You are on page 1of 61

M.Sc.

PROJECT REPORT

2008

Psychometric testing for Indian train drivers: The


relationship between experience, age and performance

Joseph Knowles

Project report submitted in part-fulfilment of the requirements


for the postgraduate degree in Ergonomics (Human Factors)

© Loughborough University
and
Joseph Knowles

The views expressed in this report are not necessarily those


of Loughborough University

i
Psychometric testing for Indian train drivers: The
relationship between experience, age and performance
Joseph Knowles
2008

ABSTRACT
There is a commonly held belief that experience in any skill based area, such as driving,
will improve performance. This is supported by various research into the area. In
regards of age, there is a split in the literature, with some suggesting that intellectual
performance declines with age and others suggesting that it can be avoided. Generally
the biggest decline is most apparent in those over the age of 60. In one area, reaction
time, research suggests there will be a decline in later middle ages. This paper
investigates the relationship between experience, age and performance on psychometric
tests. Data from test performance of 448 Indian train drivers were combined with a
control group of 51 non-train drivers who undertook the test battery. This test battery
includes tests on complex reaction time, form perception, a group Bourdon test
(involving continuous addition), speed perception, and a visual differentiation test. An
overall test score was also analysed. From using ANOVA tests it was found that there
were significant effects of age and experience on reaction time, form perception, and the
group Bourdon test. Experience also had a significant effect for speed perception; age
had a significant effect on the visual differentiation score and for the overall score.
Tukey post hoc tests revealed that for two of the tests the control group scored
significantly higher than those with experience; in one test those with over 10 years
experience performed better; and in one test those with any experience performed better
than the controls. For age those in the age groups up to and including 45 years of age
performed significantly better than those in the three groups from 46 to 60. These
results have been discussed, along with limitations and implications, stressing the
importance of inclusive design for older adults based on ergonomic principles, and
recommending companies consider other factors before promoting people purely based
on experience.

ii
ACKNOWLEDGEMENTS

The author would like to acknowledge Dr Neil Mansfield and Dr O. P. Srivastava for
their supervision and support, Shyam Dev Ram for his technical support, Mr R. K. Lal
for providing the provision of equipment, staff and accommodation, those at RDSO
who took part in the study, and those directors who gave their staff the time to take part.

This research was supported by the European Commission. EC project reference


number: Asia Link-ASIE/2005/111000.

iii
TABLE OF CONTENTS
 
Abstract.............................................................................................................................ii 
Acknowledgements..........................................................................................................iii 
Table of contents.............................................................................................................iiv 
1  Introduction............................................................................................................... 1 
1.1  Safety and the railways in the UK and India .................................................... 2 
1.2  Driving .............................................................................................................. 3 
1.3  Psychometric testing ......................................................................................... 4 
2  Literature Review ..................................................................................................... 5 
2.1  Experience ........................................................................................................ 5 
2.1.1  General experience and performance ........................................................... 5 
2.1.2  Experience and car driving ........................................................................... 6 
2.1.3  Experience and specific skills....................................................................... 7 
2.2  Age.................................................................................................................... 8 
2.2.1  Age and reaction time ................................................................................. 10 
3  Aims and Objectives............................................................................................... 11 
3.1  Experience ...................................................................................................... 11 
3.2  Age.................................................................................................................. 11 
4  Methods .................................................................................................................. 13 
4.1  Participant characteristics ............................................................................... 13 
4.2  Materials ......................................................................................................... 13 
4.3  Experimental procedure.................................................................................. 15 
4.4  Analysis .......................................................................................................... 18
4.5 Ethical Approval……………………………………………………………...18
5  Results..................................................................................................................... 19 
5.1  Complex Reaction Time ................................................................................. 19 
5.2  Form Perception.............................................................................................. 20 
5.3  Group Bourdon Test ....................................................................................... 22 
5.4  Speed Perception........................................................................................... 244 
5.5  Visual Differentiation ..................................................................................... 24 
5.6  Overall Scores (Mean scores) ......................................................................... 25 
5.7  Summary......................................................................................................... 26 
6  Discussion............................................................................................................... 27 
6.1  Experience ...................................................................................................... 27 
6.1.1  Poor performance by experienced drivers ................................................ 288 
6.1.2  Good performance of the control group ..................................................... 29 
6.1.3  Experience and expertise ............................ Error! Bookmark not defined. 
6.2  Age.................................................................................................................. 30 
6.2.1  Age and the hypothesis ............................................................................... 31 

iv
6.2.2  Sleeping and aging...................................................................................... 32 
6.2.3  Age, physiology and attitude ...................................................................... 33 
6.2.4  Age and personality .................................................................................... 35 
6.3  Implications .................................................................................................... 35 
6.4  Limitations ...................................................................................................... 36 
7  Conclusion .............................................................................................................. 39 
8  References............................................................................................................... 41

Figure 1 Picture of testing equipment 13


Figure 2 Graph showing reaction time and experience 19
Figure 3 Graph showing reaction time and age 20
Figure 4 Graph showing form perception and experience 21
Figure 5 Graph showing form perception and age 22
Figure 6 Graph showing group Bourdon and experience 23
Figure 7 Graph showing group Bourdon and age 23
Figure 8 Graph showing speed perception and experience 24
Figure 9 Graph showing visual differentiation and age 25
Figure 10 Graph showing overall (mean) scores and age 26
Figure 11 Picture of testing equipment close-up 46

Table 1 Summary of results 26


Table 2 Reaction time raw scores and corresponding t-scores 47
Table 3 Form perception raw scores and corresponding t-scores 48
Table 4 Group Bourdon raw scores and corresponding t-scores 49
Table 5 Speed perception raw scores and corresponding t-scores 50
Table 6 Visual Differentiation raw scores and corresponding t-scores 51
Table 7 Visual Differentiation raw scores and corresponding t-scores 52

Appendix A The complex reaction time test 46


Appendix B The form perception test 48
Appendix C The group Bourdon test 49
Appendix D The speed perception test 50

v
Appendix E The visual differentiation test 51
Appendix F The controls needed and sought table 53
Appendix G Letter for distribution around RDSO 54
Appendix H Control group consent form 55

Word Count: 10,605

vi
1 INTRODUCTION
In both the United Kingdom and India psychometric testing is used to test potential
train drivers; the results on these tests give an indication of an individual’s abilities
relating to train driving and therefore help to aid employers when choosing successful
and safe candidates. In India they are used prior to deployment on high speed trains.

The rail system itself is a classic domain for ergonomics, as work from all types of
different areas, including psychology, can be researched and developed with a view to
improving human-machine interaction. These areas include cab design, control room
design and planning, for example, and involve those users directly involved in the
system such as drivers and passengers. In the UK a good deal of research was carried
out into rail human factors in the 1960’s and 1970’s; this research did continue into
the 1980’s and 1990’s, but this was a small amount compared to other transport
ergonomics research (Wilson and Norris, 2005). Since the mid-1990’s this has
changed, with the Chief Engineer of Network Rail highlighting the influence of
radically changing public and government perceptions and relatively fast technical
systems in an industry where nothing much has changed for 250 years (McNaughton,
2003). He believed ergonomics could bridge the gap between what the railway system
could offer and what the industry and society want.

Research into the various areas associated with the rail system has been concerned
with improvements in the business itself with a recent upsurge in passenger demand
and also into areas of safety. Serious accidents in the UK as well as ongoing problems
in other parts of the world have meant that safety has become the main focus for
businesses to improve. As a consequence of this much work in ergonomics has been
carried out focussing on the safety critical aspect of the railway system. Further afield
in India, safety is also seen as the most important facet of a train organisations
domain, and continuing effort is made to improve safety. As will be highlighted, one
of the key areas of any rail system is the driver.

1
1.1 Safety and the railways in the UK and India
For many years, investigations into accidents and incidents concerning safety critical
systems identified the ‘human element’ as the major contributing factor (Pasquini et
al., 2004). Statistics from the UK rail industry attribute around 70% of accidents to
human error (Hall, 1999). There were 23 reported fatalities for the year 2007 from
various aspects of the train service (excluding suicide), which most would agree is far
too many (Rail Safety and Standards Board, 2008).

The Rail Safety and Standards Board’s (RSSB) role is to provide leadership in the
development of the long-term safety strategy and policy for the UK rail industry. It is
responsible for managing railway group standards; it measures and reports on safety
performance; and it provides safety intelligence, data and risk information to inform
safety decisions on the railways (Mills, 2005). The RSSB are involved in a broad
range of human factors research including work on signals, and work on driver visual
strategies.

In India safety is seen as a key business theme, recognising it as a special attribute,


significant not only for passenger traffic but also for the transportation of materials in
today’s highly competitive environment (Sharma and Amitabh, 2004). India has one
of the largest railway structures in the world with over 63,000 kilometres of track,
over 1.4 million in the workforce, and running over 13,000 trains per day. Safety is
taken seriously and has improved dramatically over the last 50 years. Despite this, in
2005-2006 there were 234 train accidents in total, including both passenger and
freight trains that were both serious and relatively minor, and 177 fatalities related to
those incidents. Like in the UK, many of these incidents are attributed to human error:
88% in 2005/06, up from 84% in 2004/05. Therefore, whilst improvements have been
made over a period of many years, it appears management in India has still not
mastered either the current issues, or the ones developing as their economy grows and
technology improves.

The principle organisation responsible for research into safety and technological
improvements in India is the Research Designs and Standards Organisation (RDSO)
2
which was established in 1957 under the Ministry of Railways in Lucknow. RDSO is
the sole research and design organisation of Indian Railways and functions as the
technical advisor to the Railway Board, Zonal Railways and Production Units. RDSO
performs functions such as development of new and improved designs; development,
adoption, absorption of new technology for use on the railways; and inspection of
critical safety items such as rolling stock and signalling. They also have a Psycho-
Technical department concerned with the drivers and other staff operating within the
railways system where assessment of cognitive abilities takes place (psychometrics),
alongside biofeedback tests. This allows RDSO to test drivers regularly to ensure they
are fit to drive and to measure potential improvements or decline in abilities over
time.

1.2 Driving
Thus far, systems in both the UK and India have been described and presented
alongside a series of facts concerning accidents and incidents over recent years. From
this, one key theme has emerged: that is, safety is a serious and important facet of any
rail system and that human error accounts for a large number of the aforementioned
accidents and incidents. Therefore, drivers need to be considered. Furthermore, the
aspects that define the drivers’ ability must be explored, alongside consideration of
how these abilities are measured.

There are many aspects that influence a drivers’ performance. In an investigation into
train driver visual strategies, for example, it was found that different designs of
signals are responded to differently by drivers, and argued that this is where the focus
should lie, as train driving is a predominantly visual task (Luke et al., 2006).
According to McLeod et al. (2005), strategic behaviours, situation analysis,
distributed cognition and attention all have important roles to play in driving
performance. Furthermore, Hamilton and Clarke (2005), suggest memory, perception
and anticipation are all key features. Jansson et al. (2005) used a cognitive work
analysis (CWA) in order to apprehend the constraints imposed in the train driver.
Initial analysis showed that the driver has to assimilate information from a number of
sources which sometimes create attention conflicts. However, the main weakness was
3
not too much information; it was the drivers struggling to find relevant information.
The drivers sometimes found themselves in an information vacuum.

Therefore as shown, various aspects of cognition and intelligence have been linked
with train driving. In both the UK and India psychometric testing aims to investigate
drivers’ aptitude in several of these areas.

1.3 Psychometric testing


Intelligence tests are a good way to measure ability and any tests accurately
measuring this could be indicators of good performance. It follows that those that are
better drivers will perform better in aspects of intelligence tests relevant to driving.
Likewise, those that perform well in tests should be better drivers and thus tests can
be used to predict good performance. One way to measure intelligence and
personality is to use psychometric tests. Psychometric tests are concerned with
factors, which is a mathematical concept derived from statistical analyses of multiple
correlations among groups of tests (Colman, 2003). The factors are then given
meaningful names such as verbal ability or perceptive ability. The number of factors
considered to account for intelligence varies from one (Spearman, 1923), to one
hundred and fifty (Guildford, 1967). In this case however, general intelligence is not a
measure needed, and it would therefore be a case of formulating the measures needed
such as perception and attention to measure driving performance.

One of the main criticisms of psychometrics is that they fail to provide any
understanding of the underlying cognitive processes: they tell us what the different
abilities are, but not how they work (Kahney, 2003). This is an area of interest to
cognitive psychologists, but in a practical working environment in the case of driving,
it may be more important to know what factors are involved in improved performance
before trying to understand the processes underlying them.

4
2 LITERATURE REVIEW
As can be seen therefore, one of the most important aspects of the rail system is
safety. From data from both the UK and India, a large percentage of accidents and
incidents are attributed to human error. Drivers have the main role in this of course
and there have been various suggestions as to what aspects or abilities are crucial for a
train driver. Furthermore, psychometric tests have been developed and are in use for
recruitment and development purposes in the UK and India, and these tests aim to
indicate which aspects or abilities a particular driver excels at or performs poorly in.
So what other aspects are important in determining how well a driver may perform in
these tests? This leads to the focus of the present study.

2.1 Experience

2.1.1 General experience and performance

Experience is something gained over a period of time and can often lead to
improvement of skills and expertise in any given area. Those who have experience
with a certain type of task (such as driving a train) will have learnt about the
intricacies involved and will have established rules and routines, as well as further
developing skills in that task. When people begin to learn about tasks they generally
start out learning what they have to do. This is called declarative knowledge because
it consists of declarative statements, such as knowing that red means stop (Colman,
2003). The other type of knowledge is procedural knowledge and consists of a
building up of knowing parts to form a skilled whole pattern (Colman, 2003). With
practise, skilled performance becomes better integrated and speeds up considerably
and it has been suggested that 10 years is the number at which someone will gain
expertise in skill-based activities (Kahney, 2003). This suggests that experience in a
particular domain will lead to improved performance. The experience acquired over
time may form the basis of the complex knowledge of that domain. As well as years
on the job however, it may be that the quality of those experiences may also be a
factor in differentiating experts from novices (Seifert et al., 1997).

5
Rasmussen (1987) created a model of human control and behaviour called the skill-
rule-knowledge-based framework. This suggests that during training in a particular
task, such as driving, control moves from the knowledge or the rule-based levels
towards the skill-based level. Therefore, there is a reduction in the mental/cognitive
workload required for the operations involved in the driving task, meaning that a
larger amount of attention can be directed towards other tasks or operations. The level
of available attention the driver has at any given moment is partly dependent on the
driver’s prioritisation between different tasks, whether primary or secondary. This
work makes an interesting point relating to distraction, but also suggests that
experience will lead to an intrinsic shift, allowing the driver to attend to more whilst
in the cab using honed skills and thus perform better. Furthermore, there has to be an
ability to plan for future situations for any train driver, which could be described as
‘opportunistic planning’ (Hayes-Roth and Hayes-Roth, 1979; Birnaum and Collins,
1984). This is the ability to think of a number of opportunities and plan for them with
varying degrees of abstraction. This ability will improve will practice and experience
and again, therefore, should result in improved performance.

2.1.2 Experience and car driving

The majority of research into driving has been into that of motor vehicle driving and
its focus primarily on younger drivers and accident statistics. It is relevant however as
there a number of parallels between driving a car and driving a train.
Methodologically, it has been suggested that it is difficult to separate experience and
age in relation to accident statistics, as there are many more 17 year olds with six
months experience as there are 24 year olds with six months experience (Clarke et al.,
2006). It has been found that accidents decrease as time passes from full licence
attainment (Waller et al., 2000), however, age and experience correlate highly in this
case and it is therefore difficult to differentiate between them. Jonah (1986) observed,
“the attempt to separate the two concepts may prove fruitless”. Efforts have been
made to separate the two though, with some suggesting that with new drivers, whether
the license age is 16 or 18, a two-year learning curve can be seen in the years after the
licence has been acquired (Twisk, 1996).

6
There has been more recent research that has found in the UK that novice drivers are
twice as likely to have an accident as middle-aged experienced drivers (Underwood,
2007). The same study suggested that drivers’ understanding of the driving tasks
develops with experience, such that roads that demand increased monitoring receive
more extensive scanning than roads that are simpler. Novice drivers do not show
sensitivity to road complexity, with experienced drivers performing better, and trained
police drivers performing better than that. Interestingly, it was found that training
intervention could help novice drivers improve their performance. Whilst for train
drivers tracks do not become more complex as such in regards to lots of traffic, there
are simpler parts (one line through the countryside), and more complex parts
(approaches to busy stations; station yards in India), and this study suggests that those
with more experience will perform better in these situations, and that training might
be of benefit to inexperienced drivers.

2.1.3 Experience and specific skills

Whilst it may be difficult to differentiate age and experience, there has been research
investigating the specific skills involved in driving. For example, experienced drivers
have been found to have faster reaction times to hazards than novice drivers
(McKenna and Crick, 1991; Quimby and Watts, 1981; Sexton, 2000), although,
differences have not been found for all hazard perception tests (Chapman and
Underwood, 1998; Crundall et al., 2003; Sagberg and Bjornnskau, 2006). In a study
on experience and cognitive workload it was found that non-experienced, low-
mileage drivers had at least 250 milliseconds longer reaction time to a peripheral
stimulus than experienced drivers (Patten et al., 2006). The authors reported a
statistically significant difference in cognitive workload between the two sets of
drivers and concluded that more resources should be invested in training and
experience for inexperienced drivers.

7
It has been shown also that trained novices perform better on hazard perception tests
than untrained controls (McGowan and Banbury, 2004; McKenna and Crick, 1994;
McKenna at al., 2006; Mills at al., 1998). Therefore hazard perception, which can
predict crash risk, increases with driving experience and is susceptible to training.
This has implications for both road safety and train safety.

The arguments on driving performance and experience seem to suggest that


experience can improve performance of drivers. Furthermore, various pieces of
research highlight specific aspects of driving which may be susceptible to training and
therefore may help to improve performance. Assuming that psychometric tests are a
valid and accurate representation of real-world skills, it can therefore be proposed that
experience will also lead to improved performance in these tests. In other words,
training and experience should improve performance on psychometric tests and this
therefore indicates that training and experience will improve the performance of
drivers.

Age will now be considered to explore the research on intelligence and declining or
improving abilities as this, alongside experience, is an important factor in the
performance of train drivers and relevant to Indian railways whereby the people
involved tend to be in the middle-aged bracket. Age is also important when
considering training needs and recruitment.

2.2 Age
There are links to aging and intellectual decline. Primary mental abilities are
intellectual abilities that can be studies as groups of related skills (e.g. memory) and
may often improve until the early 40’s before declining in the 50’s (Kail and
Cavanaugh, 2007). Fluid intelligence consists of abilities related to flexible and
adaptive thinking and these skills may decline during adulthood (Horn, 1982).
Crystallised intelligence reflects knowledge acquired throughout life experience and
education in a particular culture and it has been suggested that these skills improve
during adulthood until late life (Horn, 1982).

8
Middle age is the time when people tend to develop expertise (Dixon et al., 1985).
This suggests that there is some variation across age in mental ability. However, it has
been suggested that there are not massive changes in cognitive function until older,
retirement age (Woods, 1999). This has been supported by cross-sectional studies, as
well as longitudinal studies such as the Seattle Longitudinal Study which found that
the steepest decline appeared after the age of 60 (Schaie, 1996). This study also found
however, that many not only maintain intellectual performance but also improve as
they age.

As can be seen therefore, it appears that fluid intelligence does appear to decline
throughout adulthood and that crystallised intelligence does not decline and can
compensate for these other intellectual declines. There is no conclusive argument to
these views, as some experts (“optimists”) see that intellect need not decline into later
life as various measures can be undertaken to reduce the damage, including this
stability or improvement in crystallised intelligence (Baltes and Schaie, 1974). Other
experts (“realists”) however disagree with this and state that decline in intelligence is
an empirical fact that needs to be accepted (Horn and Donaldson, 1977).

The Wechsler Adult Intelligence Scale (WAIS) is one of the most influential
measures of general intelligence today. It includes a verbal scale which focuses on
learning knowledge including comprehension, arithmetic, and vocabulary; and a
performance scale which focuses on ability to solve puzzles involving blocks and
pictures. As people grow older their verbal scores remain stable, but their
performance scores decline (Sattler, 1982). This is referred to as the classic aging
pattern. However, critics argue that intelligence is a more complex, multidimensional
capacity than these tests measure (Gardner, 1985). This issue of validity and how
much these tests relate to real life is ongoing and continuous.

9
2.2.1 Age and reaction time

Generally, then, a mixed picture is presented in terms of declining general abilities


over the lifespan. It has been suggested that some aspects do change and that they
made be compensated for by other factors. One facet that may change is that of
reaction time. In a study completed on drivers’ reaction time and age, participants
were split into three groups with ten drivers in each (Makishita and Matsunaga, 2008).
These three groups were: 20-29, 41-54, and 61-64. This split in age groups was added
to try and highlight differences in performance, and several tests were carried out on
reaction times. It was found that the more difficult the task, the greater the difference
related to age. It was suggested that every drivers ability declines with age but the
speed of that decline differs amongst individuals. Therefore, differences in individual
ability become large according to that age and this may explain large individual
differences in terms of reaction times. Other studies in reaction time have shown that
information processing slows with age and that older adults do not perform as well as
younger adults in these tests (Salthouse, 1985b).

10
3 AIMS AND OBJECTIVES

3.1 Experience
From the literature discussed it has been suggested that experience improves
performance of driving. Psychometric tests aim to replicate real life cognitive
challenges and therefore it can be expected that experience will improve performance
on these types of tests, especially for those with more than 10 years experience.
Specifically, the literature suggests reaction time is faster for experienced drivers
compared to inexperienced drivers and experienced drivers also perform better at
hazard perception tests.

The objective of the study in this case will therefore be to test drivers with various
degrees of experience plus a control group with no experience to see differences. It
can be expected in this current study that those with more experience will perform
better at psychometric testing, will record significantly higher scores for all tests but
especially that which tests reaction time. This has important consequences for training
administered to staff, and by implication may help to ensure safer drivers are
operating on the railway.

3.2 Age
Literature on aging has produced mixed messages. Overall however, it can be
generally summarised that most cognitive changes associated with aging occur over
the age of 60. In India 60 is the retirement age and consequently no staff are working
past this age on the train network. Reaction time, on the other hand, appears to be one
facet of human ability that does decline through the years of adulthood.

The objective of the study in this case will therefore be to test drivers and non-drivers
in various age categories to see if there are differences. It can be expected in this
current study that age will have no bearing in any psychometric test scores other than
that of reaction time which will decrease for the older groups of drivers. Overall
scores will be similar due to compensation from crystallised intelligence factors. This
11
has important consequences for training of different age groups, and may highlight
aspects of the rail system which could be altered from an ergonomic perspective to be
more inclusive for older drivers.

Tests will also be carried out to discover if there is any interaction between experience
and age; tests run to see if one has an effect on the other or vice-versa. This may help
to shed some light on whether the two can be adequately separated and have
implications therefore for further research.

12
4 METHODS
In order to determine whether psychometric scores improve with experience, and to
determine whether age has an effect on psychometric scores, a laboratory study was
conducted using a computerised test system at RDSO, Lucknow, India.

4.1 Participant characteristics


All of the participants involved in the study were employees of RDSO. All were
Indian and male. 448 current drivers who had previously completed the psychometric
testing as part of the company’s routine and structure were used. To this 51 controls
who are non-drivers were added. There was a mean participant age of 50.96 (6.23).
The 499 participants were all given a unique reference code but names were not
recorded.

4.2 Materials
The equipment used to test various abilities at RDSO is the Computer-Aided Drivers’
Aptitude Test (CADAT). This system was developed for screening of locomotive
pilots prior to deployment on high speed trains. A photograph can be seen in Figure 1.

Figure 1 – The testing equipment

13
From job analysis a total of fifteen psychometric criteria were produced. From these
five tests were selected by RDSO as they were deemed most relevant to train drivers.
It is a test battery that is used as RDSO in this case would not accept the validity of
one piece of information; rather, they use five tests, plus a personality test not used in
the current study, to form an all-round view of an individual’s ability (Kaplan and
Saccuzzo, 1997). These tests are:

1. Complex reaction time (speed of reaction, agility). See Appendix A.


2. Form perception (perceptual speed, discrimination of forms). See Appendix B.
3. Group Bourdon (concentration). See Appendix C.
4. Speed perception (perception of moving vehicles, attention). See Appendix D.
5. Visual differentiation (sustained attention). See Appendix E.

Job analysis including critical incident interviews, observation, task inventory’s,


competency profiles, visionary interviews, and profiling were undertaken at RDSO to
establish the selection criteria which formed the basis for selection of these tests.
These selection criteria were found to be very similar to those used in the UK
(Occupational Psychology Centre, 1993). From these British Rail established four
tests that they use to test their drivers as opposed to RDSO’s five.

The tests are taken systematically by many train drivers all over India, working for
any one of the 16 railway zones. The results of each test are mapped to produce a ‘t-
score’. The test was completed originally by 100 drivers which produced a normal
distribution curve. From this, as dependent on the needs at the time, it was deemed
that a t-score of 46 or over would equate to a pass of the test. This corresponds
approximately with the 30th percentile, with the aim being to reject the worst cases, as
opposed to accepting the best. Those that record an overall score t-score of 46 or
above become eligible to be drivers for the high speed trains (over 110 kph). Those
that do not record a high enough score continue in their current posts driving a variety
of trains up to 110 kph. The higher the t-score the better the performance, be that as
more correct responses, fewer incorrect responses, or a lower reaction time, dependant

14
on the test. Each test maps on to a t-score differently due to what it measures.
Examples of how they map, taken from the current data set, can be seen in
Appendices A-E.

4.3 Experimental procedure


Pre-recorded test data from the 448 drivers was taken and directly inputted into SPSS.
Also included was age and experience in years. For experience data was not already
on the system. Therefore, information for all 448 drivers was sent to their
corresponding line managers in all 16 divisions across India by fax. The managers
then completed the forms indicating when a particular driver took the step from being
a co-pilot to actually driving the train and returned the fax. However, only
information for 270 drivers was achieved this way. To calculate the experience for the
others, the mean time of service in their current post was calculated (this data was on
the system), and this (3.44) was subtracted from the mean experience of the 270
(12.93). This therefore meant adding nine years to experience of those whose data
was not returned by line managers. Nine years was assumed to be the average length
of time it takes for drivers to start driving and move through to be locomotive pilots
and take their tests. This was confirmed by RDSO staff who agreed anecdotally that
this number is reasonable.

From this accumulation of test scores, age and experience the latter two groups were
subdivided into categories. For experience the groups were:
1. Control group with no train driving experience (n = 51)
2. 0-9 years of train driving experience (n = 30)
3. 10-12 years of train driving experience (n = 226)
4. 13+ years of train driving experience (n = 192).
This was done as research suggests that after 10 years expertise is achieved in most
skill based activities (Kahney, 2003). Therefore, it was hypothesised that there would
be a difference between the groups. Another separation is made after 12 years to see if
those with much more experience make any difference to scores.

15
From the data five age groups were also established:
1. Up to and including 40 years old (n = 35)
2. 41-45 years inclusive (n = 77)
3. 46-50 years inclusive (n = 61)
4. 51-55 years inclusive (n = 194)
5. 56-60 years inclusive (n = 132).
This was done to be able to see differences across varying middle-age groups. This
also represented the age groups which the drivers fell in to. From this 51 controls
were tested, weighted with the main data. For example, in the 0-40 group there were
30 drivers, representing 6.6% of the experimental group. Therefore 6.6% of 51 (4)
controls were sought for that age bracket (see Appendix F).

For the controls, individuals were brought to the testing area (the laboratory of the
Psycho-Technical Department, RDSO) with only a vague idea of the project aims and
objectives. This information was available to them but was not given initially due to
language barriers. The sample was based on opportunity and therefore came from a
variety of different departments. The Psycho-Technical department released a letter
(Appendix G) which was then used to get people to take part. In all cases their
speciality was administration and they may have been more familiar with
computerised tasks. Once in the testing room the participants filled in the consent
form (Appendix H) and then filled in their data on the computer with the aid of a
Hindi speaking technician who was able to explain the project and answer any
questions. Once all data was inputted into the system the test began in either English
or Hindi, according to their preference. Whilst the instructions were in one language
or the other, the tests themselves did not have a language component and were
identical for all. Each participant had an unlimited amount of time to complete the
tests and therefore had time to work through any examples given before each test as
many times as they wanted. This enabled the participant to get used to the equipment
and the buttons and commands used in each test.

16
As well as the same instructions and procedure for each participant, the environment
was similar for each person and the temperature was controlled using a climate
control facility. In this way it was ensured that test bias was minimised. Furthermore,
the testing room was the same used for any driver who arrives to take the test. The
room is away from any corridors and consequently free from most distracting noise.

4.4 Analysis
After the data from the controls was collected the test scores were entered into SPSS.
There were six scores overall for each participant (five tests plus the overall, or
average, score); as well the participants reference number, age and experience. As
noted experience in years and age in years were sub-divided into groups. The scores
formed the dependent variables, the sub-divided age and experience groups formed
the independent variables.

The repeated measure of analysis of variance (ANOVA) was used to test for any
significant main effects and interactions of age (of five groups) and experience (of
four groups). A post hoc Tukey test was used to determine the exact nature of any
significance between the individual conditions. Statistical significance was accepted
at the 5% level (p < 0.05). The tests were run initially just using the 270 (plus
controls) whose experience data was confirmed. The nine years were then added to
the other data and tests run as a whole.

Due to the criteria involved in the study ANOVA was the most appropriate test. To
tests differences between pairs of conditions a t-test is usually used, so it may seem
logical to run multiple t-tests of the levels involved in this study. However, by doing
multiple tests the chance of errors increases essentially as 5% accepted for each test is
multiplied and therefore increases (Goodwin, 2005). So, an ANOVA is used to look
for overall significance that could exist somewhere amongst the variables. Post hoc
tests then determine where that significance lies. The ‘F’ scores produced examine the
extent the obtained mean differences could be due to chance or are the result of some
other factor. Tukey post hoc test was used as it is a popular test and one which clearly
highlights direction of difference. Various post hoc tests are available but Tukey was
17
deemed acceptable in this case. Graphs also help to highlight differences, if found,
and therefore using a simple post hoc test in this case is appropriate.

4.5 Ethical Approval


This project has been approved by the Loughborough University Ethical Advisory
Committee. Approval was also given from RDSO.

18
5 RESULTS
The tests were run initially on the 270 participants (plus controls) whose experience
data was known. The results were very similar to those in this section. Only one test
changed from being significant in the initial test, to not being significant in the main
test. This was due to a small change in the F number and concerned the visual
differentiation test and experience. Overall, the F numbers changed slightly; only in
this case the slight difference was enough to change the significance value.

5.1 Complex Reaction Time


There was a significant main effect of experience on reaction time scores (F (3, 479) =
9.653, p < 0.0005), see Figure 2. Tukey post hoc tests indicated that those with over
10 years experience performed significantly better that those with less experience, and
the control group (p < 0.05).

]
]
50
] ]
Reaction Time T-Scores

40

30

20

10

Control 1-9 10- 12 13- 40

Ex pe rience (Ye ars)

Figure 2 – Graph showing experience groups and reaction time scores

19
There was a significant main effect of age on reaction time scores (F (4,479) = 2.8, p <
0.026), see Figure 3. Tukey post hoc tests indicated that the two younger groups
performed significantly better than the three groups above them (p < 0.05). There was
a significant interaction between the factor of age and the factor of experience (F (12,

479) = 2.087, p < 0.016).

60
]
]

] ]
Reaction Time T-Scores

40

20

0
0-40 41- 45 46- 50 51- 55 56- 60

Age

Figure 3 – Graph showing age groups and reaction time scores.

5.2 Form Perception


There was a significant main effect of experience on form perception scores (F (3, 479)
= 8.729, p < 0.0005), see Figure 4. Tukey post hoc tests indicated that the control
group performed significantly better than those with any experience (p < 0.05). Those
with over ten years experience performed similarly; those in the 1-9 group performed
worst.

20
60
]

Form Perception T-Scores


]
]
]

40

20

0
Control 1-9 10- 12 13- 40

Ex pe rience (Ye ars)

Figure 4 – Graph showing experience groups and form perception scores.

There was a significant main effect of age on form perception scores (F (4, 479) = 2.539,
p < 0.039), see Figure 5. Tukey post hoc tests indicated that the two youngest groups
performed significantly better than the three older groups (p < 0.05) There was no
significant interaction between the factor of age and the factor of experience (F (12, 479)
= 1.382, p = 0.170).

21
60

]
]

Form Perception T-Scores


] ]
]

40

20

0
0-40 41- 45 46- 50 51- 55 56- 60

Age

Figure 5 – Graph showing age groups and form perception scores.

5.3 Group Bourdon Test


There was a significant main effect of experience on the group Bourdon test (F (3, 479)
= 11.398, p < 0.0005), see Figure 6. Tukey post hoc tests indicated that the control
group performed significantly better than those with any experience (p < 0.05).
Performance is similar for those with over 10 years experience, however those in the
1-9 year experience group performed worst.

There was a significant main effect of age on the group Bourdon test (F (4, 479) =
13.741, p < 0.0005), see Figure 7. Tukey post hoc tests indicated that the youngest
age group perform significantly better than the three groups 46-50, 51-55 and 56-60.
The 40-45 group perform significantly better than the groups above them as well (p <
0.05). There was no significant interaction between the factor of age and the factor of
experience (F (12, 479) = 1.257, p = 0.241).

22
60
]

] ]

Group Bourdon T-Scores


]

40

20

0
Control 1-9 10- 12 13- 40

Ex pe rience (Ye ars)

Figure 6 – Graph showing experience groups and group Bourdon scores.

60 ]
]
Group Bourdon T-Scores

]
] ]

40

20

0
0-40 41- 45 46- 50 51- 55 56- 60

Age

Figure 7 – Graph showing age groups and group Bourdon scores.

23
5.4 Speed Perception
There was a significant main effect of experience on speed scores (F (3, 479) = 2.834, p
< 0.038), see Figure 8. Tukey post hoc tests indicated that the control group
performed the worst in this test, with the other groups very similar. The main effect of
age on speed scores was not significant (F (4, 479) = 2.349, p = 0.053). There was a
significant interaction between the factor of age and the factor of experience (F (12, 479)
= 2.547, p < 0.003).

50
] ]
]
Speed Perception T-Scores

40

30

20

10

Control 1-9 10- 12 13- 40

Ex pe rience (Ye ars)

Figure 8 – Graph showing experience groups and speed perception scores.

5.5 Visual Differentiation


The main effect of experience on the visual scores was not significant (F (3, 479) =
2.308, p = 0.076). There was a significant main effect of age on the visual scores (F (4,
479) = 7.795, p < 0.0005), see Figure 9. Tukey post hoc tests indicated that those in the
0-40 group and the 41-45 groups performed significantly better than the other older
groups (p < 0.05). There was no significant interaction between the factor of age and
the factor of experience (F (12, 479) = 0.957, p = 0.490).

24
60
]
]

Visual Differentiation T-Scores


]
] ]

40

20

0
0-40 41- 45 46- 50 51- 55 56- 60

Age

Figure 9 – Graph showing age groups and visual scores.

5.6 Overall Scores (Mean scores)


The main effect of experience in the overall scores was not significant (F (3, 479) =
1.023, p = 0.382). There was a significant main effect of age on the visual scores (F (4,
479) = 9.868, p < 0.0005), see figure 10. Tukey post hoc tests indicated that those in the
0-40 group and the 41-45 groups performed significantly better than the other older
groups (p < 0.05). There was no significant interaction between the factor of age and
the factor of experience (F (12, 479) = 1.127, p = 0.336).

25
] ]

50 ]
] ]

Overall (Mean) T-Scores


40

30

20

10

0-40 41- 45 46- 50 51- 55 56- 60

Age

Figure 10 – Graph showing age groups and overall scores.

5.7 Summary
As many tests were run, Table 1 shows a brief summary of the results:

TEST RUN EXPERIENCE AGE INTERACTION


SIGNIFICANCE SIGNIFICANCE SIGNIFICANCE
FOUND? FOUND? FOUND?
Reaction Time Y Y Y

Form Perception Y Y N

Group Bourdon Y Y N

Speed Perception Y N Y

Visual N Y N
Differentiation

Overall N Y N

26
6 DISCUSSION
Broadly speaking, the research reviewed in the earlier part of this paper suggests that
age-related performance is stable until older, retirement age (excepting possibly
reaction times), and that those who are more experienced will perform better than
those who are not. A study was therefore designed taking 448 drivers from the
database at RDSO, India, finding their experience in years, and assessing their
performance in various psychometric tests alongside a control group of 51 with no
experience. This type of research is important for Ergonomics as we work towards a
safer, more efficient railway system, both in the UK and in India. The results of the
tests were reported and will now be discussed.

6.1 Experience
Research suggests that those with more experience will perform better than those
without (Kahney, 2003; Rasmussen, 1987; Seifert et al., 1997; Twisk, 1996;
Underwood, 2007). However this was the result in only two of the tests (reaction time
and speed perception; Figures 2 and 8), supporting current thinking in that area
(McKenna and Crick, 1991; Quimby and Watts, 1981; Sexton, 2000). In two other
tests where experience was found to have a significant effect (form perception and the
group Bourdon test; Figures 4 and 6), post hoc tests revealed that rather than more
experienced participants producing higher results, it was the control group that
performed best. The issue here is that the control group may be better skilled and
better at psychometric tests, i.e. they would all make good train drivers. At RDSO it is
widely accepted that more experience makes better drivers and this is used as a factor
in promotion from goods drivers up to locomotive drivers. However the research here
suggests that contrary to past research and popular opinion, experience does not
necessarily improve performance on psychometric tests. This provides the first
discussion point.

27
6.1.1 Poor performance by experienced drivers

In this study drivers were tested using CADAT, a computerised system designed to
test those skills necessary to train driving. From a large amount of research job
profiles were created and the tests were then used to test these skills. The tests are not
designed to exactly replicate those in the cab and so are not providing realistic settings
and are using different controls from those used on the trains. This would explain how
research and widely held beliefs about experience have not been supported here: it
could be that experienced drivers are better drivers, but this is not being supported
because these tests are not designed to exactly replicate driving conditions. Therefore,
it would be expected that those with no experience would perform better as they
would have no schema, or idea as to what should be expected. Their views of the
testing equipment would be neutral and would therefore approach it that way. On the
other hand, train drivers may arrive at the testing very much used to a different set of
controls and therefore find it harder to adjust. Furthermore, they may have preferences
about testing equipment from other train cabs and designs and therefore have a
subjective disapproving view of the testing equipment used in this study. Mental
models are an important factor to consider in any programme or equipment design
(Dul and Weerdmeester, 2001). The issue here is that the skills tested relevant to
driving are being undermined by a difference in testing equipment from real life
settings.

There is a bias called ‘belief perseverance’ (Lepper et al., 1986), whereby people can
be motivated by a desire to be certain about their knowledge so hang on to it
doggedly, despite evidence that this ‘truth’ is indeed false. In this case, it could be
suggested that those with more experience will perform better: it is a widely held
belief and may therefore be a ‘truth’ that the drivers have accepted and hold on to,
leading to complacency and a reduction in motivation and interest in proving
themselves.

28
6.1.2 Good performance of the control group

A contributing factor to the good performance of the control group may have been the
Hawthorne effect. In this, a study was undertaken to investigate the factors that
influence worker productivity in an electric plant. In the most famous of the various
experiments that took place, six female workers were selected from a larger group in
the plant to assemble relays, a task important to the company. During one period from
May 1927 to September 1928 (Gillespie, 1988), there were various changes to work
schedule, rest periods, total hours of work, and bonuses. The standard account has it
that productivity soon reached high levels and stayed there even when working
conditions were worsened. The reason given for this was that the workers remained
productive because they believed they were a special group and the focus of attention.

This may to some extent apply to the control group used in this study. For them, their
main day-to-day tasks are administrative and therefore they may have enjoyed the
change. Many groups were taken from within one department and this may have
caused an element of competition between participants, which may or may not exist
with the train drivers. They certainly could perform without the pressure that the
drivers are under and this may have contributed. The controls may well have taken on
the role of the ‘good subject’ (Orne, 1962), tending to be very cooperative to the point
of persevering through repetitive and boring tasks. In short, the drivers may find the
test boring, mundane and part of the job, whereas the controls may have been
competitive completing a novel task and therefore made a real effort to perform
throughout the 90 minute test. The controls also knew that their data would be
compared to the drivers, but the drivers were not aware when completing the test that
they would be compared to non-drivers. This effort to be a ‘good subject’ could have
also been influenced by the fact a white western student was carrying out the study. In
Lucknow, not many of these exist, eliciting near constant attention. Whilst it may
seem unlikely, it may be that some participants made an extra effort to impress the
researcher, or because they would be taking part in an international study.

29
6.1.3 Experience and expertise

It should be mentioned that despite most of the results failing to support the
hypothesis that experience would lead to an improvement in test scores, in two tests
this did actually occur, supporting current research in the area. For reaction time there
was a marked improvement for the groups 10-12 and 13+ (Figure 2), and for speed
perception those with any experience performed better than the control group (Figure
8). This supports the idea that expertise is obtained in skill-based tasks such as driving
after 10 years (Kahney, 2003). This is interesting for training needs as training may
improve performance for less experienced drivers and can help those drivers to come
up to the standard of the more experienced drivers more quickly. For high speed train
driver’s reaction time and speed perception are possibly the two tests that will come
under more stress and this is therefore a useful finding.

It should also be noted that in the cases where the control group performed
significantly better than the other groups (form perception and group Bourdon;
Figures 4 and 6), the group with 1-9 years experience performed much worse than all
the groups. This suggests that in those tests, having either no or at least ten years
experience is better than having some. This may help to reinforce the idea of schemas
taken from the train cab helping to produce poor results. The control group performed
the best, but it may be that those with over ten years experience had more ability to
adapt than those with 1-9 years experience. They have more variety in terms of their
schemas and this may arguably support research suggesting ten years is the cut-off for
expertise, despite the control group performing the best overall in those tests.

6.2 Age
Research suggests that there will be no decline in abilities for older people until
retirement age at around 60, except in one area where research suggests that reaction
times may be worse for people as they age (Baltes and Schaie, 1974; Makishita and
Matsunaga, 2008; Salthouse, 1985b; Schaie, 1996; Woods, 1999). Reaction time
research was supported in the present study as there was a significant dip after the age

30
of 45 in the complex reaction time test (Figure 3). One of the aims of this study was to
try and differentiate between age and experience and this is why age was accounted
for when control participants were sought. Consequently it is not correct to suggest
that the two are intrinsically linked and that because reaction scores improved for
those with more experience, that means those who are older (and possibly with more
experience) performed better. However, in this test and in the speed perception test
there was a significant interaction between age and experience showing some support
for the link.

6.2.1 Age and the hypothesis

In the speed perception test no significant differences were found for age, supporting
the hypothesis that scores will not deteriorate as an individual ages. Currently at
RDSO there is a general belief that aging is a serious and relevant factor that takes
hold at around 45 years of age. However, for those who feel this is slightly unfair, this
present study will be welcomed as one of the tests involved in the main psychometric
tests indicate that age is not an issue all of the time, with older participants performing
as well as younger participants. Unfortunately this data may get lost in the overall
results and other tests where performance was not so good, as shall be seen.

The overall test scores indicated a significant effect of age, with those in the three
groups 46-50, 51-55, and 56-60 showing significantly worse scores than the younger
two groups. This was a result of four tests showing significant effects of age. In all of
these tests, as with the overall scores, the two youngest groups performed
significantly better than the three older groups above them. It may seem easy to say,
then, that these older groups should have their driving hours restricted, for example,
or that they should receive additional training. However, Ergonomics as a discipline is
focussed on user-centred design and has to therefore be inclusive. It is not sufficient
to suggest that older people are not performing as well as they might so they must
have a problem. The answer lies in the system itself, whether that is organisational
(involving training lapses), operational hours, or control and cab design and layout.

31
6.2.2 Sleeping and aging

Shift work and sleeping may be a factor to consider. It has been suggested that age
and sleeping is an issue as people age the length of time they sleep reduces, as does
the quality of that sleep (Kail and Cavanaugh, 2007). Past 50, it seems that people
suffer with their sleep, and shift workers are also fighting against issues related to
their biological clock and circadian rhythms. Neurologically 50 does not appear to be
the cut off. In a study involving two groups (one with a mean age of 21.2 and the
other with a mean age of 43.9), it was shown that the older group had results
consistently lower than the younger group over several days of shifts in hourly
neurobehavioral tasks (Reid and Dawson, 2001). Furthermore, there was a significant
change in performance across shifts from day to night, and the younger group were
able to maintain performance between day shifts and the second night shift. One of
the main things to suffer from sleepiness (perhaps resulting from shift work), is the
brain and cognitive performance.

In a study looking at cognitive performance and working conditions, it was found that
both physical activity and working before 6 in the morning and or after 10 in the
evening on the previous day were significantly associated with poorer cognitive
performance (Ansiau et al, 2007). This suggests that night workers will struggle to
function properly after work, and the research also suggests that sleep fragmentation
was also a problem for the 52 year olds who suffered shorter sleep length and greater
overall sleep satisfaction. In this study, the testing was carried out during the day and
after presumably all the controls had slept that night. For the drivers however, who
made up the bulk of the data, many of them had come from all over the country
perhaps on sleeper trains, where sleeping can be hard. This may have had an effect on
driver’s performance in the tests if they were more tired than the younger drivers for
the aforementioned reasons.

32
6.2.3 Age, physiology and attitude

Two major changes occur in older people in the eye (Kail and Cavanaugh, 2007).
Firstly, there is a decrease in the amount of light allowed into the eye meaning more
illumination is needed and can cause problems with displays with poor contrast.
Going from dark to light places or vice-versa requires adaptation which can take
longer as a person gets older. The second problem with eyesight is to do with the lens
within the eye. Often older adults become far-sighted. Adjusting to near and far
objects can be difficult, so when driving, adjusting to the controls on the dashboard
and the road out of the windscreen can be difficult. Good contrast and labeling are
crucial to the success of a display (Dul and Weerdmeester, 2001), and therefore poor
design will affect the usability of the service. Issues to do with the environment were
controlled as much as possible in the room where testing took place. Climate control
was used to ensure similar temperatures and humidity for all, and the lighting was
kept constant. Whilst the author of these lines failed to be affected by glare, that does
not mean there were not issues related to it affecting the older participants.

It is interesting to note that two of the tests where older people performed less well are
the two tests that take the longest to complete. The group Bourdon test involves
searching through pages of groups of dots (hence possible problems with acuity), and
striking through those with a cluster of four dots. This task lasts for approximately 10
minutes and involves a lot of concentration. The visual differentiating task takes
approximately 30 minutes and is designed to sustained attention.

A possible explanation for poor performance could be physical and the problems
encountered with aging. Whilst undertaking the tests an unnatural body position is
adopted whereby users sit up and hunch over the control panel. This can cause local
muscle fatigue, knocking on to other parts of the body, causing pain, discomfort and a
dip in concentration (Dul and Weerdmeester, 2001). The control panel itself is
approximately two inches in height and on top of a normal desk (see Figure 1).
Therefore users must get up and over it to use it, especially when the precise controls
needed are at the back and middle of the panel. Indians in nature are shorter in stature

33
and consequently this adopted poor posture may be uncomfortable for users. Often the
arm is near to shoulder level for the duration of the test: not a posture recommended
(Dul and Weerdmeester, 2001).

As people age mobility can suffer as major muscle groups start to weaken, meaning
even simple tasks can be difficult (Rantanen et al., 1994). This issue may occur
mainly for those over 60, however declining physical abilities and movement could be
an issue earlier, perhaps especially for those who sit predominantly at a desk for a lot
of their time. The consequences of this are that if older people are more susceptible to
aches and pains, sitting in an uncomfortable position for as long as these tests require
may cause pain and therefore a dip in concentration. Tension from concentrating will
only exacerbate the issue.

Anecdotally, many members of staff mused that older people would not perform as
well in the tests due a shift in attitude of older people in India. It was said that “once
the individual has gone a certain distance in their career, they become primarily
concerned with the settlement of their children and family and become less interested
in their work”. Although this is difficult to support without honestly answered
questionnaires, it may be an issue here.

If older participants were not as concerned as to whether they passed the test (by
achieving a t-score of 46), or not then they would have struggled to maintain interest
in the longer tests, and therefore not performed as well as they could have. The issues
confronting older train drivers, such as safety and their need to keep their job to
support their family, has no bearing on whether they pass the test to become a high
speed train driver or not. They can therefore be as relaxed about as they like; their
current job is still safe. Furthermore, the t-score of 46 is so set to ‘weed-out’ the worst
drivers, rather than accepting the best, as it is set at around the 30th percentile point.
Consequently, even with a lackluster performance there is still a chance an individual
may pass anyway.

34
6.2.4 Age and personality

Following on from this point about changes in life attitude, it has been suggested that
personality may play a part in intelligence changes in later life. It has been suggested
that basic personality changes remain stable after the age of 30 (Costa and McCrae,
1980; McCrae and Costa, 1990). Longitudinal studies have shown that these traits
remain stable even when bereavement, health problems and economic setbacks are
considered (Costa et al., 1994). However, it has been suggested that some
psychological traits can be strengthened throughout life. For example, middle-aged
people with flexible attitudes are less likely to be prone to intelligence decline through
their lifespan than those who could be described as cognitively rigid (Schaie, 1984).
Thus, this difference in psychological make-up may help to explain why older people
generally performed less well overall, and in two of the tests.

6.3 Implications
Currently RDSO promote people based partially on experience. Train drivers start out
in their careers as firemen, for example, before gradually working their way up. When
they start driving they climb due to good performance and accumulation of years
experience behind them. This study suggests that this is not the ideal way to promote
people. Only reaction time and speed perception improved as a result of experience,
and in two tests experience made performance worse. This suggests some disparity
between the mental models drivers are developing in their job and the testing
equipment. This study also suggests that the widely-held belief that experience
improves performance may not be as straightforward as it appears, and that employers
should think twice before indiscriminately promoting people based purely on
experience.

As one of the tests showed no significant differences in regards to age, the view held
by some people that after the age of 46 people deteriorate should be treated with
caution. What needs consideration is that two of the tests where older groups
performed badly were long, as is the overall duration of the test. Issues here could

35
relate to inclusive design, whereby the set up of the equipment is ‘designing out’ older
users and hindering their performance. This could be physiological or psychological.
Independent tests on concentration of older people may yield different results. They
may be a more widespread attitude shift in those older participants and therefore may
need more incentives to perform well and concentrate in longer tests in order to reveal
their true abilities.

In only two of the tests an interaction was found between the age and experience
factors. This shows some support for the idea that “the attempt to separate the two
concepts may prove fruitless” Jonah (1986). However, it is not enough support to not
give hope to future researchers that the two can be analysed and conclusions drawn
without one factor significantly affecting the other. With a large sample and a large
enough range in years for both age and experience, they can be separated. Some
studies have supported this, and the ones that have suggested an intrinsic link are
normally those concerned with young motorists and accident statistics. This has
implications for future research and perhaps highlights an area where differences can
be analysed.

6.4 Limitations
The first limitation may be that of the experience. While 270 people’s actual
experience was discovered, that meant that 178 did not have necessarily the correct
amount of years for their experience. By subtracting the mean years that the
remaining 178 had been in their current post from the mean years of experience of the
270, a nine year figure was arrived at indicating that this is the approximate time for a
driver to progress through the ranks to their current post. Then that nine was added to
the length of time the 178 had been in their current post. So, for an individual who has
been their current post for two years at time of testing, their total experience is
assumed to be 11 years. Anecdotally, senior management agreed that nine years was
about right, and by running statistical analysis on the initial 270 before adding the rest
of the data, it could be seen that the results were similar and only the F numbers
changed slightly. Only one significance number changed: in the visual differentiation
test there was significance after the initial test, but due to a drop of 0.5 in the F
36
number significance was not found in the main test. However, despite this, it would
have been preferable to have the correct experience data for all the drivers. Adding
nine years to the remaining data meant that many more people moved into the 10-12
category and out of the 1-9 category. While this may be accurate; more people
probably have been a driver for over ten years, it may have precluded a more even
comparison between those with less than ten years experience and those over the
watershed mark.

Research has suggested that personality may predict how older people deal with tests
like these and how their intellect changes. However in this study, due to its size and
complexity as it was, personality was not considered. Consequently, as it is, the
results indicate that the older groups performed less well in some tests than the
younger groups and provide no causation to explain it. On the test bank a personality
test is completed and by analysing this in regards to particular groups, scores and
performance it may be possible to make some definite assumptions about the nature
of personality and performance, rather than suppose and hypothesise here.

There may have been some bias in regards to the control group and consequently a
rational to explain why those with less or no experience performed significantly better
on two of the tests. This involves the control group performing over and above
perhaps their natural ability due to a desire to impress, for whatever reason. They
were told the nature of the study beforehand and this may have led to an element of
competition to prove that they had what it takes to be a high speed train driver. This
was used as an encouraging factor to get them to take part in the first place, as was the
idea that they could compare their results against their colleagues in their offices. It
may have been beneficial to deceive, or rather, to not emphasise the competition
element of their participation. This may have led to a more natural performance and
slightly different results.

Different results may have been found had the raw scores been used instead of the t-
scores. The reason they were not used was because each person’s score was mapped
onto the t-score equally and therefore any differences would be relatively the same.

37
Another benefit was that using milliseconds for reaction time and correct responses in
the form perception test, for example, would have made it difficult to gain an overall
average score. Thus, t-scores made the analysis simpler and equal throughout, and
provided an important statistic in the overall score. Reaction time would be over a far
greater parameter (perhaps 500ms to 1500ms) and thus differences would have been
larger than the t-score scale which is from 20 to 80. On the other hand, correct
responses may have only been from 0 to 30 on the form perception test so differences
would have been smaller than the t-score scale. It may be beneficial to look at the raw
data analysis, but in this case t-scores were seen as the best way to proceed.

38
7 CONCLUSION
The aims of this study were to investigate whether experience and age had any effects
on performance in psychometric tests. Using a data bank including 448 Indian train
drivers and by collecting 51 participants for a control groups with no experience,
various statistical analyses were run on t-scores taken from six psychometric tests to
find significant effects. From a review of the literature in the area it was suggested
that experience should improve performance across the board on all tests; that age
should have no effect on the results except for one test, reaction time. From the results
significant effects were found for four of the tests, not the overall test, in regards to
experience and that in two of these tests post hoc analysis revealed that those with
little or no experience outperformed those with experience. For age, no significant
effect was found in one of the tests, but it was found that those in the two youngest
groups (up to 46 years) outperformed those older groups for four of the tests and the
overall (mean) test.

Various ideas have been explored to explain these results, including differences of
testing equipment and real life equipment, participant bias, belied perseverance, and
issues concerning the competition prior information may have created. It is
recommended that employers consider whether promoting individuals predominantly
based on their experience is wise (as happens at RDSO). A wider set of criteria taking
into account other factors concerning what their experience has taught them would be
more beneficial.

In regards to age, further investigation is needed as to why older people performed


significantly worse on the testing in this study. Some ideas regarding concentration
have been mooted and the next logical step is to discover why performance dips after
46, when research suggests that for most people dips in performance do not occur
until older, retirement age. Research involving attitudes to work in India and
performance may reveal areas which can be improved upon. A systematic, organised
review of equipment, both used for testing and that used in the cabs, should be
ergonomically reviewed to ensure the designs are as inclusive as possible. This will

39
eradicate or support the possibility that older users are being designed out and thus not
performing as well. Consideration of older drivers needs in regards to sleeping should
be considered as well as any reduction is physical ability.

Finally, the CADAT system or any other future system needs to be as close to the
system used in the cabs in India as possible, or the tests must be as relevant as
possible. Otherwise the testing is irrelevant in terms of predicting safety behaviour
and competence when actually out driving, and of course it is the safety and
competence of the train drivers that is the primary concern of any rail network, be that
in the UK or India.

40
8 REFERENCES
Ansiau, D., Wild, P., Niezborala, M., Rouch, I. & Marquie, J. C. 2007, Effects of
working conditions and sleep of the previous day on cognitive performance. Applied
Ergonomics. 39, 99 - 106.
Baltes, P. B. & Baltes, M. M. 1990, Psychological Perspectives on Successful Aging:
The Model of Selective Optimization and Compensation. In P. B. Baltes and M. M.
Baltes (Eds.) Successful Aging: Perspectives from the Behavioural Sciences. (New
York: Cambridge University Press).
Baltes P. B. & Schaie K. W. 1974, The myth of the twilight years. Psychology Today,
31, 720 - 725.
Birnbaum, L. A. & Collins, G. C. 1984, Opportunistic planning and Freudian slips.
Proceedings of the Sixth Cognitive Science Society. (Boulder: CO).
Chapman, P. & Underwood, G. 1998, Visual search of driving situations: danger and
experience. Perception. 27, 951 – 964.
Clarke, D. D., Ward, P., Bartle, C. & Truman, W. 2006, Young driver accidents in the
UK: The influence of age, experience, and time of day. Accident Analysis and
Prevention. 38, 871 – 878.
Colman, A. M. 2003, Oxford Dictionary of Psychology. (Oxford: Oxford University
Press).
Costa, P. T. Jr. & McCrae, R. R. 1980, Still stable after all these years: Personality as
a year to some issues in aging. In P. B Baltes and O. G. Brim (Eds.). Lifespan
Development and Behavior (Vol. 3). (New York: Academic Press). 65 - 102.
Costa, P. T. Jr., Metter, M., & McCrae, R. R. 1994, Personality stability and its
contribution to successful aging. Journal of Geriatric Psychiatry. 27(1), 41 - 59.
Crundall, D., Chapman, P., Phelps, N. & Underwood, G. 2003, Eye movements and
hazard perception in police pursuit and emergency response driving. Journal of
Experimental Psychology. 9 (3), 163 – 174.
Dixon, R. A., Kramer, D. A. & Baltes, P. B. 1985, Intelligence: A life-span
developmental perspective. In B. B. Wolman (Ed.), Handbook of intelligence:
Theories, measurements, and application (New York: Wiley). 301 - 350.

41
Dul, J. & Weerdmeester, B. 2001, Ergonomics For Beginners: A Quick Reference
Guide. 2nd Ed. (London: Taylor & Francis).
Gardner, H. 1985, Frames of Mind: The Theory of Multiple Intelligence. (New York:
Basic Books).
Gillespie, R. 1988, The Hawthorne experiments and the politics of experimentation.
In J. G. Morawski (Ed.), the rise of experimentation in American psychology. (New
Haven: Yale University Press). 114 – 137
Goodwin, C. J. 2005, Research in Psychology: Methods and Design. 4th Ed. (John
Wiley & Sons: USA).
Guildford, J. P. 1967, The Nature of Human Intelligence. (London: McGraw-Hill).
Hall, S. 1999, Hidden Dangers – Railways Safety in the Era of Privatisation.
(London: Ian Allen).
Hamilton, W. I. & Clarke, T. 2005, Driver performance modelling and its practical
application to railway safety. Applied Ergonomics. 36, 661 – 670.
Hayes-Roth, B. & Hayes-Roth, F. 1979, A cognitive model of planning. Cognitive
Science. 3, 275 – 310.
Horn, J. L. 1982, The Theory of Fluid and Crystallized Intelligence in Relation to
Concepts of Cognitive Psychology and Aging in Adulthood. In F. I. M. Craik and S.
Trehub (Eds.). Aging and Cognitive Processes. (New York: Plenum).
Horn, J. L & Donaldson, G. 1977, Faith is not enough: A response to the Baltes-
Schaie claim that intelligence does not wane. American Psychologist. 32, 369 – 373.
Jansson, A., Olsson, E. & Kecklund, L. 2005, Acting or Reacting? A Cognitive Work
Analysis Approach to the Train Driver Task. In J. R. Wilson, B. J. Norris, T. Clarke
and A. Mills (Eds.), Rail Human Factors. (Aldershot: Ashgate).
Jonah, B. A. 1986, Accident risk and risk taking behaviour among younger drivers.
Accident Analysis and Prevention. 16, 255 – 271.
Kahney, H. 2003, Problem solving: Current issues. 2nd Ed. (Maidenhead: Open
University).
Kail, R. V. & Cavanaugh, J. C. 2007, Human Development: A life-span view. (United
Kingdom: Thomson Wadsworth).

42
Kaplan, R. M. & Saccuzzo, D. P. 1997, Psychological Testing: Principles,
Applications, and Issues. 4th Ed. (London: Brookes/Cole Publishing Company).
Lepper, M. R., Ross, L. & Lau, R. R. 1986, Persistence of inaccurate beliefs about the
self: Perseverance effects in the classroom. Journal of Personality and Social
Psychology. 50, 482 – 491.
Luke, T., Brook-Carter, N., Parkes, A. M., Grimes, E. & Mills, A. 2006, An
investigation of train driver visual strategies. Cognitive Technical Work. 8, 15 – 29.
Makishita, H. & Matsunaga, K. 2008, Differences of drivers’ reaction times according
to age and mental workload. Accident Analysis and Prevention. 40, 567 – 575.
McCrae, R. R. & Costa, P.T. Jr. 1990, Personality in Adulthood. (New York:
Guilford).
McGowan, A. & Banbury, S. 2004, Evaluating interruption-based techniques using
embedded measures of driver anticipation. In S. Banbury, S. Tremblay (Eds.), A
Cognitive Approach to Situation Awareness. (Aldershot: Ashgate)
McKenna, F. P. & Crick, J. L. 1991, Hazard perception in drivers: a methodology for
testing and training. Final Report. Transport Research Laboratory. (Crowthorne, UK).
McKenna, F. P. & Crick, J. L. 1994, Developments in hazard perception. Contractor
Report 313. Transport Research Laboratory. (Crowthorne, UK).
McKenna, F. P., Horswill, M. S. & Alexander, J. L. 2006, Does anticipation training
affect drivers’ risk-taking? Journal of Experimental Psychology. 12(1), 1 – 10.
McLeod, R. W., Walker, G. H. & Moray, N. 2005, Analysing and modelling train
driver performance. Applied Ergonomics. 36, 671 – 680.
McNaughton. 2003, Opening address by Chief Engineer of Network Rail to the First
European Conference on Rail Human Factors, York, 13-15 October.
Mills, A. 2005, The RSSB Human Factor Programme. In J. R. Wilson, B. J. Norris, T.
Clarke and A. Mills, (Eds.) Rail Human Factors. (Aldershot: Ashgate).
Mills, K. L., Hall, R. D., McDonald, M. & Rolls, G. W. P. 1998, The effects of hazard
perception training on the development of novice driver skills. Final Report.
http://www.dft.gov.uk. Department for Transport, UK. Accessed 27th June 2008.
[online]

43
OPC. 1993, Train Driver Assessment: The local manager’s resource pack.
Occupational Psychology Centre. (London: British Railway Board).
Orne, M. T. 1962, On the social psychology of the psychology experiment: With
particular reference to demand characteristics and their implications. American
Psychologist. 17, 776 – 783.
Pasquini, A., Rizzo, A. & Save, L. 2004, A methodology for the analysis of SPAD.
Safety Science. 42, 437 – 455.
Patten, C. J. D., Kircher, A., Ostlund, J., Nilsson, L. & Svenson O. 2006, Driver
experience and cognitive workload in different traffic environments. Accident
Analysis and Prevention. 38, 887 -894.
Quimby, A. R. & Watts, G. R. 1981, Human factors and driving performance.
Laboratory Report 1004. Transport and Road Research Laboratory. (Crowthorne,
UK).
Rail Safety and Standards Board. 2008. Press Release.
http://www.rssb.co.uk/press/2008/ASPR_2007_-_1_May_08.pdf. Accessed on 27th
June 2008. [online]
Rantanen, T., Era, P. & Heikkinen, E. 1994, Maximal Isometric Strength and Mobility
among 75-year-old Men and Women. Age and Ageing. 23, 132-137.
Rasmussen, J. 1987. Cognitive control and human error mechanisms. In J.
Rasmussen, K. Duncan K. and J. Leplat (Eds.) New Technology and Human Error.
(Chichester: John Wiley and Sons).
Reid, K. & Dawson, D. 2001, Comparing performance on a simulated 12 hour shift
rotation in young and older subjects. Occupational and Environmental Medicine. 58,
58-62.
Russell, S. & Long, B. 2005, The Role of the Future Train Driver. In J. R. Wilson, B.
J. Norris, T. Clarke, and A. Mills (Eds.) Rail Human Factors. (Ashgate: Aldershot).
Sagberg, F. & Bjornskau, T. 2006, Hazard perception and driving experience among
novice drivers. Accidental Analysis and Prevention. 38, 407 – 414.
Salthouse, T. A. 1985b, A Theory of Cognitive Aging. (Amsterdam: North Holland).
Sattler, J.M. 1982, Age effects on Wechsler Adult Intelligence Scale-Revised Tests.
Journal of Consulting and Clinical Psychology. 50, 785 – 786.

44
Schaie, K. W. 1984, Midlife influences upon intellectual functioning in old age.
International Journal of Behavioural Development. 7, 463 – 478.
Schaie, K. W. 1996, Intellectual Development in Adulthood: The Seattle Longitudinal
Study. (Cambridge: Cambridge University Press).
Seifert, C. M., Patalano, A. L., Hammond, K. J. & Converse, T. M. 1997, Experience
and expertise: the role of Memory in Planning for Opportunities. In P. J. Feltovich, K.
M. Ford and R. R. Hoffman (Eds.) Expertise in Context. (London: The MIT Press).
Sexton, B. 2000, Development of hazard perception testing. In Proceedings of the
Novice Driver Conference (1st – 2nd June), Bristol, http://www.dft.gov.uk. Accessed
22nd May 2008. [online]
Sharma, P. C. & Amitabh. 2004, Safety as Key Business Theme! – Indian Railways
Perspective. In International railway safety conference, Perth.
Spearman, C. 1923, The Nature of ‘Intelligence’ and the Principles of Cognition.
(London: Macmillan).
Twisk, D. A. M. 1996, Young driver accidents in Europe, magnitude and nature. In H.
Simpson (Ed.) New to the Road: Reducing the Risks for Young Motorists. (Los
Angeles: Youth Enhancement Service, UCLA School of Medicine). 27 – 33.
Underwood, G. 2007, Visual attention and the transition from novice to advanced
driver. Ergonomics. 50, 1235 – 1249.
Waller, P. F., Elliott, M. R., Shope, J. T., Raghunathan, T. E. & Little, R. J. A. 2000,
Changes in young adult offence and crash patterns over time. Accident Analysis and
Prevention. 33 (2), 117 – 128.
Wallis, T. S. A. & Horswill, M. S. 2007, Using fuzzy signal detection theory to
determine why experienced and trained drivers respond faster than novices in a hazard
perception test. Accident Analysis and Prevention. 39, 1177 – 1185.
Wilson, J. R. & Norris, B. J. 2005, Rail Human Factors: Past, Present and Future. In J.
R. Wilson, B. J. Norris, T. Clarke and A. Mills, (Eds.), Rail Human Factors.
(Aldershot: Ashgate).
Woods, R. T. 1999, Psychological Problems of Ageing: Assessment, Treatment and
Care. (Chichester: John Wiley & Sons Ltd).

45
Appendix A

Complex reaction time (speed of reaction, agility)

This is the second test undertaken on the test bank. The participant must rest one
finger upon the ‘L’ button (see Figure 11), and watch the screen. On the screen is a
traffic light which turns yellow, green or red. Upon seeing a colour the participant
must take their finger from the ‘L’ button and press the button corresponding to the
colour on the screen (e.g. by pressing the yellow button when a yellow light is
shown). The participant must then return their finger to the ‘L’ button. Reaction time
is recorded for the initial movement to the coloured button from the ‘L’ button, and
from the coloured button back to the ‘L’ button. This test runs for approximately 90
seconds. Reaction time is averaged and then translated to the t-score.

Table 2 displays a list of reaction time raw scores and their corresponding t-scores.
The higher the t-score the better and that means a lower number for reaction time.
There are gaps as this is taken from the actual data used, and with only 499
participants not every time was recorded. From this list however it is possible to see
that if the control group recorded a mean t-score of 46, reaction time was around 925-
915ms.

Figure 11 – Close-up of the control panel

46
Raw Score (ms) T-Score Raw Score (ms) T-Score Raw Score (ms) T-Score
1362-1218 20 995-990 40 798-787 57
1211 21 983-976 41 786-776 58
1199 22 971-961 42 775-764 59
1181 24 960-950 43 763-752 60
1165 25 946-939 44 751-741 61
1149-1147 26 937-926 45 740-730 62
1134-1124 28 925-915 46 727-718 63
1120 29 913-903 47 717-706 64
1108-1107 30 900-892 48 701-698 65
1087-1081 32 889-880 49 604-685 66
1072-1066 33 879-868 50 681-674 67
1065-1055 34 867-857 51 668-660 68
1053-1043 35 856-846 52 658-653 69
1038-1034 36 844-834 53 647-638 70
1030-1023 37 833-822 54 634-628 71
1017-1009 38 821-810 55 623 72
1006-996 39 809-799 56

Table 2 – Reaction time raw scores and corresponding t-scores

47
Appendix B

Form perception (perceptual speed, discrimination of forms)

This is the first test undertaken on the test bank. On the screen flashes five images
simultaneously in a row, one of which is irregular from the others. This could include
a row of four perfectly formed ‘6’s and one that is misshapen. The participant must
then press the corresponding number on the panel (e.g. if the misshapen ‘6’ is the
third in the row, then ‘3’ must be pressed in the panel – see Figure 11). The test lasts
around two minutes. The number of correct responses is translated to the t-score.

Table 3 displays raw form perception scores and t-scores. This is drawn from the data
set so has missing raw scores. However, it is possible to see that if the control group
had a mean of around 55 they were getting 19 answers right on the test.

Form Raw Scores Form Raw Score


T-Score T-Score
(Correct Answers) (Correct Answers)
3 20 15 44
6 20 16 47
7 21 17 50
8 24 18 52
9 27 19 55
10 30 20 58
11 32 21 61
12 35 22 64
13 38 23 67
14 41

Table 3 – Form perception raw scores and corresponding t-scores

48
Appendix C

Group Bourdon (concentration)

This is the third test undertaken on the test bank. A screen full of clusters of dots
comes up and the participant must move from left to right crossing through each
cluster that has a total of four dots. The cursor buttons are used (see Figure 11), and
participants can move left or right at any time, and strike through the four dot clusters
but pressing the up arrow. The strike can be cancelled by pressing the down arrow.
The test lasts for approximately ten minutes. The total number of correct responses is
translated to the t-score.

Table 4 displays raw scores and t-scores. The higher the raw score the better, thus the
t-score is higher. It is possible to see from this table that if the control group had a
mean t-score of around 57, that they gave around 211-216 correct responses.

Group Raw Group Raw Group Raw


(Correct T-Score (Correct T-Score (Correct T-Score
Responses) Responses) Responses)
2 21 129-134 43 217-221 58
21 24 135-140 44 223-227 59
26 25 141-146 45 228-233 60
53 29 147-152 46 234-237 61
57 30 153-158 47 240-243 62
66 32 159-163 48 246-250 63
71-74 33 164-169 49 253-256 64
79-82 34 170-175 50 257-262 65
89-93 36 176-181 51 264-268 66
96-99 37 182-187 52 269-272 67
100-105 38 188-192 53 275-279 68
106-111 39 193-198 54 286 69
112-117 40 199-204 55 288-290 70
118-123 41 205-210 56 308-309 73
124-128 42 211-216 57 331 77

Table 4 – Group Bourdon raw scores and corresponding t-scores

49
Appendix D

Speed perception (perception of moving vehicles, attention)

This is the forth test undertaken on the test bank. On the screen is a train and a tunnel.
To make the train move the ‘H’ button must be pressed once (see Figure 11). The
train then moves at various speeds and enters the tunnel. The ‘L’ button must then be
pressed by the participant at the point they think the train will exit the tunnel. During
practise it is possible to see how close the estimate is, but in the test it is not, and
therefore estimates cannot be adjusted accordingly. The test lasts for around two
minutes. The average error of estimation is translated to the t-score.

Table 5 displays the average error of estimation: as it goes down, the t-score goes up.
From this it is possible to establish that if the control group had a mean t-score of 43,
their average error of estimation would be around 451-436.

Speed Raw Speed Raw Speed Raw


(Avg. Err. T-Score (Avg. Err. T-Score (Avg. Err. T-Score
Estimation) Estimation) Estimation)
8713-851 20 591-577 35 327-312 50
842-830 21 569-562 36 310-294 51
824 22 558-546 37 293-278 52
805-793 23 539-531 38 275-258 53
786-775 24 522-507 39 257-242 54
770-757 25 498-491 40 240-223 55
750-737 26 485-474 41 222-206 56
727-719 27 469-454 42 204-188 57
718-705 28 451-436 43 186-171 58
696-685 29 434-419 44 169-152 59
673-667 30 414-401 45 151-136 60
661-654 31 395-382 46 133-117 61
646-636 32 380-366 47 115-100 62
626-614 33 361-350 48 96-83 63
610-597 34 346-329 49 79-68 64

Table 5 – Speed perception raw scores and corresponding t-scores

50
Appendix E

Visual differentiation (sustained attention)

This is the last test undertaken on the test bank. On the screen a number of coloured
dots appear in various places and of various combinations. Then another dot appears
and disappears in various places and combinations. Another dot appears and
disappears. Then a third dot appears and disappears. The participant must press the
‘H’ button when, at any point during the dots appearing, the colours red, green and
yellow appear in sequence (see Figure 11). Furthermore, the participant must press the
‘L’ button when five dots of the same colour (any colour) appear on the screen at any
time. After the three dots have appeared on the screen, a new screen appears with
different dots and the process happens again. The test last for approximately 30
minutes. The number of correct detections (of sequences and groups of five) is
translated to the t-score.

Tables 6 and 7 display the translation between the raw scores and the t-scores. The
first table is for the correct detection of sequences of dots. The second table is for
correct detection of five dots of the same colour. T-scores are taken from the two tests
and averaged for an overall visual differentiation test score.

Visual Raw Visual Raw Visual Raw


T-Score T-Score T-Score
1 1 1
0 20 13 42 25-26 52
1-2 33 14 43 27 53
3 34 15 44 28 54
4 35 16-17 45 29 55
5 36 18 46 30-31 56
6-7 37 19 47 32 57
8 38 20 48 33 58
9 39 21-22 49 34 59
10 40 23 50 35-36 60
11-12 41 24 51 37 61
38 62

Table 6 – Visual differentiation sequence raw scores and corresponding t-scores


51
Visual Raw 2 T-Score Visual Raw 2 T-Score Visual Raw 3 T-Score
0 20 7 42 14 53
1 32 8 43 15 55
2 33 9 45 16 57
3 35 10 47 17 58
4 37 11 48 18 60
5 38 12 50 19 62
6 40 13 52

Table 7 – Visual differentiation field raw scores and corresponding t-scores

52
Appendix F

Sample of the Study

Group  Age  Experimental %  Controls  Controls  Controls 


Range  Group  Requested  Acceptable  Acquired 

Group 1  0‐40  30  7  4  4  5 

Group 2  41‐45  70  16  9  8  7 

Group 3  46‐50  53  12  7  6  8 

Group 4  51‐55  177  39  23  20  17 

Group 5  56‐60  118  26  16  14  14 

Total    448    59  52  51 

53
Appendix G

Psycho-Technical Directorate

Note No.PT/AP/4.1 Dated: 11.07.2008

Sub: Development of norms on Psychological Test Battery.


---

Psycho-Technical Directorate is responsible for screening of High Speed


Train Drivers on Indian Railways. In this connection we have created a data bank of
Locomotive Pilots from different age groups. In order to evaluate the performance of
Locomotive Pilots on psychological test batter we need a reference group drawn from
non-driving population representing age groups in the sample of drivers in the
normative group. In this connection we are inviting volunteers from different
directorate of RDSO in the indentified age-groups to take psychological test on
Computerized Aptitude Test for Drivers (CADAT). The tests are expected to be of 90
minutes duration.

2. You are requested to kindly permit the staff (who is a volunteer) to leave
office for about 2 hrs. to undertake the test in Psycho-Technical Laboratory, Annexe-
II. The age group in which the volunteers are required along with approximate
number of subjects is indicated in the annexure.

(R.K. Lal)
ED/Traffic(P)
Distribution

Sr. ED/Civil Sr. ED/Signal Sr. ED /Elec. Loco

Sr. ED/Finance ED/TI ED/Traffic

ED/Carriage ED/Wagon ED/QA-Mech

ED /Testing ED /ED ED /B&S

ED /Track ED /QA/Civil ED /TM

ED /GE ED/PS&EMU ED/Research

ED/MP ED/M&C ED/Admn.

ED/Stores CPO ED/VDG

Copy to:
1. Secy. to DG for kind information of DG.
2. PPS to ADG for kind information of ADG.

54
Appendix H Reference Number:

Psychometric Testing

We are currently engaged in a project investigating the effects of age and experience
on psychometric performance, and would appreciate your participation in this study.
Please complete the following details before taking the psychometric tests. Thank
you.

Age:
:

Date of Birth:……………………………………………………..
:
Current Designation:…………………………………………..
:
Start date of current post:…………………………………..
?

Please confirm you have NO experience of train driving by signing your name below:

…………………………………………

Date of testing:……………………… Tested by:……………………………

55

You might also like