of numbers and letters, but they were overlaid on a background that showed moving traffic, versus the blank background used in the paper-and-pencil version of the test. The dynamic traffic scene was stored on the hard drive of the computer as an MPEG file. Examinees' responses were registered on a touch screen using a light pen. If an error was made, an audible buzz sounded, and the computer prompted the examinee to return to the last correct letter or number in the sequence, which was identified for the examinee. This protocol was designed to remove an element of subjectivity in the way instructions are delivered and errors are corrected by test administrators.

A shorter set of test stimuli was used for the Dynamic Trails procedure, containing only the numbers 1-7 and the letters A-G. This reduced the anticipated test duration to under 3 minutes. A data file of the examinee's performance was generated which included the time-to-complete, as well as the exact time of every response during the test (including error responses); plus identifying information entered by the test administrator.

The final perceptual-cognitive measure applied to collect functional status information in the Maryland Pilot Older Driver Study was subtest 2 of the Useful Field of View battery that has been shown through prior research to be significantly related to crash involvement (Owsley, Ball, Sloane, Roenker, and Bruni, 1991; Owsley, Ball, McGwin, Sloane, Roenker, White, and Overley, 1998). Aspects of visual attention addressed by this procedure include the detection, localization and identification of suprathreshold targets in complex displays. Using a PC-based test apparatus that displays central and peripheral targets within a 35-degree radius visual field, three variables can be manipulated--target presentation duration, the competing attentional demands of the central and peripheral detection tasks, and the salience of the peripheral targets.

The complete Useful Field of View protocol includes three subtests that, together, provide a measure of the percentage reduction of the maximum 35-degree radius field. An abbreviated version of the first subtest, which measures processing speed capability and vigilance, was used as practice for the second subtask, which measures divided attention capabilities; this was the specific measure of interest for this study. During the practice, examinees were required to identify a centrally-located object which varied in duration, by pressing an icon of a truck or a car (whichever was presented) on the touch-screen display, after the target was presented. Subtest 2 required this same identification; but in addition, the examinee was required to locate a simultaneously-presented peripheral target of varying eccentricity (these could appear in eight locations, spaced near the edge of the computer screen every 45 degrees around the central target) as quickly as possible. The response format for subtest 2 is presented in figure 5. Subtest 3 was not used in this research. Differences in response latency measured via a touch screen for targets presented in a visual field of constant size was therefore the dependent variable acquired in the Maryland research, not differences in the size of drivers' "useful field of view" per se.

Figure 5. 
Response Format for the Useful Field of View Subtest 2

Figure 5. Response Format for the Useful Field of View Subtest 2

Examinees used a light pen to identify central targets and to locate peripheral targets. They were given a chance to practice with the device before being tested-an important control to minimize differences between drivers on Subtest 2 due to a "practice effect." The instructions for the example exercise were as follows: "For this test, you will see an object--either a car or a truck--inside of the box in the middle of the screen. The object will stay on for a short period of time and then disappear. You will need to decide whether it was a car or a truck and then touch the light pen on the car or truck icon on your screen in the answer section." After the examinee completed the practice session, he/she was given the opportunity for more practice or to begin the test. The instructions for Subtest 2 were as follows: "The car or truck icon will again be presented in the middle of the screen, and will disappear. After the object disappears you will see another object presented on one of the 8 spokes radiating from the center of the screen. You will need to identify whether the object in the center was a car or a truck, by touching your answer with the light pen, and then touch the light pen to the location where the second object was presented on one of the 8 spokes." The length of the display varied in duration, depending upon performance; it was adjusted until a 75% correct threshold was attained. Accordingly, the examinee's score which was the briefest duration of peripheral target presentation that could be correctly identified, in milliseconds; this measure was recorded to a data file by the computer. It was anticipated that the instructions, practice, and completion of Subtest 2 as described above could be accomplished in approximately 5 minutes.

Physical Measures

Measures of physical ability incorporated into GRIMPS included tests of lower limb strength, endurance, and coordination; and upper body flexibility. These procedures, as they were administered in the Maryland Pilot Older Driver Study, are described below.

Leg strength, endurance, and coordination were measured using two different but related procedures: the Rapid Pace Walk and the Foot Tap tests. Each procedure was designed to be completed in less than 1 minute. The physical abilities targeted in these tests were those needed to sustain pedal control without fatigue and to quickly and accurately shift back and forth from the accelerator to the brake pedal when circumstances demand it. Also, a gait that is slowed significantly could be indicative of a higher risk of falling, which is related to crash risk as well (Marottoli, Cooney, Wagner, Doucette, and Tinetti, 1994).

For the Rapid Pace Walk, an examinee walked along a 3-m (10-ft) path marked with tape on the floor then returned along the same path to starting point, as quickly as possible. The instructions to the examinee were as follows: "I want you to walk along the side of this tape line to the end, turn around, and walk back here as quickly as you can." The test administrator demonstrated the walk and path, then said: "I am going to time you. Go as fast as you feel safe and comfortable. If you use a cane or walker, you may use it if you feel more comfortable. Ready, begin." Timing started when the examinee picked up the first foot, and stopped when the last foot reached start/finish point. The total time to traverse the 3-m (10-ft) path up and back (total of 6 m [20 ft] walked) was manually recorded by the test administrator, using a stopwatch.

For the Foot Tap test, a 75-mm (3-in) binder was used. The open binder was placed on the floor in front of a chair, where the examinee was sitting, oriented such that the rings were crosswise in front of the examinee, and such that the examinee could place his/her foot flat on the floor beside the rings while seated in the chair with the (right) foot extended slightly forward of the knee. The examinee was instructed: "Please place your feet on each side of this binder. Now move your left foot under the chair so it will be out of the way." The test administrator then tapped across the rings, alternately touching the floor on each side, as an example, while continuing with the following instructions: "When I say go, move your right foot back and forth, lifting it over the rings, alternately tapping the floor on each side of the binder. Tap each side five times for a total of 10 taps. I will time how quickly you can do this. Ready? Go." The test administrator manually recorded the time to complete the foot tapping task with a stopwatch.

To measure upper body flexibility, the GRIMPS battery included an Arm Reach and a Head-Neck Rotation test. Each procedure was designed to be completed in less than 1 minute. Upper body flexibility is needed to turn the steering wheel quickly in an emergency, and to look to the sides and over the shoulder to the rear to check for approaching traffic when merging and changing lanes.

For the Arm Reach test, an examinee was asked to raise each arm as high as possible over his/her head. To pass, the arm had to be raised so that the elbow was above shoulder height. Instructions for this procedure were as follows: "Please raise your right arm as high as you can over your head. You may put your arm down... Now please raise your left arm as high as you can over your head." The test administrator recorded "pass" and "fail" scores manually, for each arm.

For the Head Neck Rotation test, the examinee sat in a chair equipped with a seat belt; this restricted his/her ability to pivot at the waist to look to the rear instead of turning the head, neck, and upper torso, as required when driving. The test administrator checked the seat belt to insure that it was tightly fastened, then moved to a position 3 m (10 ft) behind the examinee at a pre-marked location and held up a cardboard clock face with the hands set to either the 3:00 or 9:00 position. The instructions to the examinee were as follows: "Just as you would turn your head and upper body to look behind you to back your car or change lanes, please turn and read the time on the clock face I am holding behind you." If the examinee could not turn far enough in one direction to read the clock, he/she was asked to try turning the other way. The test was scored as a "pass" if the examinee could read the clock. The test administrator manually recorded the "pass" or "fail" outcome.

Mobility Questionnaire

Self-report data describing driving habits and providing categorical estimates of the level of driving exposure were obtained for the study samples in the Maryland Pilot Older Driver Study using a "mobility questionnaire." The Mobility Questionnaire sought information about avoidance of specific driving situations, plus mobility-related health issues such as falls, or difficulty in walking or in climbing stairs, that have been correlated with crashes in previous research (Sims, Owsley, Allman, Ball, and Smoot, 1998; Koepsell, Wolf, McCloskey, Buchner, Louie, Wagner, and Thompson, 1994; Marottoli, Cooney, Wagner, Doucette, and Tinetti, 1994). The data collection instrument is shown in figure 6.

The intent behind the development and application of the Mobility Questionnaire was to characterize the study samples in terms of their self-imposed limitations in the amount of miles driven, and/or in the situations they choose to drive, and if possible, to identify relationships between such self-regulating behaviors and the indicators of functional status and the safety measures obtained in the pilot study. Notwithstanding inherent problems with the precision and reliability of self-report data, the responses of the older drivers to the questionnaire items could provide valuable insight into the extent to which safety problems linked to functional decline might be mitigated by self-regulation.

The Mobility Questionnaire was administered after completion of the functional testing battery for the License Renewal and Medical Referral samples, and either before or after functional testing for the Residential Community sample depending upon whether participants in this study group had to wait to be screened. Responses were recorded on the paper form shown in figure 6, then were entered into a spreadsheet program for later analysis.

Figure 6. 
Mobility Questionnaire

Figure 6. Mobility Questionnaired

1 Provided for use in the Maryland Pilot Older Driver Study by the Roybal Center for Applied Gerontology at the University of Alabama at Birmingham, in collaboration with Western Kentucky University.

2 As per Dr. Karlene Ball, Director, Roybal Center for Applied Gerontology.