ISO 26022:2010

INTERNATIONAL ISO
STANDARD 26022
First edition
2010-09-01
Road vehicles — Ergonomic aspects of
transport information and control
systems — Simulated lane change test to
assess in-vehicle secondary task demand
Véhicules routiers — Aspects ergonomiques des systèmes de
commande et d'information du transport — Essai du changement de
voie simulé pour évaluer la demande de tâche secondaire à bord du
véhicule
Reference number
©
ISO 2010
PDF disclaimer
This PDF file may contain embedded typefaces. In accordance with Adobe's licensing policy, this file may be printed or viewed but
shall not be edited unless the typefaces which are embedded are licensed to and installed on the computer performing the editing. In
downloading this file, parties accept therein the responsibility of not infringing Adobe's licensing policy. The ISO Central Secretariat
accepts no liability in this area.
Adobe is a trademark of Adobe Systems Incorporated.
Details of the software products used to create this PDF file can be found in the General Info relative to the file; the PDF-creation
parameters were optimized for printing. Every care has been taken to ensure that the file is suitable for use by ISO member bodies. In
the unlikely event that a problem relating to it is found, please inform the Central Secretariat at the address given below.

©  ISO 2010
All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying and microfilm, without permission in writing from either ISO at the address below or
ISO's member body in the country of the requester.
ISO copyright office
Case postale 56 • CH-1211 Geneva 20
Tel. + 41 22 749 01 11
Fax + 41 22 749 09 47
E-mail copyright@iso.org
Web www.iso.org
Published in Switzerland
ii © ISO 2010 – All rights reserved

Contents Page
Foreword .iv
Introduction.v
1 Scope.1
2 Terms and definitions .1
3 Lane change test (LCT).3
3.1 Principle and overview.3
3.2 Participants .4
3.3 Equipment .4
3.4 Scenario design.5
3.5 Experimental design .8
3.6 Procedure.9
3.7 Performance measures and data analysis.11
3.8 Validation .14
Annex A (normative) Instructions to participants .15
Annex B (informative) Experimental plan.18
Annex C (informative) Background and rationale .20
Annex D (informative) Adapting LCT to a real vehicle .23
Annex E (normative) Calculation of LCT metric using an adapted path trajectory for each
participant .24
Annex F (informative) Interpretation of LCT measures.32
Annex G (informative) Initial validation of LCT.38
Bibliography.40

Foreword
ISO (the International Organization for Standardization) is a worldwide federation of national standards bodies
(ISO member bodies). The work of preparing International Standards is normally carried out through ISO
technical committees. Each member body interested in a subject for which a technical committee has been
established has the right to be represented on that committee. International organizations, governmental and
non-governmental, in liaison with ISO, also take part in the work. ISO collaborates closely with the
International Electrotechnical Commission (IEC) on all matters of electrotechnical standardization.
International Standards are drafted in accordance with the rules given in the ISO/IEC Directives, Part 2.
The main task of technical committees is to prepare International Standards. Draft International Standards
adopted by the technical committees are circulated to the member bodies for voting. Publication as an
International Standard requires approval by at least 75 % of the member bodies casting a vote.
Attention is drawn to the possibility that some of the elements of this document may be the subject of patent
rights. ISO shall not be held responsible for identifying any or all such patent rights.
ISO 26022 was prepared by Technical Committee ISO/TC 22, Road Vehicles, Subcommittee SC 13,
Ergonomics applicable to road vehicles.
iv © ISO 2010 – All rights reserved

Introduction
Many advances are being made in introducing a wide range of information, communication, entertainment and
driver assistance systems in motor vehicles. Navigation aids, emergency messaging systems and wireless
communication, including e-mail and internet access, are all possible. Since many of these features require
the driver's attention it is important to recognise that, on one hand, these systems provide information and
assistance but, on the other hand, have the potential to distract the driver as well.
The lane change test (LCT) described in this International Standard, is a dual-task method that is intended to
estimate secondary task demand on the driver resulting from the operation of an in-vehicle device in a
laboratory setting. The method is simple and inexpensive so that it can be used by vehicle manufacturers,
in-vehicle device manufacturers, and other organizations.
The driver behaviour and attentional demand principles embodied in the LCT only apply to the operation of a
typical passenger car, as the vehicle dynamics model, driver eye height, and lane change dimensions and
geometries are scaled for such vehicles.
The test procedure specified in this International Standard uses software to set up the LCT task on a
computer, and to calculate the primary task performance measures. Appropriate software is available from the
ISO Central Secretariat.
INTERNATIONAL STANDARD ISO 26022:2010(E)

Road vehicles — Ergonomic aspects of transport information
and control systems — Simulated lane change test to assess
in-vehicle secondary task demand
1 Scope
This International Standard describes a dynamic dual-task method that quantitatively measures human
performance degradation on a primary driving-like task while a secondary task is being performed. The result
is an estimate of secondary task demand.
The method is laboratory based, and this International Standard defines the method, the minimum
requirements for equipment to support the method, and procedures for collecting and analyzing data derived
from the method.
The method is applicable to all types of interactions with in-vehicle information, communication, entertainment
and control systems; manual, visual, haptic and auditory, and combinations thereof. Secondary tasks requiring
speed variations to be performed cannot be tested with this method. It applies to both Original Equipment
Manufacturer (OEM) and aftermarket in-vehicle systems. It also applies to systems either portable or
integrated into the vehicle. The driver behaviour principles, the specific task procedures and driving task
correspond only to the operation of a passenger car.
2 Terms and definitions
For the purposes of this document, the following terms and definitions apply.
2.1
adaptive model
reference path trajectory adapted to each participant
2.2
baseline
test condition with the primary task only
2.3
basic model
nominal reference path trajectory is the same for all participants
2.4
calibration task
type of reference task used for the purpose of comparing different tests or test results between sites, or over
time at a given site
2.5
course
path along which the simulated vehicle actually travels
2.6
dual task
two tasks concurrently performed, primary task plus a secondary task
2.7
environment
physical surroundings in which data are captured
2.8
goal
system end state sought by the driver and which is meaningful in the context of a driver's use of an in-vehicle
system
EXAMPLE Obtaining guidance to a particular destination; magnification of a map display; or cancelling route
guidance.
2.9
integrated system
two or more in-vehicle devices, which provide information to, or receive output from, the driver of a motor
vehicle, whose input and/or output have been combined or harmonized
EXAMPLE 1 An in-vehicle entertainment system and route guidance system which use the same visual, manual and
auditory interface.
EXAMPLE 2 An in-vehicle entertainment system whose auditory output mutes when a mobile phone call is made or
received.
2.10
lane change
lateral displacement of a vehicle from current lane to another lane, including crossing one or two lanes
2.11
lane change task
series of prescribed lane change manoeuvres that are the main component of the primary task in the lane
change test
2.12
outlier
an observation that lies outside the overall pattern of a distribution
2.13
manufacturer
organisation or person designing, developing, producing, integrating and/or supplying in-vehicle equipment
2.14
path deviation measure
mdev
difference between a reference path trajectory and an actual driven course
2.15
portable system
nomadic device
device which provides information to, or receives output from the driver of a motor vehicle, that can be used
within the vehicle without installation or can be rapidly and easily installed in and removed from the vehicle
2.16
primary task
course following and manoeuvring control activity which a participant performs throughout the duration of a
test (simulated substitute for driving)
2.17
reference task
a standardized secondary task which can be used to compare different levels of performance degradation
2 © ISO 2010 – All rights reserved

2.18
run
driving used to collect LCT data, typically consisting of 18 lane changes accomplished over a 3 minute period
in either single task or dual task conditions
2.19
secondary task
interaction with an in-vehicle information, communication, entertainment, or control system, carried out
concurrently with the primary task
2.20
secondary task demand
sum of perceptual, cognitive and motoric activity required by a secondary task
2.21
single task
one task (primary or secondary task) without additional activity (as opposed to dual task)
2.22
task
process of achieving a specific and measurable goal using a prescribed method
NOTE Ultimately, it is for the users of this International Standard to determine tasks that are meaningful in the
context of a driver's use of an in-vehicle device.
EXAMPLE 1 Obtaining guidance by entering a street address using the scrolling list method, continuing until route
guidance is initiated (visual-manual task).
EXAMPLE 2 Determining where to turn based on a turn-by-turn guidance screen (visual task).
2.23
track
three-lane, straight, simulated roadway
2.24
trial
investigation of one participant undertaking one repetition of one secondary task
3 Lane change test (LCT)
3.1 Principle and overview
The lane change test (LCT) is a simple laboratory dynamic dual task method that quantitatively measures
performance degradation in a primary driving task while a secondary task is being performed. The primary
task in the LCT is a simulated driving task which resembles the visual, cognitive and motor demands of
driving.
In the LCT, a test participant is required to do a primary task consisting of driving at a constant, system-limited
speed of 60 km/h along a simulated straight three-lane road containing a series of lane changes defined by
signs displayed on a screen. Simulated vehicle position is controlled by means of a steering wheel.
Participants are instructed in which of the lanes to drive by signs that appear at approximately regular intervals
on both sides of the track. The LCT is performed by participants according to pre-test instructions contained in
this International Standard (see Annex A). The method may be implemented in a laboratory, in a driving
simulator, in a mock-up or in a real vehicle.
There is no limitation to the definition of a secondary task according to this International Standard as long as
the secondary task is compatible with the LCT procedure.
EXAMPLE Secondary tasks requiring speed variations to be performed are not accommodated.
3.1.1 Application of the LCT
In a typical application of the LCT, the primary task performance degradation resulting from a certain
secondary task can be compared to the performance degradation resulting from a reference task
(see Annex B). A reference task is a standardized secondary task which can be used to compare different
levels of performance degradation. Such standardized reference task can also be employed to compare
different test sites or to verify consistency in repeated testing.
For product development purposes, the LCT can also be used to compare alternative HMI candidates or
solutions rather than comparing each of these solutions to a reference task (see Annex C).
3.2 Participants
Participants shall be licensed drivers having a similar level of familiarity with the secondary task under
investigation. Other relevant characteristics of the participants shall be recorded (gender, age, driving
experience and previous experience with the LCT). At least 16 participants shall take part in the evaluation of
a single secondary task or in the comparison of two or more secondary tasks for a within-subject design.
3.3 Equipment
3.3.1 Display of visual driving scene
The LCT visual driving scene shall be realised with a monitor or projector with a net refresh rate of at least
50 Hz. A minimum resolution of 1024 × 768 pixels with a colour depth of 24 bit (also called True Color) is
required. The size of the display device must meet the requirements defined in 3.4.3.
3.3.2 Testing environment
The illumination level in the testing environment shall be appropriate to the secondary task. Simulated engine
sound is optional and, if used, shall be adjusted to a low level.
Participants shall be comfortably seated directly in front of the primary task visual display while performing
each test. The seat shall not swivel or rock. A seat belt shall be used if it is assumed that the use of a seat belt
might influence the test results (e.g. restrictions in posture and reach).
3.3.3 Steering wheel and simulation characteristics
For the lateral control of the simulated vehicle, a computer game steering wheel can be used for a
laboratory/monitor setup. For seating buck or real vehicle testing, movement of the steering wheel shall
provide signals to replicate movements of the game steering wheel. The steering wheel force displacement
characteristics shall be approximately linear, and the tangential force at the rim to turn the steering wheel shall
be no more than 20 N. A centering (breakout) force, which is optional, shall be no more than 12 N. Adapting
LCT to a real vehicle is discussed in Annex D.
The overall transport delay between an initial steering input and a visual display response shall be < 120 ms
(not including the simulated vehicle model delay).
The steering wheel sensor shall have a resolution less than 1,5 degrees.
With the steering wheel turned 90 degrees from the straight ahead position, the simulated vehicle shall, at a
constant speed of 60 km/h, complete a 360 degree turn in between 13 s and 17 s, corresponding to a turning
circle of 70 m to 90 m in diameter.
3.3.4 Vehicle dynamics
The simulated vehicle shall be programmed to respond approximately like a typical passenger car in terms of
its size and inertial properties. Yaw rate to steer angle characteristics shall be represented by a first order lag
4 © ISO 2010 – All rights reserved

with an equivalent time constant of approximately 0,15 s (±0,03 s), including vehicle mathematical model
response and other simulation transport delays (see 3.3.3).
3.3.5 Measurement recording
All vehicle parameters (i.e. position on track, steer angle, heading angle, heading (yaw) rate, etc.) are
recorded at a sampling rate of at least 10 Hz. Additional parameters and information such as a unique time
stamp, track number and markers for tasks (typically set manually by the experimenter) shall also be
recorded. The LCT simulation shall have a minimum update rate of 100 Hz. In addition, secondary task
performance shall be recorded (see 3.7.3). These data are needed to ensure the participant's compliance with
the instructions (see Annex A).
3.4 Scenario design
3.4.1 Simulated roadway and surroundings
The simulated roadway (see Figure 1) consists of a straight, three-lane track. A series of lane changes are
defined by roadside signs placed at approximately regular intervals. The road is located on a plain green field.
Figure 2 shows the dimensions of the lanes and the markers (delineators). The road surface is grey (like a
typical paved road). Lane markings are white.
The simulated roadway shall be displayed as if viewed from an eye point 120 cm ± 10 cm over the road
surface. This is consistent with that of a typical passenger car.
NOTE 1 The roadway and lane change geometries and the vehicle dynamics are not scaled to correspond to a truck or
other large commercial vehicle.
Minimum track length shall be 3 000 m corresponding to 3 min of driving at 60 km/h. This is sufficient in length
to collect 2 min of LCT data.
NOTE 2 According to the signs in Figure 1, the participant has to change from the current (middle lane) to the right
lane.
Figure 1 — Simulated roadway
Dimensions in metres
3,85
0,14
3,85
1,50
3,85
11,20
Key
1 road edge line
Figure 2 — Dimensions of the road markers (not to scale)
3.4.2 Signs
Both “START” signs and “Lane Change” signs appear in pairs on both sides of the simulated roadway
(see Figure 1 for example of sign positions). “Lane Change” signs are rectangular and have a thin black
border and black symbols on a white background (see Figure 4). The signs are 2,0 m wide and 1,0 m high
and the lower border is 1,0 m above the ground. The downwards arrow on the “Lane Change” sign indicates
the target lane. In Figure 4, (a), (b), and (c) show the left, middle and right lanes, respectively. Dimensions of
the “Start” sign can be the same as the “Lane Change” sign, but the colours shall be different from those of
the “Lane Change” signs (see Figure 3). The “START” sign shall appear approximately 150 m before the first
“Lane Change” sign.
Figure 3 — “Start” sign
Figure 4 — “Lane Change” signs
6 © ISO 2010 – All rights reserved

3.4.3 Experimental and measurement setup
3.4.3.1 Participant's view of display
The horizontal viewing angle to the display for the road scenery (monitor or screen) shall be between 20° and
55°. The eye-to-display distance shall be no less than 60 cm. The horizon of the visual scene shall be
between −5° and +5° from the participant's eye point height.
3.4.3.2 Steering wheel location
The steering wheel shall be located where it does not interfere with the visual scene of the LCT. It shall be
capable of being handled comfortably.
3.4.3.3 Standard scenario
The standard scenario for a 3 000 m test track is as follows:
⎯ only the scripted instructions are given to the participants (see Annex A),
⎯ speed is specified and limited to 60 km/h, resulting in a duration of about 180 s per track,
⎯ there are 18 pairs of “Lane Change” signs along a track.
3.4.3.4 Lane change sign spacing
The lane change signs are always visible but blank until the lane indications on the signs appear (i.e. pop-up)
at a distance of 40 m before the signs. The mean distance from sign to sign is 150 m (a minimum of 140 m
plus an exponentially distributed random variable with a mean of 10 m), so that the mean duration between
two lane changes is about 9 s (at a speed of 60 km/h).
3.4.3.5 Ordering the lane change directions
The number of the six possible lane changes (from left lane to middle lane, from left lane to right lane, etc.)
shall be balanced within 18 pairs of signs. This balancing also applies if tracks longer than 3 000 m are used.
If the total number of pairs of signs on a track is not a multiple of 18, balancing shall be done to the extent
possible. The presentation order of lane change signs in tracks used for different runs shall be randomised to
avoid learning effects. At least five different orders shall be used randomly for different runs.
3.4.3.6 Secondary task equipment
The equipment for evaluating the secondary task shall be positioned where the participants can properly
interact with it. The position depends on the purpose of the test, e.g. systems intended for vehicle use shall be
positioned in their intended locations in the vehicle relative to the participant.
3.4.3.7 Document the setup
Because experimental setting effect may have an impact on the results, description of the setup, viewing
distance and angle to road scene and type of road scene presentation, testing environment (personal
computer, buck or complete car), and type of seat shall be reported.
It has been shown that better lane change trajectories have resulted when using a driving simulator instead of
a personal computer [12], presumably due to the increased visual realism in a driving simulator. This
emphasises the need to describe precisely each equipment setup before comparing the results of different
experimental contexts.
3.5 Experimental design
3.5.1 General
A within-subject design shall be used (see Clause B.1). This allows for each participant to experience both the
single and dual task conditions and facilitates statistical comparison between the two conditions. Furthermore,
the experimental design shall be designed to minimize carry over or training effects between the two
conditions (for an example see Figure 5).
The secondary task(s) is (are) conducted in the dual task runs. The secondary task can either be performed
block wise (the same secondary task repeated within one track is a blocked run) or in a mixed design
(different secondary tasks given within one track is a mixed run). See Annex B.2 for rationale. Dual task runs
[primary task plus secondary task(s)] are conducted between baseline runs.
Each participant performs identical secondary tasks of equal difficulty but with different presentation order.
The order of secondary tasks, the track parts used and the combinations between secondary task and track
parts used shall be counterbalanced or randomised.
Start and end points of each secondary task operation are recorded, as well as driving performance during
single and dual task conditions for later data analysis. Start point is typically as soon as the verbal instructions
have been given. End point is the last operation made by the participant on the secondary task (e.g. a button
press) which leads to a predefined final state.
Total time on task for each secondary task shall be no less than 2 min. This means that at least 12 lane
changes during which the secondary task is being performed shall be taken into account.
During the LCT in dual task conditions, the experimenter shall monitor participant behaviour to ensure that the
participant is following the instructions and engaging in both tasks as instructed.
If the participant does not follow the experimental instructions (as described in Annex A) this part of the data
shall be excluded from data analysis. Furthermore, if the participant does not perform the secondary task, this
shall be recorded and considered in the evaluation of the interface design.
3.5.2 Baseline runs
In the experimental design, and once training is completed, baseline run data are collected at the beginning, in
the middle and at the end of the experimental phases. If only 2 baseline runs are carried out to avoid overly
long experimentation time, these runs shall be at the beginning and at the end of the experimental phases.
Mean deviation values (mdev) in the baseline runs can later be compared to check for learning effects. The
participant is free to use both hands for steering. One track with 18 signs (about 3 min) shall be used for each
baseline run.
A participant shall accomplish at least one baseline run at the beginning of the experimental phase without a
lane change error (wrong lane or missed lane change). This will ensure an adequate level of skill and training
(see Annex A). It will also provide at least one baseline suitable for calculating the participant's adaptive model
(see Annex E). To establish baseline performance for comparison with dual task performance, the mdev
values from several error-free baseline runs can be averaged.
3.5.3 Blocked runs
Given a blocked design, either the same secondary task with different parameters or a new task of the same
type is repeated within a single run for a duration of at least 2 min (see Figure 5).
With a blocked design a driving distance of 3 km (corresponding to 3 min) should be sufficient to collect data
for at least 2 min, with time for instructions for each secondary task item considered.
8 © ISO 2010 – All rights reserved

3.5.3.1 Mixed runs
In a mixed design, different secondary tasks (ST) are repeated within a single run (see Figure 5). With a
mixed design, each secondary task shall be repeated enough times to record 2 min of data for each
secondary task. This will require more than one run if there is more than one secondary task.
For analysis, the combination of all of the results of a given secondary task are used to calculate a single
mean deviation.
3.6 Procedure
3.6.1 Participant instructions and training
All participants shall receive identical instructions. These shall be read to the participant. As described in
Annex A, the participant shall receive:
⎯ information on the general purpose of the test, and in particular instructions on the lane change task,
⎯ training on the primary task only,
⎯ training on the secondary tasks only,
⎯ training on the dual task situations,
⎯ instructions before first baseline run,
⎯ instructions before dual task testing, and
⎯ if required, instructions during dual task testing.
Baseline run (3)
Baseline run (3)
Dual Task Run 4 Dual Task Run 4
ST#4, ST#4, ST#4, . ST#1, ST#3, ST#4, .
Instruction ST#4
Dual Task Run 3
ST#2, ST#4, ST#2, .
Dual Task Run 3
ST#3, ST#3, ST#3, .
Baseline run (2)
Instruction ST#3
Dual Task Run 2
Baseline run (2)
ST#3, ST#1, ST#2, .
Dual Task Run 1
Dual Task Run 2
ST#4, ST#1, ST#3, .
ST#2, ST#2, ST#2, .
Instruction ST#2
Baseline run (1)
Dual Task Run 1
ST#1, ST#1, ST#1, .
Instruction on ST#1
Instruction ST#1
Instruction on ST#2
Instruction on ST#3
Baseline run (1)
Instruction on ST#4
blocked mixed
design design
Figure 5 — Example of blocked and mixed experimental design
If a participant fails to meet the baseline criterion for minimum mdev (see Clause A.2) due to improper lane
positioning, the importance of staying in the middle of the lane between lane changes shall be emphasized.
Upon request, participants shall be instructed to perform the primary task and the secondary task as
described in Clause A.4.
3.6.2 Allocation of attention in dual task conditions
3.6.2.1 General
The interpretation of LCT performance measures depends upon the way that participants allocate attention
between the primary task and the secondary task they are performing at the same time. Because the lane
change test is a divided attention method, in order for the measures generated by the LCT (mdev) to be
consistently interpretable, an assumption is made that the participant is allocating attention in such a way that
if the secondary task demand increases, it will lead to degradations in primary task performance.
However, participants may allocate attention differently than this, even when carefully instructed. If they do, it
may lead to LCT results that obscure important differences between the secondary tasks in their demands on
participant attentional resources. For example, two tasks may produce similar mdev values, but differ
significantly in the number of errors made on each secondary task, the number of extra (unnecessary)
secondary task entries made on each task, or the number of secondary tasks completed during the run [10].
Information on the interpretation of LCT measures is given in Annex F.
There are at least three different ways in which a participant might allocate attention between the tasks during
an LCT.
3.6.2.2 More attention to the primary task
Participants might allocate attention so as to try to optimize their performance on the primary task, and let the
secondary task performance suffer when they become overloaded. It is possible, for example, that because
the demand associated with the lane changes is approximately periodic (rather than continuous), a participant
might use a strategy in which activity on the secondary task is suspended during the lane change, and
resumed when the lane change is completed. Such a strategy might lead to better mdev scores, but degrade
secondary task performance, by slowing it down, increasing errors (if task information is forgotten or there is
difficulty resuming it), or leading to extraneous control inputs or display search time.
3.6.2.3 More attention to the secondary task
Participants might allocate attention so as to maintain or optimize their performance on the secondary task
and let their performance on the primary task suffer. Under this attentional strategy, overload from dual task
demands would become apparent in mdev measures, while secondary task performance may not degrade
from that observed in a single task condition.
3.6.2.4 Attention to both primary task and secondary task
Participants are asked to manage both tasks concurrently according to the instructions in Annex A
(e.g. through frequent task-switching), with similar attentional allocations to both tasks. Under this attentional
allocation, degradations in performance would be expected to show up on both tasks when their combined
demands exceed a participant's momentary resource capacity. Therefore, both mdev measures and
secondary task performance measures would be expected to show degradation.
3.6.2.5 Interpretation of dual task results
Attentional allocation is not yet fully understood, and it is not clear that instructions or incentives will ensure
that the allocation of attention between tasks will be the same for all participants. The literature on dual-task
methods has demonstrated that results are best interpreted when measurements of both tasks performed
concurrently are obtained together with measurements of each of the tasks done in a single task condition.
10 © ISO 2010 – All rights reserved

3.6.3 Test procedure details
The main purpose of the lane change test is to compare performance degradations resulting from each of two
or more secondary tasks. One of the secondary tasks could typically be a reference task (see 3.1). The test
shall be performed as a “within-participant” procedure. This means that, for each participant, performance
shall be established and compared under dual task conditions for runs with the secondary task(s) under test,
and for runs with reference task(s) if appropriate.
The experimenter shall give the participant instructions on when to start the secondary task (see A.6). Start
and end point of operation of the secondary task shall be recorded. A secondary task starts immediately after
the experimenter's verbal instruction is finished, or when the participant starts performing the secondary task,
whichever event occurs first. If there is a delay between end of instructions and participant start of secondary
task, start of task shall be considered to be at end of instructions.
3.7 Performance measures and data analysis
3.7.1 Path deviation measures of secondary task demand
Two path deviation measures can be used to quantify the effect of secondary task demand on primary task
performance. These path deviation measures can be calculated using an adaptive model (adaptive mdev) or a
basic model (basic mdev). The adaptive model calculates a reference path trajectory (adapted curve) for a
baseline run for each participant. The adaptive model and method shall be used to quantify and compare task
demand for various dual task conditions. The basic model can also be used to compare a participant's path
performance to a nominal reference path trajectory that is the same for all participants. Use of the basic model
is optional.
These mean path deviation measures cover important aspects of the participant's performance. These include
detection (late detection of a sign or missing a sign), quality of the manoeuvre (slow lane change results in
larger deviation), and lane keeping quality, all of which may result in an increased deviation.
3.7.1.1 Adaptive model
The adaptive model shall be used to quantify the differences between experimental conditions. The effect of
secondary task demand is measured by the adaptive mdev which is the mean deviation between the actual
driving course of the participant along the track and the reference path trajectory from the adaptive model (see
Figure 6 for a symbolic example of the model path and actual driving data). This reference path is adjusted to
each participant in order to better match their actual driving behaviour in response to the signs in the baseline
task. The result of the adaptation calculation is that the start of the lane change, the lane change length, and
the overall lateral position in the lane (as shown by the solid line in Figure 6 and the solid red line in Figure 7)
are adjusted to match the baseline behaviour of the respective participant. The details of these adaptation
procedures and the calculation of the resulting adapted path trajectory are given in Annex E.
3.7.1.2 Basic model
The basic model is similar in concept and application to the adaptive model, except that the reference path
trajectory is nominal and the same for each participant (see Figure 8). The measure of the effect of
secondary-task demand is the basic mdev which is the mean deviation between this nominal basic path model
trajectory and the actual driving course of the participant along the track (see Figure 6). The nominal basic
path model trajectory is based on the sign geometry and spacing as described in 3.4.3.4. Once the instruction
content of the sign becomes visible (pops-up), a reaction time of 600 ms to initiate the lane change is
assumed. A participant would cover a distance of 10 m within this time (16,67 m/s × 0,6 s = 10 m). Therefore,
the basic model trajectory turns 30 m before a lane change sign to take reaction time into account. The basic
lane change has a length of 10 m regardless of whether it is a single or a double lane change, as shown in
Figure 8.
3.7.2 Performance calculation using the adaptive mean deviation value
The mean deviation (adaptive mdev) between the adaptive model path trajectory and the participant's actual
driving course shall be calculated as the performance parameter of main interest. Only the section(s) of the
track in which the participant is performing the respective secondary task shall be used for analysis, i.e. those
sections where the experimenter gave instructions are excluded from the analysis if such instructions last
more than 1 s.
To facilitate this, the relevant sections shall be marked by the experimenter during each run with appropriate
markers in the recordings. Trials with secondary tasks which were not completed (due to participants inability
or time limits), however, can be included in the analysis as long as the participant worked on the secondary
task to their best ability, as instructed.
NOTE According to the signs, the driver has to change from the middle to the right lane and then to the left lane.
Note that Figure 6 is not to scale (driving direction is from left to right). The average path deviation is the total area
between the reference path trajectory of the adaptive or basic model and the driving course (m ) divided by distance
driven (m).
Figure 6 — Comparison of reference path trajectory (solid line) and actual driving course (dotted line)

Key
1 sign content pop-up (40 m) 4 adapted right lane position
2 adapted left lane position 5 adapted start of lane change
3 adapted centre lane position 6 adapted lane change length
NOTE: Dashed line shows an incorrect lane change.
Figure 7 — Adaptive model trajectory parameters relative to sign content pop-up and sign position
12 © ISO 2010 – All rights reserved

Dimensions in metres
10 10 20
Key
1 sign content pop-up
Figure 8 — Basic model trajectory values relative to sign content pop-up and sign position
The mean deviation (mdev) between either the adaptive or basic model path trajectory and the actual driven
course can be calculated as follows:
⎡⎤
1 ⎛⎞yy−
ii+−11
x (1)
⎢⎥deviation,i⎜⎟
∑
S 2
⎝⎠
⎣⎦
xx=−x
deviation,iiposition, reference,i
where
x is the lateral deviation;
deviation,i
S is the length of the data segment analysed;
y is the longitudinal vehicle position along the track;
x is the actual vehicle lateral position on the track;
position,i
x is the lateral position of the reference (adaptive or basic) path trajectory.
reference,i
This calculation is done across all data segments which are relevant for a certain experimental condition. In
other words, the invalid data segments (e.g. time for instructions by experimenter) are removed and the
remaining valid data are handled as if they were one continuous set of data. The same applies to an
experimental design where secondary tasks of one experimental condition are distributed over several
experimental runs. In baseline runs mdev is calculated for the section between the start sign and 50 m after
the last of 18 lane change signs.
The resulting mdev values shall be calculated for each participant and each experimental condition of interest,
for example, for baseline driving, for dual task conditions with each secondary task under test, and with a
reference task.
3.7.3 Measures of secondary task performance
Because participants may allocate attention in multiple ways during the LCT, measurements of primary task
performance (adaptive mdev) and secondary task completion time, number of secondary task completions
and number of errors, shall be obtained. These secondary task measures are described in Clause F.1.2 and
their relevance will depend on the type of task.
3.7.4 Analyzing differences between dual task conditions
An evaluation is accomplished by comparing mdev values. This can involve comparing performance in two
different dual tasks, or comparing dual task performance with the baseline condition. The mdev differences
can be statistically analyzed (e.g. ANOVA, t-Test, Friedman Test, etc., depending on the specific design and
research question). All statistical parameters relevant to the comparison shall be reported (e.g. t-values,
F-values, p-values, degrees of freedom, sample size, and outliers). In data diagrams, error bars with standard
errors are recommended. The data should be examined for outliers because they can influence the
distribution of mean deviation. Any treatment of outliers shall be reported.
In general, while comparisons of mdev values (as discussed above) provide an overall measure of dual task
differences, mdev alone may not be sufficient to explain the differences between two secondary tasks with
similar mdev values, as discussed in Annex F. Similarly, not finding a statistically significant difference
between mdev values does not prove that the tasks are equivalent. Additional performance variables for the
primary task and the secondary task are discussed in Annex F and can be considered in order to provide a
more complete assessment of those tasks. These measures can also help to determine whether the main
performance degradations from resource overload are reflected in the primary task measure (mdev), or
whether some changes in dual task performance are reflected in secondary task measures.
3.8 Validation
Initial validation results are summarized in Annex G. Data are shown from a study comparing dual task
performance with eight secondary tasks of varying levels of difficulty, using both LCT and an over-the-road
vehicle. The results show a correlation between mdev in LCT and standard deviation of lane position over-the-
road. The results of driving simulator studies are also discussed. Additional validation studies should be
carried out, as experience is gained with the application of LCT.
14 © ISO 2010 – All rights reserved

Annex A
(normative)
Instructions to participants
A.1 Before training on the primary task
Outline for instructions to participants prior to primar
...