Train punctuality: the passenger perspectived3cez36w5wymxj.cloudfront.net/migrated/Train... ·...
Transcript of Train punctuality: the passenger perspectived3cez36w5wymxj.cloudfront.net/migrated/Train... ·...
Train punctuality: the passenger perspective
November 2015
1
ContentsForeword 1
Key findings 2
Transport Focus conclusions and recommendations3
2
3
6
8
Train punctuality: the passenger perspective1
Office of Rail and Road conclusions 4
Punctuality performance measurement 115
Background, objectives and methodology 12Passengers’ top of mind views on train performance 17Reaction to PPM and CaSL performance measures 22Speed, punctuality, frequency trade-offs 35Performance measurement trade-off exercises 41Summary 49
Punctuality and passenger satisfaction 536
2
Foreword
As part of this push to give more weight in the process to passengers’ perspectives, Transport Focus and
the Office of Rail and Road (ORR) have co-funded twostudies to understand more about passengers’ views on train punctuality. The first is new qualitative researchinto views about train performance – punctuality andcancellations – and how it should be measured. Thesecond is a quantitative analysis of the relationshipbetween a passenger’s actual punctuality and theirsatisfaction with the journey in question.
The research involved 10 focus groups acrossEngland, Scotland and Wales in March 2015 (takingplace in Cardiff, Glasgow, London and Manchester). Thepassengers taking part reflected a cross-section of age,gender and socio-demographics, as well as differentsegments of the rail market. Meanwhile the quantitativeanalysis looked at the relationship between actualpunctuality experienced and how satisfied 10,849 AbellioGreater Anglia passengers were when they took part in
the National Rail Passenger Survey (NRPS) between2012 and 2014. It is clear from both studies that greaterfocus on punctuality will be needed to deliver whatpassengers expect in the five years from 2019 (railwayControl Period 6, or CP6). The key findings of the twostudies are set out below, followed by the conclusionsand recommendations that Transport Focus and ORReach draw from them.
We hope that this report will help support the process of determining the Control Period 6 passengerperformance targets and ensuring that they focus onpassengers’ needs.
Anthony Smith Richard PriceChief Executive Chief ExecutiveTransport Focus Office of Rail and Road
The process to determine what Britain’s railways should deliver in the five yearsfrom 2019 is already underway. It is vital that passengers’ views are at the heartof the process – and with passengers paying over 60 per cent of the cost of therailway through fares, their views must count in decisions on how rail serviceimprovements and investments are prioritised.
Train punctuality: the passenger perspective1
3
Key findings from the passenger researchThe full research report by Illuminas is included in thisdocument as section 5. The key findings are:
• Passengers in the focus groups saw a punctual service as thekey success criterion for their journey. They saw performancemeasurement as an important part of ensuring that railcompanies focus sufficiently on punctuality. Punctuality is avital prerequisite for building trust between passengers and a train company.
• In general, passengers want a tougher Public PerformanceMeasure (PPM) target, but feel that the Cancellations andSignificant Lateness (CaSL) target is more reasonable. Oncemade aware of how PPM is calculated, passengers oftenregarded a very high target as not unreasonable – given thefive minutes ‘grace’ (10 minutes for long-distance trains) for a train to be considered ‘on time’ by the industry. Thepassengers also suspected that the current target level of 92.5 per cent is regarded as a ceiling rather than a minimum.Some argued that an industry focused on its customers should always aim to deliver 100 per cent, even if it is never achieved.
• Some passengers suggested that a ‘significantly late’ trainshould be 10 minutes for short-distance and 20 minutes forlong-distance, which is a tighter definition than the 30 minutescurrently used by the industry. For some, combining
cancellations and significant lateness was confusing.Passengers in the Glasgow groups could not understand why there is no CaSL target in Scotland.
• Passengers in the focus groups defined ‘on time’ as beingwithin one minute of scheduled time, although some wouldaccept leeway of one or two minutes. The passengers believedpunctuality should be measured at all stations, not just where a train terminates. There was low awareness of the currentperformance measures and lack of trust in how the rail industrymeasures punctuality. Once aware of how PPM is calculated,the passengers felt it offered the rail industry too muchflexibility in meeting targets they already regarded as too low; it also confirmed their suspicions about statistics generally, in that a train that is late can count as ‘on time’.
• The research explored if it would be better to measure theproportion of passengers who arrive on time, rather than the proportion of trains that are on time. Some commuters in the focus groups found this attractive, recognising that there would be greater incentive for train companies to runtheir trains punctually. On balance, however, passengers in the focus groups felt that all trains should count equallytowards punctuality targets. There was a strong view thatpassengers travelling on a lightly-loaded train should not be treated as less important when it comes to focus onpunctuality.
Train punctuality: the passenger perspective2
4
Trade-offsVarious ‘trade-offs’, described below, were explored in theresearch. Significantly, passengers in the focus groups felt that if the railway is properly-planned, properly-maintained and properly-run it should not be necessary to make trade-offs between thequality of different aspects of their journey. Many passengers felt that not striving to deliver on all fronts is an example of thecustomers interest being insufficiently to the fore of rail industrythinking. Three key points emerged:• Punctuality versus speed and frequency. Overall, the
preference expressed was for trains to be more punctualrather than faster or more frequent. There were exceptions:on very long-distance trips journey time reduction isimportant, as is frequency to some leisure passengers.
• Punctuality versus fewer cancellations and days ofwidespread disruption. Commuters regarded punctuality as the priority because it affects them twice a day every day, whereas leisure and business passengers felt that ashort delay has less impact on them than a cancellation.
• Punctuality versus running an extra train. Most sawpunctuality as more important than seeking to run an extratrain. This was particularly true of commuters, unless theyexperienced crowding that prevented them from getting on the train at all.
Other findings• Many passengers taking part in the research felt it is not
acceptable for train companies to add extra time into thetimetable on approach to the destination station. Those whothought managing expectations in this way was acceptableoften observed that it would be more honest if the extra time was distributed throughout the journey.
• On very high frequency routes, for example with 10 trains per hour or more, the idea of measuring the interval betweentrains (rather than measuring each train at its destination)found favour with most passengers. However some still feltthat trains should be measured against the timetable even in these circumstances.
• Passengers in the research favoured the current approach by some train companies of adjusting the timetable in theautumn leaf-fall season to deliver a predictable, reliableservice, even if journey times are slightly extended.
• There was strong opposition to the current practice that a train cancelled before 22:00 the day before does not count as a cancellation in official statistics.
Passengers were presented with what some regarded as a trickydilemma. Namely, whether – on a day of widespread disruption – the rail industry should seek to reinstate the normal timetableas quickly as possible, even if there is short-term disadvantageto some passengers as a result. The sort of things that might be done include: trains missing out stations so the next journeyruns on time; cancelling a train so the next one runs on time; and terminating trains before the normal destination so most ofthe next trip runs on time. While the effect of the focus groupenvironment may have influenced the responses given in front of peers, many participants felt it was fairer to ‘spread the pain’than cause disadvantage to some for the ‘greater good’.However, passengers in London felt that short-term disruption to restore the timetable was reasonable because alternatives are more likely to exist for those affected.
Train punctuality: the passenger perspective2
5
The executive summary of the report by consultantsGHD is included as section 6.
This analysis updated work Transport Focus published in 20101
examining the relationship between actual train performance andpassenger satisfaction. The 2015 work looked at the actualpunctuality experienced by 10,849 Abellio Greater Anglia passengerswho took part in the National Rail Passenger Survey (NRPS)between 2012 and 2014. We are grateful to Abellio Greater Angliafor their co-operation in providing detailed train performanceinformation for the three years in question.
The key findings are:• 82 per cent of passengers were satisfied with punctuality
when their train arrived early or within 59 seconds of the timegiven in the public timetable. This varies by journey purpose:75 per cent of commuters were satisfied with punctualitywhen on time or early, while 94 per cent of leisurepassengers and 91 per cent of business passengers weresatisfied with punctuality when on time. A possible reason forthe variation is that commuters’ assessment of punctuality‘today’ is affected by past experience – and, put simply,commuters have more ‘past experience’ than leisure andbusiness passengers. The effect of cancellations may alsohave a marginal effect. For example where a passenger’sintended train is cancelled and they catch the next one – thepassenger is late but the train they travelled on is on time.
• For every minute of lateness (that is, after scheduled arrivaltime), passenger satisfaction with punctuality declined by threepercentage points. This also varies by journey purpose: commutersatisfaction declined by five percentage points per minute oflateness, while among leisure passengers the decline was onepercentage point per minute. Therefore a significant degree of passenger satisfaction is ‘lost’ when trains are officially ‘ontime’ according to the industry measure of PPM, but late inpassengers’ eyes. For example, passengers travelling on a train
that arrives four minutes 59 seconds late – ‘on time’ accordingto PPM – will record satisfaction with punctuality over 12percentage points lower than had that train been truly on time.
• For every minute of lateness, overall satisfaction declines by oneand a half percentage points. Among commuters the decline in overall satisfaction is steeper at three percentage points perminute of lateness. Therefore overall satisfaction on a train thatis four minutes 59 seconds late – that is, on time according to PPM – will have fallen by over six percentage points.
• On trains leaving London (where the majority of passengersare likely to get off before the train’s terminating station), 56 per cent of passengers arrived at their station on time or early. This compares with 69 per cent of trains reachingtheir destination on time.
Key findings from analysis of the relationship between punctualityand passenger satisfaction
1What passengers want – Towards a ‘right time’ East Anglian railwayhttp://www.transportfocus.org.uk/research/publications/what-passengers-want-towards-a-right-time-east-anglian-railway
6
Train punctuality: the passenger perspective3
2Presenting ‘right-time’ performance information to rail passengers, May 2013http://www.transportfocus.org.uk/research/publications/presenting-righttime-performance-information-to-rail-passengers
Transport Focusconclusions and recommendationsThe overall conclusion Transport Focus draws from this researchand analysis is that a key strategic focus for Control Period 6must be on delivering a railway that is markedly more punctualthan at present. That is, markedly more punctual in the waypassengers understand punctuality – on time to within a minuteof the timetable (‘right time’, as the rail industry calls it). Unlessan incident is specifically disrupting the service, passengersshould arrive ‘right time’ significantly more often than nowwherever they get off. Trains which never run on time, or whichare late more often than not, must become a thing of the past.And we note that it can be done: Arriva Trains Wales, ChilternRailways and c2c regularly achieve a ‘right time’ arrival score of over 80 per cent.
The research shows that passengers want a robust,dependable timetable so they arrive on time at their destination.It appears that passengers – particularly commuters – are moreinterested in arriving punctually than in how long the journeytakes or in the industry striving to run extra trains. There areexceptions: there is appetite for reduced journey time on verylong-distance trips; for having more trains on routes with lowfrequencies; and for extra trains among commuters who strugglejust to get on the train – never mind getting a seat. It should bekept in mind that this is qualitative research, but it suggests thatpassengers do not want a ‘fudge’ where the timetable just aboutworks on paper, but rarely if ever in practice. This should beexplored further in quantitative research.
It is clear that punctuality measured by the PublicPerformance Measure (PPM) does not accurately reflectpassengers’ experience, mainly because of the five or 10 minute leeway and because trains are measured only at theirdestination. PPM also undermines trust: counting trains as ontime when they are late plays badly to passengers, many ofwhom are already suspicious of statistics from their traincompany. Transport Focus therefore advocates a move awayfrom PPM as the single regulatory measure of train punctuality.Continuing for another five years with a regime where late isconsidered ‘OK’ is simply not tenable. But Transport Focusrecognises that moving instead to a single pass/fail measure of‘arrival within 59 seconds’ could have unintended consequences.Might the answer be targets to encourage movement of thewhole ‘arrivals curve’ (example below) to the left on the graph,incentivising better ‘right time’ performance, but also givingincentives to achieve a three minute delay rather than a fourminute delay, or a nine minute delay rather than an 11 minutedelay and so on?
The industry needs to determine how punctuality should bemeasured at intermediate stations and then reflected in theregulatory targets. A new metric could try to combine punctualityat destination with that at some or all intermediate stations. Itwould be more difficult to explain, but it might focus attention onwhat really matters to passengers. If accompanied by completetransparency about performance, passengers would still be ableto see how their trains performed, but the metric would bedifficult to explain to passengers and staff – potentially limiting its effectiveness as a tool to drive improvement.
Example of an ‘arrivals curve’
Pro
po
rtio
n o
f tr
ain
s
Lateness in minutes
30%
25%
20%
15%
10%
5%
0%-5 -3 -1 1 3 5 7 9 11 13 15 17 19 21 23 25 27
7
Control Period 6 should seek to deliver better punctuality for commuters – but not at the expense of giving train companiesless incentive to run off-peak trains on time. The fact that poorpeak punctuality can currently be off-set by good off-peakperformance should be addressed, but not by creating theopposite problem. Would setting a target for peak trains, as wellas for the timetable as a whole, be a way to increase focus ontrains carrying a large numbers of commuters, without riskingperverse incentives elsewhere?
A single pass/fail measure does not paint a full picture. A trainwith a 'right time' record of zero might arrive within two minuteson every occasion - but equally it might always be nine minuteslate. It should therefore be made easy for a passenger to see agraduated picture of how often their train is ‘right time’, arriveswithin two minutes, three minutes, four minutes and so on. Thiswould give a richer, more rounded view of train punctuality to both passengers and commentators.
The definition of ‘significant lateness’ within the Cancellationsand Significant Lateness (CaSL) measure should be reviewed for Control Period 6. To passengers, particularly those makingshort journeys, a train becomes significantly late after less than30 minutes delay. Separately, the target for CaSL should bereduced further for long-distance operators. Passengers makinglong-distance leisure and business trips said cancellations causethem more trouble than late running, yet the target for VirginTrains (that is, Intercity West Coast) and Virgin Trains East Coast is more forgiving than for other operators. A CaSL target should also beconsidered for Scotland.
The industry shouldconsider how underlyingpunctuality can be measuredand better understood – thatis, on days when there havebeen no specific incidents.Perhaps this could beachieved by tracking ‘righttime’ performance on the ninebest days in every 10 (with theleast good removed)? Orperhaps by tracking ‘right time’punctuality on days when thereare very few cancellations orsignificantly late trains (as a proxy for the absence of disruption)?
Finally, in looking forward to 2019 the industry shouldconsider if the level of punctuality required by previous High LevelOutput Specifications has been insufficiently ambitious to driveradical thinking and innovation. If the governments and regulatorwere to seek substantially higher performance in Control Period 6and beyond, could it be the catalyst for fundamental change thatwould ultimately lead to exceptionally high levels of punctualityand highly efficient use of track, station and fleet capacity?
Summary• When it comes to performance, a main objective for 2019-
2024 should be to markedly improve true ‘on time’ punctualityacross the rail network, including at intermediate stations. That means robust timetables that are neither overly heroic nor excessively padded. It means accurate sectional runningtimes and accurate station dwell times.
• Transport Focus advocates a move away from PPM as thesole measure of success. It is not trusted by passengers anddoes not drive sufficient focus on punctuality as passengersdefine it.
• Whatever replaces PPM must strongly incentivise ‘right time’punctuality, but seek to avoid unintended consequences ofhaving a different, more demanding, sole measure of success.The industry should investigate if targets could be developedthat would incentivise better performance generally, including‘right time’. How can the industry be incentivised in ControlPeriod 6 to move the whole ‘arrivals curve’ to the left on thegraph?
• Whatever the chosen metric, there should be a separatepunctuality target for peak services to incentivise the industryto get commuters to and from work on time. Givenpassengers’ concerns about reducing the incentive to run off-peak trains on time, this is preferable to weighting trains by volume of passengers within punctuality statistics.
• To give a more meaningful picture of punctuality, the industryshould make it easy for passengers to see a graduated pictureof how often their train (or group of trains) is ‘right time’, within two minutes, three minutes, four minutes and so on.
• The industry must become properly transparent and granularwith all performance information – and through the ‘front door’via train company websites and apps. If a passenger uses onlythe ‘07:29’ and the ‘18:17’, data for those trains alone shouldbe readily available to them. The industry did not respondeffectively to recommendations following research in 20132
(funded jointly by Transport Focus, ORR and National RailEnquiries) and it should now do so.
• There should be a review of the Cancellations and SignificantLateness (CaSL) measure and targets with a view to: reducingthe definition of ‘significant lateness’, particularly for short-distance journeys; further tightening the target forlong-distance train companies; and Transport Scotland shouldconsider specifying a CaSL measure in Control Period 6.
• The industry should tackle causes of distrust in statistics by:eliminating differences between the public timetable and the working timetable; reforming the system wherebycancellations made before 22:00 the day before do not countas cancellations in statistics; and being transparent about the number of days on which a formal amended timetable (one involving fewer trains running than normal) operated and about the number of trains involved.
As the regulator for rail in Great Britain, we want to ensure thatpassengers are at the heart of Britain’s railways and we are keen to see improvements in the levels of punctuality and reliability forpassengers achieved by the industry to the greatest extent possiblewith the funding available. These studies are an important startingpoint as part of our preparations for Control Period 6 (CP6).
The qualitative study reflects the views of 80 passengers using therail network. Their views provide us and the industry with a valuableinsight into the priorities of different passenger groups, their level ofawareness of performance targets and how they understand andinterpret the existing performance measures. The quantitative analysisis useful in helping the industry better understand the real relationshipbetween actual train performance and passenger satisfaction, althoughit must be noted that the results for Abellio Greater Anglia passengersmay not be reflective of all train passengers, especially those usingdifferent types of services, e.g. long distance.
ORR is working to get the best deal for passenger and taxpayers.ORR’s role is to create incentives and set targets for passenger
punctuality and reliability that reflect the interests of both governmentand the passenger (and recognising also the role of freight operators).We must develop a framework that genuinely encourages NetworkRail and the industry to deliver improvements in punctuality, and which also aligns the objectives of train operators and Network Rail.
We are pleased that the industry National Task Force (NTF) isleading a review of the existing measures and generating alternativeoptions for CP6 and that it has found these studies useful in thatprocess. We would like the industry to deliver an outcome thatprovides the industry with measures which drive the right behaviours,which government would feel comfortable using for setting its highlevel targets for the industry and for its train operator franchises, and which we could then also use to set the regulatory targets forholding Network Rail to account.
We particularly welcome that the industry has taken on board the need to provide passengers with more accessible and transparentinformation about punctuality. The National Task Force has endorseda “My Journey” work stream to make additional information available
Office of Rail and Roadconclusions
8
Train punctuality: the passenger perspective4
to passengers about the reliability of their individual journeys. We willmonitor progress with this work.
Before introduction of any new industry punctuality and reliabilitymeasure, there are a range of options and features now underdiscussion, in part informed by these studies and also from industrydiscussion and debate. This includes measurement to ‘right time’(arrival within one minute of scheduled time), measurement at everystation, use of banding, and some sort of passenger weighting. We consider that it is likely that we will need to achieve a balance of these desirable features across a small number of measures,rather than seeking to achieve one single measure which reflects all the features identified.
The assessment by the industry, by government, and by us of anypotential new industry measures must be a carefully considered oneand so we do not propose to make recommendations as to particularmeasures at this stage. We particularly want to understand how anynew measure would translate into management behaviour in the dayto day operation of the railway, and also in how the railway is planned
and timetabled, so that any unwanted outcomes are avoided. Forexample a ‘right time’ measure has the potential to result in extendedjourney times as a result of longer station dwell times to enablegreater volumes and different types of passengers to get on and offthe train. We must take into account the impact of any changes topassenger punctuality measures on the freight market (and vice versa).
There is merit in continuing to measure PPM for comparison andcontinuity purposes, as this is a measure that has been in place for a number of years. New measures, which may help the industryaddress the issues raised by these studies, could require improvementsin technology (or the availability or distribution of technology on thenetwork). We believe we must not be limited by the currenttechnology but set the industry on a path that enables the importantnew technology on the horizon, such as GPS train fitment and smartticketing, to be utilized to further improve measures over time.
We are acutely aware that there is a very difficult balance to strike between a comprehensive measure that captures all elementsof punctuality and one that is transparent and easy for users tounderstand.
The NTF is leading an industry trial of a selection of newmeasures over the coming months, and we are eager to see theoutcome of this trial. We consider that there needs to be flexibility to develop a strong set of measures that would meet the criteria set out above. We do not consider it necessary to wait until the nextControl Period to implement a new measure or set of measures.
In terms of next steps, we will continue to engage with the NTF workstream, while developing our views on what the regulatedoutputs will be. This will inform our initial Periodic Review 2018(PR18) consultation early in 2016, and wider consultation which we intend to carry out in relation to outputs around that time.
We look forward to continuing to work with the industry todevelop proposals in this area.
9
10
11
Pre
pare
d in
com
plia
nce
with
the
Inte
rnat
iona
l qua
lity
stan
dard
cov
erin
g m
arke
t res
earc
h, IS
O 2
0252
(201
2), T
he M
RS
Cod
e of
Con
duct
, and
the
Dat
a P
rote
ctio
n A
ct 1
998
by Il
lum
inas
, 183
-203
Eve
rsho
lt S
treet
, Lon
don
NW
1 1B
U, U
K
T +4
4 (0
)20
7909
092
9 F
+44
(0)2
0 79
09 0
921
E in
fo@
illum
inas
-gl
obal
.com
w
ww
.illu
min
as-g
loba
l.com
PREP
AR
ED B
Y IL
LUM
INA
S a
glob
al te
am b
ased
in L
ondo
n, N
ew Y
ork
and
Aus
tin
Punc
tual
ity p
erfo
rman
ce m
easu
rem
ent
Res
earc
h D
ebrie
f
Prep
ared
for:
Tra
nspo
rt F
ocus
D
ate:
8th
Apr
il 20
15
12
Bac
kgro
und,
obj
ectiv
es a
nd
met
hodo
logy
13
Bac
kgro
und
Trai
n pe
rform
ance
in B
ritai
n is
sub
ject
to a
num
ber o
f qua
lity
serv
ice
targ
ets
cove
ring
a ra
nge
of a
spec
ts o
f the
ser
vice
whi
ch th
e ra
ilway
s pr
ovid
e
Thes
e ta
rget
s ar
e se
t by
a pr
oces
s in
volv
ing
the
Offi
ce o
f Rai
l and
Roa
d (O
RR
), th
e ra
il in
dust
ry it
self
and
the
Wes
tmin
ster
and
Edi
nbur
gh
gove
rnm
ents
. The
se b
odie
s ar
e ab
out t
o be
gin
a re
view
pro
cess
that
will
le
ad to
the
prod
uctio
n of
a s
et o
f Hig
h Le
vel O
utpu
t Spe
cific
atio
ns th
at w
ill
incl
ude
targ
ets
for t
rain
per
form
ance
from
201
9 to
202
4
To h
elp
info
rm th
e de
velo
pmen
t of t
hese
targ
ets,
Tra
nspo
rt Fo
cus
and
the
OR
R h
ave
com
mis
sion
ed re
sear
ch to
ens
ure
that
pas
seng
ers’
prio
ritie
s fo
r tra
in p
erfo
rman
ce a
re c
lear
ly u
nder
stoo
d an
d in
form
the
abov
e pr
oces
s
14
The
obje
ctiv
es o
f the
rese
arch
are
to e
stab
lish
wha
t pas
seng
ers
know
abo
ut ra
il pe
rform
ance
mea
sure
men
t an
d to
iden
tify
thei
r asp
iratio
ns fo
r how
(and
wha
t ele
men
ts o
f) pe
rform
ance
sho
uld
be m
easu
red
in th
e fu
ture
Key
obj
ectiv
es fo
r the
rese
arch
are
to u
nder
stan
d:
How
pas
seng
ers
judg
e tra
in c
ompa
nies
’ ‘pe
rform
ance
’ P
asse
nger
s’ e
xper
ienc
e of
dis
rupt
ion
incl
udin
g de
lays
, can
cella
tions
, div
ersi
ons,
sho
rt-fo
rmat
ions
, etc
H
ow th
ese
expe
rienc
es tr
ansl
ate
into
pas
seng
ers’
ass
essm
ent o
f the
railw
ay’s
‘per
form
ance
’ P
asse
nger
s’ k
now
ledg
e an
d un
ders
tand
ing
of h
ow p
erfo
rman
ce is
cur
rent
ly m
easu
red
offic
ially
P
asse
nger
s’ u
nder
stan
ding
of a
nd re
actio
ns to
the
Pub
lic P
erfo
rman
ce M
easu
re a
nd th
e ge
nera
l con
cept
of
‘lat
enes
s’
Pas
seng
ers’
und
erst
andi
ng o
f and
reac
tions
to th
e C
ance
llatio
ns a
nd S
igni
fican
t Lat
enes
s M
easu
re
The
exte
nt to
whi
ch p
asse
nger
s un
ders
tand
and
are
pre
pare
d to
mak
e al
low
ance
s fo
r the
inte
r-re
latio
nshi
p of
ser
vice
freq
uenc
y, jo
urne
y tim
e (s
peed
), ca
paci
ty (a
vaila
bilit
y of
sea
ts) a
nd p
erfo
rman
ce
Inte
rest
am
ongs
t pas
seng
ers
in h
avin
g of
ficia
l per
form
ance
mea
sure
s pu
blis
hed
and
avai
labl
e fo
r pub
lic
scru
tiny
Diff
eren
ces
in u
nder
stan
ding
and
exp
ecta
tions
by
pass
enge
r typ
e, jo
urne
y pu
rpos
e, jo
urne
y di
stan
ce,
serv
ice
frequ
ency
, etc
Th
e im
pact
of p
oor p
erfo
rman
ce o
n pa
ssen
gers
’ tru
st in
the
railw
ays
and/
or s
peci
fic T
rain
Ope
ratin
g C
ompa
nies
(TO
C)
Wha
t crit
eria
sho
uld
form
the
basi
s of
any
qua
ntita
tive
asse
ssm
ent o
f pas
seng
er p
erce
ptio
ns o
f TO
C
perfo
rman
ce.
Obj
ectiv
es
15
Met
hodo
logy
Ten
qual
itativ
e fo
cus
grou
ps w
ere
cond
ucte
d ac
ross
four
loca
tions
: Lo
ndon
D
erby
G
lasg
ow
Car
diff
Four
focu
s gr
oups
wer
e co
nduc
ted
in L
ondo
n an
d tw
o in
eac
h of
the
othe
r loc
atio
ns:
Two
of th
e gr
oups
in L
ondo
n sa
mpl
ed A
belli
o G
reat
er A
nglia
pas
seng
ers
only.
Thi
s w
as s
o th
at th
e fin
ding
s of
thes
e gr
oups
can
be
cons
ider
ed a
long
side
a s
epar
ate
stud
y th
at T
rans
port
Focu
s an
d O
RR
hav
e co
mm
issi
oned
look
ing
at th
e im
pact
of a
ctua
l del
ay e
xper
ienc
ed b
y pa
ssen
gers
on
how
th
ey a
nsw
er p
artic
ular
Nat
iona
l Rai
l Pas
seng
er S
urve
y qu
estio
ns. F
or th
at e
xerc
ise
Abe
llio
Gre
ater
A
nglia
has
bee
n us
ed a
s a
case
stu
dy.
Eac
h fo
cus
grou
p co
ntai
ned
eigh
t res
pond
ents
who
use
d th
e tra
in fo
r eith
er c
omm
utin
g or
bus
ines
s/le
isur
e pu
rpos
es
Res
pond
ents
wer
e al
so c
ateg
oris
ed b
y ag
e an
d so
cio-
dem
ogra
phic
s.
16
Gro
up d
iscu
ssio
n pr
ogra
mm
e
The
focu
s gr
oup
disc
ussi
on p
rogr
amm
e is
det
aile
d be
low
:
LOCA
TIO
N
DEM
OG
RAPH
ICS
& T
RAIN
USA
GE
TRA
IN U
SAG
E
Lond
on
Youn
ger,
C1C
2D
Com
mut
er
Old
er,
ABC1
Bu
sine
ss/L
eisu
re
Lond
on (
Abel
lio G
reat
er A
nglia
pass
enge
rs)
Youn
ger,
ABC
1 Bu
sine
ss/L
eisu
re
Old
er,
C1C2
D
Com
mut
er
Man
ches
ter
Youn
ger,
C1C
2D
Busi
ness
/Lei
sure
Old
er,
ABC1
Co
mm
uter
Card
iff
Youn
ger,
ABC
1 Co
mm
uter
Old
er,
C1C2
D
Busi
ness
/Lei
sure
Gla
sgow
Yo
unge
r, C
1C2D
Co
mm
uter
Old
er,
ABC1
Bu
sine
ss/L
eisu
re
Focu
s gr
oups
took
pla
ce w
eeks
com
men
cing
9 a
nd 1
6 M
arch
201
5.
Methodology
Ten qualitative focus groups were conducted across four locations: London Derby Glasgow Cardiff
Four focus groups were conducted in London and two in each of the other locations: Two of the groups in London sampled Abellio Greater Anglia passengers only. This was so that the findings of these groups can be considered alongside a separate study that Transport Focus and ORR have commissioned looking at the impact of actual delay experienced by passengers on how they answer particular National Rail Passenger Survey questions. For that exercise Abellio Greater Anglia has been used as a case study.
Each focus group contained eight respondents who used the train for either commuting or business/leisure purposes
Respondents were also categorised by age and socio-demographics.
6
17
Pas
seng
ers’
top
of m
ind
view
s on
trai
n pe
rform
ance
18
Wha
t mak
es fo
r a “s
ucce
ssfu
l” jo
urne
y?
Rel
iabi
lity
is p
asse
nger
s’ m
inim
um re
quire
men
t fro
m th
eir t
rain
ser
vice
P
asse
nger
s de
fine
relia
bilit
y pr
imar
ily in
term
s of
pun
ctua
lity
and
seco
ndar
ily in
term
s of
not
bei
ng
canc
elle
d or
dis
rupt
ed (r
efle
ctin
g th
e fre
quen
cy w
ith w
hich
thes
e is
sues
are
typi
cally
exp
erie
nced
)
Pun
ctua
l ser
vice
Sta
tion
envi
ronm
ent
Trai
n en
viro
nmen
t
Get
ting
a se
at
Oth
er th
an a
relia
ble
serv
ice,
a ra
nge
of
addi
tiona
l fac
tors
info
rm p
asse
nger
s’
asse
ssm
ent o
f wha
t mak
es fo
r a s
ucce
ssfu
l jo
urne
y Th
ese
fact
ors,
suc
h as
get
ting
a se
at a
s w
ell a
s a
clea
n an
d sa
fe e
nviro
nmen
t on
the
train
and
at
the
stat
ion,
can
enh
ance
the
exte
nt to
whi
ch th
e jo
urne
y is
rate
d po
sitiv
ely
But
thes
e fa
ctor
s al
one
(i.e.
in th
e ab
senc
e of
a
punc
tual
ser
vice
) will
not
resu
lt in
a s
ucce
ssfu
l jo
urne
y
The
impo
rtanc
e of
thes
e fa
ctor
s of
cou
rse
varie
s by
jour
ney
type
Fo
r exa
mpl
e on
a lo
nger
dis
tanc
e bu
sine
ss/le
isur
e jo
urne
y, g
ettin
g a
seat
is a
lmos
t exp
ecte
d an
d th
e la
ck o
f it c
an s
ever
ely
influ
ence
sat
isfa
ctio
n H
owev
er, o
n a
shor
ter c
omm
uter
trip
in p
eak
hour
s pa
ssen
gers
do
not n
eces
saril
y ex
pect
to g
et a
se
at, s
o ge
tting
one
can
cre
ate
cons
ider
able
del
ight
.
Hyg
iene
fa
ctor
s
Qua
lity
driv
ers
Sta
ff
19
Ove
rall,
pas
seng
ers’
sat
isfa
ctio
n w
ith th
e le
vel o
f del
ay, d
isru
ptio
n an
d ca
ncel
latio
ns is
not
esp
ecia
lly
poor
H
owev
er, e
xper
ienc
es d
o va
ry b
y pa
ssen
ger t
ype
CO
MM
UTE
R
Del
ays
tend
to b
e m
inor
but
freq
uent
Fo
r exa
mpl
e, 2
-3 m
inut
e de
lays
se
vera
l tim
es p
er w
eek
Few
can
cella
tions
, but
sho
rt fo
rmat
ions
no
t unc
omm
on
NB
: Due
to th
e gr
oup
envi
ronm
ent a
n el
emen
t of o
ver c
laim
may
exi
st
BU
SIN
ESS/
LEIS
UR
E D
elay
s, c
ance
llatio
ns a
nd d
isru
ptio
n ar
e ra
re
Mos
t can
reca
ll at
leas
t one
“hor
ror s
tory
” of
a jo
urne
y, th
ough
this
was
not
ne
cess
arily
rece
nt
RU
RA
L LI
NES
D
elay
s an
d ca
ncel
latio
ns o
ccur
mor
e of
ten
durin
g ‘b
ad’ w
eath
er
Inst
ance
s of
trai
ns n
ot s
topp
ing
at
sche
dule
d st
atio
ns
Wee
kend
ser
vice
ofte
n no
n-ex
iste
nt o
r re
duce
d in
freq
uenc
y
GLA
SGO
W
Pas
seng
ers
in G
lasg
ow fe
el th
e nu
mbe
r of
min
or d
elay
s an
d ca
ncel
latio
ns is
hig
h an
d th
at th
e se
rvic
e is
infe
rior c
ompa
red
with
oth
er p
arts
of B
ritai
n Lo
ng d
elay
s ar
e ra
re, w
ith tr
ains
eith
er
2-3
min
utes
late
or c
ance
lled
outri
ght
NB
: oth
er T
rans
port
Focu
s re
sear
ch in
S
cotla
nd s
how
ed a
bel
ief t
hat S
cotR
ail i
s to
o ca
utio
us a
nd in
clin
ed to
can
cel
serv
ices
at t
he ‘f
irst s
ign
of a
sno
wfla
ke’
“The
boa
rd w
ill s
ay th
at th
e tra
in is
del
ayed
by
thre
e m
inut
es, a
nd th
en it
will
just
cha
nge
to c
ance
lled,
ther
e is
not
hing
in b
etw
een.
” G
lasg
ow g
roup
, Com
mut
er, Y
oung
er.
Exp
erie
nce
of d
elay
s, d
isru
ptio
n an
d ca
ncel
latio
ns is
mix
ed
20
Pas
seng
ers
assu
me
that
per
form
ance
is m
easu
red
prim
arily
in te
rms
of p
unct
ualit
y, a
lthou
gh th
ey
have
no
defin
itive
kno
wle
dge
of th
is
Mos
t stru
ggle
to re
call
havi
ng h
eard
or s
een
anyt
hing
abo
ut p
erfo
rman
ce m
easu
rem
ent
Ref
er to
/mak
e co
mm
ent o
n m
edia
sto
ries
e.g.
: K
ings
Cro
ss C
hris
tmas
deb
acle
S
outh
ern
com
mut
er s
ervi
ce fr
om B
right
on
to V
icto
ria th
at w
as ‘l
ate
ever
y da
y’
Som
e m
entio
ns o
f pos
ters
in ra
il st
atio
ns s
how
ing
‘num
bers
of s
ome
sort’
ab
out p
unct
ualit
y P
artic
ular
ly in
Lon
don,
pas
seng
ers
quer
y th
e re
liabi
lity
of th
ese
num
bers
as
ofte
n se
em h
igh
Ther
e is
no
clea
r und
erst
andi
ng o
f who
set
s pe
rform
ance
targ
ets
It is
pre
sum
ed th
at th
is is
don
e by
the
train
com
pani
es th
emse
lves
, the
‘Gov
ernm
ent’
or ‘s
omet
hing
of
ficia
l’.
LITT
LE/N
O K
NO
WLE
DG
E SO
ME
KN
OW
LED
GE
Lim
ited
know
ledg
e of
how
per
form
ance
is a
ctua
lly m
easu
red
21
Pre
viou
s re
sear
ch*
has
high
light
ed th
e im
porta
nce
of re
liabi
lity
for t
rain
com
pani
es in
atta
inin
g an
d bu
ildin
g tru
st
Ther
e ar
e ot
her f
acto
rs in
add
ition
to re
liabi
lity
and
man
y of
thes
e ar
e la
ckin
g am
ongs
t tra
in c
ompa
nies
Th
eref
ore
pass
enge
rs a
re o
ften
cyni
cal w
hen
talk
ing
abou
t the
ext
ent t
o w
hich
they
trus
t the
ir tra
in
com
pany
C
onse
quen
tly, t
hey
are
also
pes
sim
istic
whe
n ta
lkin
g in
term
s of
rail
targ
ets;
how
they
are
set
, who
set
s th
em a
nd th
e ex
tent
to w
hich
they
are
ach
ieve
d
Rel
iabi
lity
is k
ey to
TO
Cs
build
ing
trust
Thos
e sa
tisfie
d w
ith th
e pu
nctu
ality
of t
heir
train
ser
vice
ar
e m
ore
likel
y to
trus
t the
TO
C
For c
omm
uter
s, p
unct
ualit
y is
m
ost i
mpo
rtant
to b
uild
ing
trust
, gi
ven
that
an
on-ti
me
serv
ice
is
mos
t des
ired
For b
usin
ess/
leis
ure
pass
enge
rs, e
mot
iona
l fac
tors
ar
e al
so c
ited
as b
eing
key
to
esta
blis
hing
trus
t, su
ch a
s st
aff
enga
gem
ent,
qual
ity o
f rol
ling
stoc
k an
d cu
stom
er s
ervi
ce.
“I tru
st th
e on
e th
at I
use
beca
use
they
get
me
to
wor
k on
tim
e, a
nd it
’s v
ery
impo
rtant
that
I ge
t to
wor
k on
tim
e ot
herw
ise
I’m la
te fo
r cou
rt.”
Der
by g
roup
, Com
mut
er, O
lder
.
“In te
rms
of q
ualit
y of
the
actu
al tr
ain
carr
iage
s I
thin
k V
irgin
are
way
ahe
ad a
nd th
at in
fluen
ces
my
trust
in th
em.”
Der
by g
roup
, Bus
ines
s/Le
isur
e, Y
oung
er.
* http
://w
ww
.tran
spor
tfocu
s.or
g.uk
/rese
arch
/pub
licat
ions
/pas
seng
ers-
rela
tions
hip-
with
-the-
rail-
indu
stry
22
Rea
ctio
n to
PP
M a
nd C
aSL
perfo
rman
ce m
easu
res
23
Ove
rvie
w: r
eact
ions
to tr
ain
perfo
rman
ce m
easu
res
Pas
seng
ers
know
littl
e ab
out h
ow tr
ain
perfo
rman
ce is
mea
sure
d - k
now
ledg
e of
PP
M a
nd C
aSL
non-
exis
tent
How
ever
, mea
surin
g pe
rform
ance
is s
een
as im
porta
nt a
nd th
ere
is a
n ap
petit
e fo
r lea
rnin
g ab
out h
ow it
is m
easu
red
That
sai
d, m
ost p
asse
nger
s ex
pect
that
targ
ets
will
not
be
parti
cula
rly d
eman
ding
and
tend
to g
ive
low
est
imat
es w
hen
aske
d w
hat t
hey
belie
ve ta
rget
s w
ill b
e
For C
aSL,
the
head
line
targ
et o
f 2.2
per
cen
t was
see
n as
reas
onab
le
98 p
er c
ent n
ot c
ance
lled
was
typi
cally
sug
gest
ed a
s an
asp
iratio
n A
ccep
ted
that
‘thi
ngs
happ
en’ m
eani
ng s
ome
canc
ella
tions
inev
itabl
e
PP
M le
ss w
ell r
ecei
ved
98 p
er c
ent
- 99
per c
ent p
unct
ualit
y of
ten
the
typi
cal t
arge
t sug
gest
ed
As
such
, 92.
5 pe
r cen
t tar
get f
alls
wel
l sho
rt of
idea
l (al
thou
gh n
ot a
ltoge
ther
une
xpec
ted)
O
n re
flect
ion,
aro
und
95 p
er c
ent s
een
as ‘r
ealis
tic’ t
arge
t
How
ever
, onc
e th
e de
tail
of ta
rget
s is
reve
aled
, man
y fe
el th
at th
eir ‘
wor
st s
uspi
cion
s’ h
ave
been
con
firm
ed
Do
not b
elie
ve th
at m
easu
res
accu
rate
ly re
flect
true
per
form
ance
Fi
gure
s ‘m
assa
ged’
by
gene
rous
targ
ets
and
‘loop
hole
s’
Furth
erm
ore,
man
y pa
ssen
gers
stru
ggle
to u
nder
stan
d ho
w ta
rget
s im
pact
on
thei
r jou
rney
W
heth
er a
chie
ved
or n
ot, w
hat a
re th
e co
nseq
uenc
es?
Som
e m
ake
an e
xplic
it lin
k be
twee
n co
mpe
nsat
ion
for d
elay
or c
ance
llatio
n w
ith p
erfo
rman
ce ta
rget
s or
with
fine
s an
d pe
rform
ance
targ
ets
Whi
le th
ere
are
calls
for m
ore
strin
gent
rule
s in
resp
ect o
f fin
es a
nd c
ompe
nsat
ion,
this
is a
lso
seen
as
a do
uble
-edg
ed
swor
d –
perc
eptio
n th
at u
ltim
atel
y ‘th
e pa
ssen
ger a
lway
s pa
ys’
“And
at t
he e
nd o
f the
day
if th
at w
as o
kay,
they
wou
ld te
ll yo
u w
ould
n’t t
hey?
It w
ould
be
“Her
e’s
how
we
are
m
easu
red.
..” b
ut th
ey o
bvio
usly
don
’t w
ant t
o te
ll yo
u th
at b
ecau
se e
very
body
is g
oing
to s
ay th
at’s
not
righ
t.”
Car
diff
grou
p, C
omm
uter
, You
nger
.
“It’s
a b
it of
a c
on, t
hey’
re
mea
surin
g th
ose
figur
es in
a
way
that
mak
es th
em s
how
w
hat t
hey
wan
t the
m to
sho
w,
rath
er th
an m
easu
ring
it to
id
entif
y an
d fix
pro
blem
s, to
m
ake
it ac
tual
ly a
goo
d se
rvic
e.”
Lond
on g
roup
, Com
mut
er,
You
nger
.
24
Per
form
ance
mea
sure
men
t: in
itial
exp
ecta
tions
Pas
seng
ers
expe
ct p
unct
ualit
y to
be
mea
sure
d in
det
ail a
nd b
elie
ve th
at th
e te
chno
logy
can
/sho
uld
faci
litat
e th
is
STAT
ION
TO
C
JOU
RN
EY
Eve
ry s
ervi
ce o
pera
ted
by
ever
y TO
C s
houl
d be
m
easu
red
and
publ
ishe
d
All
jour
neys
sho
uld
be
mon
itore
d an
d pu
blis
hed
Dep
artu
re/a
rriv
al ti
mes
sh
ould
be
mea
sure
d fo
r ev
ery
train
sto
ppin
g at
eve
ry
stat
ion
Som
e su
gges
t a s
ampl
ing
appr
oach
as
long
as
the
sam
ple
still
mea
sure
d th
e tra
in’s
ent
ire ro
ute
(not
ju
st e
nd d
estin
atio
n st
atio
n).
“Eve
ry s
tatio
n it
stop
s at
bec
ause
it’s
all
very
w
ell i
f it l
eave
s G
lasg
ow a
nd g
ets
to E
dinb
urgh
at
the
right
tim
e bu
t if s
omew
here
alo
ng th
e lin
e, s
ay it
’s d
elay
ed a
nd y
ou a
rriv
e la
te o
r mis
s yo
ur c
onne
ctin
g tra
in?”
G
lasg
ow g
roup
, Com
mut
er, Y
oung
er.
“I kn
ow w
e’ve
got
the
tech
nolo
gy to
do
it b
ut if
you
can
get
a b
ig e
noug
h sa
mpl
e, it
’s s
tatis
tical
ly v
alid
an
yway
.” C
ardi
ff gr
oup,
Com
mut
er, Y
oung
er.
25
Unv
eilin
g of
PP
M to
resp
onde
nts
Pas
seng
er
thou
ghts
on
whe
n pu
nctu
ality
sh
ould
be
mea
sure
d
Pas
seng
er
perc
eptio
ns o
f w
hat ‘
on-ti
me’
m
eans
Pas
seng
er
estim
atio
n of
%
of tr
ains
that
ar
rive
on ti
me
Brie
f de
scrip
tion
of
PP
M g
iven
Initi
al
awar
enes
s of
P
PM
test
ed
Edu
cate
d th
at a
le
eway
exi
sts
for
train
s ar
rivin
g at
de
stin
atio
n st
atio
n
Edu
cate
d th
at
curr
ent t
arge
t is
92.5
% o
f tra
ins
arriv
ing
on ti
me
Ask
ed to
re
valu
ate
estim
atio
n fo
r %
of ‘
on-ti
me’
tra
ins
Edu
cate
d th
at
train
s ar
e on
ly
mea
sure
d fo
r pu
nctu
ality
at
end
dest
inat
ion
stat
ion
Edu
cate
d th
at
long
dis
tanc
e tra
ins
have
a 1
0 m
inut
e le
eway
Edu
cate
d th
at
shor
t dis
tanc
e tra
ins
have
a 5
m
inut
e le
eway
Firs
t st
age
Seco
nd
stag
e
Third
st
age
Pas
seng
ers
wer
e ed
ucat
ed a
bout
Pub
lic P
erfo
rman
ce M
easu
re (P
PM
) in
the
follo
win
g w
ay
Thei
r rea
ctio
ns w
ere
soug
ht a
fter e
ach
piec
e of
info
rmat
ion
was
giv
en
26
Ove
rvie
w: r
eact
ions
to P
PM
The
Pub
lic P
erfo
rman
ce M
easu
re (P
PM
) sho
ws
the
perc
enta
ge o
f tra
ins
whi
ch a
rriv
e at
thei
r te
rmin
atin
g st
atio
n on
tim
e.
Initi
al re
actio
ns a
re th
at P
PM
sho
uld
mea
sure
pun
ctua
lity
firm
ly a
nd s
tring
ently
B
road
con
sens
us th
at >
95 p
er c
ent o
f tra
ins
shou
ld b
e ar
rivin
g ‘o
n tim
e’
‘On
time’
sho
uld
mea
n to
the
min
ute
Pun
ctua
lity
at e
very
sto
p sh
ould
be
mea
sure
d
Initial reaction
Initi
al s
tring
ency
sof
tens
as
educ
ated
mor
e ab
out w
hat P
PM
mea
sure
s U
nder
stan
ding
that
a v
ery
smal
l ele
men
t of ‘
leew
ay’ m
ay b
e ne
eded
ar
ound
‘on
time’
mea
sure
Considered reaction
Sce
ptic
ism
set
s in
and
pas
seng
ers
ques
tion
the
valu
e of
the
mea
sure
92
.5 p
er c
ent t
arge
t fee
ls fa
r too
low
(NB
- ta
rget
tend
s to
be
seen
as
a ce
iling
rath
er th
an a
bas
e le
vel)
Leew
ay c
onsi
dere
d ov
erly
gen
erou
s i.e
. 10
min
utes
‘lat
e’ c
onsi
dere
d to
o lo
ng to
stil
l be
coun
ted
as ‘o
n tim
e’
That
trai
ns a
re o
nly
mea
sure
d at
des
tinat
ion
stat
ion
furth
er c
ompo
unds
cy
nici
sm
Reaction after detailed
explanation
Awar
enes
s of
how
PP
M m
easu
re is
cal
cula
ted
furth
er u
nder
min
es tr
ust i
n tra
in
oper
ator
s.
27
Initi
al
awar
enes
s of
P
PM
test
ed
No
awar
enes
s of
PP
M
Pas
seng
er
estim
atio
n of
%
of t
rain
s th
at a
rriv
e on
tim
e
Est
imat
es ra
nge
from
70
per c
ent –
90
per c
ent
Thos
e w
ho s
ugge
st lo
wer
figu
res
argu
e th
is re
flect
s tru
e ex
perie
nces
Th
ose
that
pre
dict
hig
her f
igur
es b
elie
ve th
at s
ome
of th
e sh
orte
r, le
ss c
onge
sted
se
rvic
es a
re m
ore
relia
ble
and
boos
t the
ave
rage
S
ome
belie
ve th
at tr
ain
com
pani
es w
ould
not
wan
t to
repo
rt lo
w fi
gure
s, a
nd
“mas
sage
” the
figu
res
acco
rdin
gly
“I th
ink
peop
le h
ave
a pe
rcep
tion
that
it’s
a lo
t lo
wer
but
ther
e’s
load
s of
diff
eren
t com
pani
es a
nd
jour
neys
that
are
muc
h sh
orte
r and
pro
babl
y on
-tim
e m
ost o
f the
tim
e.”
Car
diff
grou
p, C
omm
uter
, You
nger
.
Pas
seng
er
perc
eptio
ns o
f w
hat ‘
on-ti
me’
m
eans
For t
he m
ajor
ity, ‘
on-ti
me’
mea
ns to
the
min
ute
Som
e ac
cept
ance
that
a s
mal
l lee
way
(one
/tw
o m
inut
es) m
ay b
e ju
stifi
ed
Som
e re
cogn
ition
that
it is
like
ly to
diff
er
depe
nden
t on
the
jour
ney
leng
th
“If it
arr
ives
at 7
.01a
m in
stea
d of
6.5
9am
to m
e th
at is
late
; it’
s no
t the
tim
e th
ey
adve
rtise
d it.
Not
that
late
I w
ill a
ccep
t, bu
t it’s
stil
l lat
e.”
Lond
on g
roup
, Com
mut
er,
You
nger
.
Edu
cate
d th
at
curr
ent t
arge
t is
92.
5% o
f tra
ins
arriv
ing
on ti
me
92.5
per
cen
t ini
tially
sou
nds
high
as
a re
flect
ion
of tr
ue tr
ain
punc
tual
ity (
alth
ough
no
t par
ticul
arly
hig
h as
a ta
rget
) S
ome
sugg
est t
hat t
arge
t sho
uld
be m
ore
aspi
ratio
nal
Sho
uld
striv
e to
be
the
best
i.e.
100
per
cen
t A
nd s
houl
d be
pro
gres
sive
e.g
. sho
w p
rogr
ess
year
on
year
.
Det
aile
d vi
ew: r
eact
ions
to P
PM
28
Edu
cate
d th
at
shor
t dis
tanc
e tra
ins
have
a 5
m
inut
e le
eway
Edu
cate
d th
at
long
dis
tanc
e tra
ins
have
a
10 m
inut
e le
eway
Edu
cate
d th
at
train
s ar
e on
ly
mea
sure
d fo
r pu
nctu
ality
at
end
dest
inat
ion
stat
ion
Ask
ed to
re
valu
ate
estim
atio
n fo
r %
of ‘
on-ti
me’
tra
ins
Pas
seng
ers
ofte
n ‘o
utra
ged’
at i
ndus
try’s
def
initi
on o
f ‘on
-tim
e’
Als
o di
sbel
ief t
hat t
his
is n
ot m
ore
wid
ely
com
mun
icat
ed
Som
e su
gges
t alte
rnat
ives
: Le
eway
cal
cula
ted
as a
per
cent
age
of tr
ain
jour
ney
Le
eway
bui
lt in
to th
e tim
etab
le, p
artic
ular
ly
fo
r lon
g di
stan
ce jo
urne
ys (a
s pe
r airl
ines
)
“If y
ou’re
giv
en a
tim
e an
d it
gets
pas
t tha
t tim
e, th
en in
my
opin
ion
that
’s la
te –
ho
w c
an y
ou s
ay it
’s o
n tim
e if
it ar
rived
10
min
utes
late
?”
Gla
sgow
gro
up,
Bus
ines
s/Le
isur
e,
Old
er
Pas
seng
ers
furth
er ir
ritat
ed b
y th
is
Add
s to
the
perc
eptio
n th
at tr
ain
com
pani
es ‘m
assa
ge’ t
he fi
gure
s
Con
sens
us th
at p
unct
ualit
y sh
ould
be
mea
sure
d at
eve
ry s
tatio
n
Oth
erw
ise
not a
ccur
ate
give
n m
any
peop
le w
ill d
isem
bark
bef
ore
final
de
stin
atio
n
A pe
rcep
tion
that
92.
5 pe
r cen
t is
not a
true
refle
ctio
n of
trai
n pu
nctu
ality
G
iven
all
of th
e ca
veat
s, a
sco
re o
f 99
per c
ent i
s no
t see
n as
unr
easo
nabl
e.
“We
all s
aid
that
we
care
abo
ut if
it g
ets
to o
ur
stat
ion
on ti
me,
so
this
isn'
t rea
lly te
lling
us
anyt
hing
abo
ut th
at, i
t’s ju
st te
lling
us
whe
ther
it
gets
to w
hich
ever
the
end
stat
ion
is.”
Car
diff
grou
p, C
omm
uter
, You
nger
.
Det
aile
d vi
ew: r
eact
ions
to P
PM
29
Unv
eilin
g of
CaS
L to
resp
onde
nts
Brie
f de
scrip
tion
of C
aSL
give
n
Ask
ed to
es
timat
e w
hat
‘sig
nific
antly
la
te’ m
eans
Initi
al
awar
enes
s of
CaS
L te
sted
Ext
ent t
o w
hich
ca
ncel
latio
ns
and
sign
ifica
nt
late
ness
sho
uld
be m
easu
red
toge
ther
Ask
ed to
re
valu
ate
estim
atio
n fo
r %
of c
ance
lled/
si
gnifi
cant
ly la
te
train
s
Edu
cate
d th
at
curr
ent t
arge
t is
2.2%
of t
rain
s ca
ncel
led
or
sign
ifica
ntly
late
Pas
seng
er
estim
atio
n of
%
of tr
ains
that
are
ca
ncel
led
or
sign
ifica
ntly
late
Edu
cate
d th
at 3
0 m
inut
es la
te is
co
nsid
ered
as
‘sig
nific
antly
’ lat
e
Edu
cate
d th
at
Sco
tRai
l is
not m
easu
red
agai
nst C
aSL
Firs
t st
age
Seco
nd
stag
e
Gla
sgow
on
ly
Pas
seng
ers
wer
e ed
ucat
ed a
bout
Can
cella
tions
and
Sig
nific
ant L
aten
ess
(CaS
L) in
the
follo
win
g w
ay
Thei
r rea
ctio
ns w
ere
soug
ht a
fter e
ach
piec
e of
info
rmat
ion
was
giv
en
30
CaS
L is
def
ined
as
the
num
ber a
nd p
erce
ntag
e of
pas
seng
er tr
ains
(fra
nchi
sed
and
open
ac
cess
ope
rato
rs) w
hich
are
can
celle
d in
par
t or f
ull,
or w
hich
arr
ive
at th
eir f
inal
des
tinat
ion
30 o
r mor
e m
inut
es la
ter t
han
the
time
show
n in
the
publ
ic ti
met
able
.
Pas
seng
er d
efin
ition
of ‘
sign
ifica
ntly
late
’ con
side
rabl
y m
ore
strin
gent
th
an th
e in
dust
ry m
easu
re
10 m
inut
e ‘c
ut-o
ff’ fo
r sho
rt di
stan
ce a
nd 2
0 m
inut
es fo
r lon
g di
stan
ce tr
ains
C
ance
llatio
ns s
een
as re
quiri
ng a
mor
e st
ringe
nt ta
rget
than
la
tene
ss –
98
per c
ent n
ot c
ance
lled
emer
ges
as c
onse
nsus
figu
re.
S
ome
indi
gnat
ion
at in
dust
ry d
efin
ition
of ‘
sign
ifica
ntly
late
’ - 3
0 m
inut
es
cons
ider
ed to
o ge
nero
us
How
ever
, 2.2
per
cen
t as
head
line
figur
e co
nsid
ered
a re
ason
able
targ
et fo
r ca
ncel
latio
ns a
lbei
t mor
e de
tail
rega
rdin
g re
ason
s fo
r can
cella
tion
ofte
n de
sire
d M
easu
ring
sign
ifica
nt la
tene
ss a
nd c
ance
latio
ns to
geth
er is
con
fusi
ng fo
r so
me.
Reaction after detailed explanation
Initial reaction
Ove
rvie
w: r
eact
ions
to C
aSL
31
Initi
al
awar
enes
s of
C
aSL
test
ed
Ask
ed to
es
timat
e w
hat
‘sig
nific
antly
la
te’ m
eans
No
awar
enes
s of
CaS
L
Pas
seng
ers
estim
ate
that
‘sig
nific
antly
late
’ wou
ld in
clud
e sh
ort d
ista
nce
train
s th
at a
re m
ore
than
10
min
utes
late
and
long
dis
tanc
e tra
ins
that
are
mor
e th
an
20 m
inut
es la
te
i.e. d
oubl
e th
e al
low
ance
s th
at P
PM
mak
es fo
r lat
e tra
ins
Edu
cate
d th
at
30 m
inut
es
late
is
cons
ider
ed a
s ‘s
igni
fican
tly’
late
Incl
udin
g si
gnifi
cant
late
ness
with
in c
ance
llatio
n fig
ures
cau
ses
som
e co
nfus
ion
Som
e ag
ree
that
30
min
s de
lay
is a
s ba
d as
a c
ance
llatio
n an
d so
sho
uld
be c
ount
ed a
s su
ch
Oth
ers
susp
ect i
t may
be
a w
ay o
f ‘hi
ding
’ lat
enes
s
Pas
seng
er
estim
atio
n of
%
of t
rain
s th
at a
re
canc
elle
d or
si
gnifi
cant
ly
late
Est
imat
es a
re 9
5-98
per
cen
t of t
rain
s ar
e no
t can
celle
d or
sig
nific
antly
late
P
asse
nger
s sp
eak
in te
rms
of th
e nu
mbe
rs o
f tra
ins
that
are
not
del
ayed
or
sig
nific
antly
late
R
athe
r tha
n th
e nu
mbe
r tha
t are
can
celle
d or
sig
nific
antly
late
.
“I ca
n't u
nder
stan
d w
hy it
is 3
0 m
inut
es a
nd I
wou
ld a
im to
redu
ce
that
. I th
ink
they
hav
e gi
ven
them
selv
es q
uite
a lo
t of l
eew
ay
actu
ally
with
30
min
utes
.” G
lasg
ow g
roup
, Bus
ines
s/Le
isur
e,
Old
er.
Det
aile
d vi
ew: r
eact
ions
to C
aSL
32
Edu
cate
d th
at
curr
ent t
arge
t is
2.2%
of t
rain
s ca
ncel
led
or
sign
ifica
ntly
late
2.2
per c
ent c
onsi
dere
d a
reas
onab
le ta
rget
by
mos
t O
nly
a m
inor
ity a
rgue
for s
omet
hing
mor
e st
ringe
nt
Ext
ent t
o w
hich
ca
ncel
latio
ns
and
sign
ifica
nt
late
ness
sho
uld
be m
easu
red
toge
ther
Ask
ed to
re
valu
ate
estim
atio
n fo
r %
of c
ance
lled/
si
gnifi
cant
ly la
te
train
s N
ot a
cle
ar c
onse
nsus
on
sign
ifica
nt la
tene
ss
Sug
gest
ed s
igni
fican
t lat
enes
s sh
ould
be
mea
sure
d al
ongs
ide
PP
M
(som
e as
sum
ed th
at in
clud
ing
a tra
in w
ithin
CaS
L bo
osts
PP
M s
core
s fu
rther
) A
lso
man
y ar
gue
that
cla
ssifi
catio
n fo
r ‘si
gnifi
cant
ly la
te’ s
houl
d be
diff
eren
t for
sh
ort a
nd lo
ng d
ista
nce
30 m
inut
e de
lay
on a
10
min
ute
jour
ney
mor
e si
gnifi
cant
than
on
a 3
hour
jo
urne
y C
ould
be
calc
ulat
ed a
s a
perc
enta
ge o
f jou
rney
tim
e
Edu
cate
d th
at
Sco
tRai
l is
not
mea
sure
d ag
ains
t CaS
L
Pas
seng
ers
in G
lasg
ow c
anno
t com
preh
end
why
Sco
tRai
l is
not m
easu
red
agai
nst
CaS
L S
pecu
late
that
this
is b
ecau
se o
f the
var
iabl
e an
d “s
ever
e” w
eath
er
But
not
con
side
red
a le
gitim
ate
reas
on fo
r exc
lusi
on
Sup
ports
pas
seng
er p
erce
ptio
n th
at d
elay
s ar
e ei
ther
min
or, o
r tra
ins
are
canc
elle
d ou
trigh
t.
“Wha
t do
you
expe
ct?
99 p
er c
ent?
I w
ould
se
ttle
for 9
9 pe
r cen
t for
eve
ryth
ing.
” G
lasg
ow g
roup
, Bus
ines
s/Le
isur
e, O
lder
.
“2.2
per
cen
t? I
thin
k th
at’s
qu
ite re
ason
able
. We
can
live
with
that
.” Lo
ndon
gro
up, B
usin
ess/
Leis
ure,
Old
er
Det
aile
d vi
ew: r
eact
ions
to u
nvei
ling
of C
aSL
33
Cur
rent
ly, a
ll tra
ins
coun
t equ
ally
tow
ards
pun
ctua
lity
targ
ets
– ea
ch in
divi
dual
trai
n ei
ther
do
es o
r doe
sn’t
arriv
e w
ithin
5 o
r 10
min
utes
of s
ched
uled
tim
e. T
here
is th
eref
ore
exac
tly
the
sam
e in
cent
ive,
from
a p
erfo
rman
ce ta
rget
per
spec
tive,
to ru
n a
train
on
time
that
has
1,
000
pass
enge
rs o
n it
as o
ne w
ith 2
0 pa
ssen
gers
.
On
bala
nce,
mos
t pas
seng
ers
belie
ve th
at a
ll tra
ins
shou
ld c
ount
equ
ally
tow
ards
pun
ctua
lity
targ
ets
They
arg
ue th
at e
very
one
has
mad
e th
e sa
me
cont
ract
with
the
TOC
and
eac
h jo
urne
y sh
ould
ha
ve th
e sa
me
wei
ght
Ther
e is
som
e ar
gum
ent (
parti
cula
rly b
y co
mm
uter
s) th
at jo
urne
ys s
houl
d be
wei
ghte
d by
pas
seng
er
num
bers
A
nd s
ome
susp
icio
n th
at ta
rget
s m
ight
be
mas
sage
d by
ope
rato
rs ru
nnin
g tra
ins
on ‘e
asy’
se
rvic
es to
‘up’
the
aver
age
How
ever
, als
o be
lieve
that
wei
ghtin
g by
pas
seng
er n
umbe
rs m
ight
intro
duce
per
vers
e in
cent
ives
If
targ
ets
are
skew
ed to
bus
y tra
ins,
it is
fear
ed th
at th
e tra
in c
ompa
nies
may
dro
p se
rvic
e st
anda
rds
on q
uiet
er s
ervi
ces
and
not s
trive
to e
nsur
e op
timum
pun
ctua
lity.
“I do
n't t
hink
they
sho
uld
or o
ther
wis
e th
ey
will
end
up
mak
ing
all t
he tr
ains
for l
eisu
re
four
hou
rs la
te, a
nd th
e re
st o
n tim
e!”
Lond
on g
roup
, Bus
ines
s/Le
isur
e, O
lder
.
“But
are
they
look
ing
at th
at s
ayin
g th
ere’
s on
ly 2
0 pe
ople
on
it so
let’s
can
cel i
t bec
ause
it’
s no
t eco
nom
ical
to ru
n it?
“ G
lasg
ow g
roup
, Com
mut
er, Y
oung
er.
Det
aile
d vi
ew: r
eact
ion
to c
alcu
latio
n of
pun
ctua
lity
targ
ets
34
Kno
wle
dge
of P
PM
/CaS
L an
d ho
w e
ach
mea
sure
is c
alcu
late
d fu
rther
und
erm
ines
pas
seng
ers’
trus
t in
the
railw
ay fo
r a n
umbe
r of r
easo
ns:
Bel
ief t
hat t
arge
ts a
re n
ot s
tretc
hing
eno
ugh
over
all
Targ
ets
seen
as
cont
aini
ng n
umer
ous
loop
hole
s an
d ca
veat
s th
at a
llow
the
indu
stry
to m
assa
ge
the
figur
es
Exp
ecta
tion
that
targ
ets
will
be
treat
ed a
s a
ceili
ng fo
r ser
vice
per
form
ance
and
that
ope
rato
rs
are
only
ince
ntiv
ised
to ‘g
et o
ver t
he li
ne’.
Effe
ct o
n tru
st
“You
trus
t the
m le
ss, k
now
ing
the
deta
il be
hind
the
mea
sure
now
, it’s
like
no
t wha
t we
wou
ld e
xpec
t, ou
r pe
rcep
tion
of 9
5 pe
r cen
t on
time
is
not a
ctua
lly a
ccur
ate.
” Lo
ndon
gro
up, B
usin
ess/
Leis
ure,
O
lder
.
“I sa
id b
efor
e th
at I
trust
if I
had
a co
mpl
aint
but
now
I th
ink
they
’d p
roba
bly
do th
e sa
me
with
thei
r com
plai
nt fi
gure
s an
d sa
y “o
h w
ell,
this
com
plai
nt d
oesn
't
coun
t for
this
reas
on”,
wha
teve
r.”
Gla
sgow
gro
up, B
usin
ess/
Leis
ure,
Old
er.
“A b
unch
of t
hiev
ing
rogu
es!”
Lond
on g
roup
, Bus
ines
s/Le
isur
e,
Old
er.
35
Spe
ed, p
unct
ualit
y,
frequ
ency
trad
e-of
fs
36
Pun
ctua
lity
mos
t des
ired
over
spe
ed, f
requ
ency
Mos
t pas
seng
ers
(and
com
mut
ers
in p
artic
ular
) des
ire m
ore
punc
tual
trai
ns
This
is b
ecau
se p
asse
nger
s’ p
rimar
y ai
m is
to a
rriv
e at
thei
r des
tinat
ion
‘on
time’
, reg
ardl
ess
of
how
long
that
jour
ney
take
s A
lso,
refle
cts
belie
f / e
xper
ienc
e th
at d
elay
s oc
cur m
ore
frequ
ently
than
can
cella
tions
Ther
e ar
e a
few
exc
eptio
ns:
Spee
d de
sire
d on
ver
y lo
ng d
ista
nce
jour
neys
(Lon
don
– G
lasg
ow),
to re
duce
jour
ney
time
Freq
uenc
y de
sire
d on
rura
l jou
rney
s, to
ens
ure
train
s ru
n m
ore
regu
larly
Fr
eque
ncy
desi
red
by s
ome
leis
ure
trave
llers
, as
this
sho
uld
resu
lt in
mor
e se
ats
and
ther
efor
e le
ss c
row
ding
A w
ides
prea
d be
lief f
rom
pas
seng
ers
that
ther
e is
no
reas
on w
hy th
e tra
in c
ompa
nies
can
’t de
liver
an
impr
ovem
ent i
n te
rms
of a
ll th
ree
fact
ors:
spe
ed, f
requ
ency
and
pun
ctua
lity
The
indu
stry
’s a
rgum
ent t
hat a
ll th
ree
can’
t be
achi
eved
furth
er c
ompo
unds
a v
iew
that
pa
ssen
gers
’ bes
t int
eres
ts a
re n
ot p
rom
inen
t eno
ugh
Pas
seng
ers
also
arg
ue th
at a
‘one
siz
e fit
s al
l’ ap
proa
ch is
inap
prop
riate
and
that
diff
eren
t jou
rney
s re
quire
diff
eren
t im
prov
emen
t prio
ritie
s.
“Tha
t is
a rid
icul
ous
argu
men
t. I t
hink
bec
ause
w
e gi
ve th
em s
o m
uch
leew
ay th
at is
why
they
ca
n th
en s
ay th
ese
thin
gs.“
Lond
on g
roup
, Com
mut
er, Y
oung
er.
“To
me
it’s
too
top
dow
n, it
’s a
gen
eral
que
stio
n in
gen
eral
term
s, w
hat w
ould
you
like
? B
ut th
e sp
ecifi
c is
sues
are
spe
cific
rout
es s
o w
hy c
an’t
they
add
ress
them
rath
er th
an ju
st s
ayin
g w
e ca
n ha
ve fr
eque
ncy,
spe
ed o
r ?“
G
lasg
ow g
roup
, Com
mut
er, Y
oung
er.
37
Task
1
60%
40%
Com
mut
ers
mor
e lik
ely
to d
esire
mor
e ‘o
n tim
e’ tr
ains
, ref
lect
ing
mor
e fre
quen
t exp
erie
nce
of d
elay
s an
d m
ore
limite
d im
pact
s of
can
cella
tions
giv
en s
ervi
ce fr
eque
ncy
Thos
e ch
oosi
ng fe
wer
can
cella
tions
(mor
e of
ten
busi
ness
/leis
ure
trave
llers
) arg
ue th
at a
10
min
ute
dela
y is
not
hug
ely
sign
ifica
nt fo
r a lo
nger
jour
ney,
but
a c
ance
llatio
n co
uld
be a
‘dis
aste
r’.
“If I
was
doi
ng a
jour
ney,
say
I w
as
goin
g G
lasg
ow to
Man
ches
ter f
or a
tri
p aw
ay, i
f you
r tra
in’s
can
celle
d,
that
’s y
our t
rip ru
ined
.” G
lasg
ow g
roup
, Com
mut
er,
You
nger
.
“I pu
t the
firs
t one
just
bec
ause
I th
ink
ther
e’s
not
as m
any
canc
ella
tions
as
ther
e ar
e la
te tr
ains
. B
ut a
can
cella
tion
at th
e w
rong
tim
e co
uld
be
terr
ible
but
they
rare
ly g
et c
ance
lled,
com
pare
d to
the
amou
nt th
at a
re la
te.“
Lond
on g
roup
, Com
mut
er, Y
oung
er.
Bas
e: 8
0 C
AU
TIO
N: T
hese
figu
res
shou
ld b
e us
ed o
nly
as a
n in
dica
tion,
due
to th
e lo
w s
ampl
e si
ze
38
30%
70
%
Com
mut
ers,
aga
in a
re m
ore
likel
y to
des
ire m
ore
‘on
time’
trai
ns
Acc
usto
med
to c
row
ding
and
not
get
ting
a se
at
Pun
ctua
lity
and
getti
ng to
wor
k/m
eetin
gs o
n tim
e is
mor
e im
porta
nt
Bus
ines
s/le
isur
e pa
ssen
gers
des
ire m
ore
com
fort
– cr
owdi
ng/la
ck o
f sea
ting
is a
mor
e si
gnifi
cant
issu
e
Som
e co
mm
uter
s ar
gue
in fa
vour
of a
dditi
onal
trai
ns o
n gr
ound
s th
at s
ervi
ces
ofte
n se
vere
ly
over
crow
ded N
ot a
que
stio
n of
not
get
ting
a se
at, b
ut o
f not
bei
ng a
ble
to g
et o
n th
e tra
in a
t all.
“But
then
you
don
’t re
ally
min
d st
andi
ng o
n th
e tra
in fo
r 15
min
utes
eve
ry m
orni
ng, y
ou w
ould
n't
wan
t to
be s
tand
ing
from
Gla
sgow
to L
ondo
n.“
Gla
sgow
gro
up, B
usin
ess/
Leis
ure,
Old
er.
“I'd
rath
er g
et s
quas
hed
for a
littl
e bi
t. B
ecau
se
we’
re u
sed
to g
ettin
g sq
uash
ed.“
Lond
on g
roup
, Com
mut
er, Y
oung
er.
Task
2
Bas
e: 8
0 C
AU
TIO
N: T
hese
figu
res
shou
ld b
e us
ed o
nly
as a
n in
dica
tion,
due
to th
e lo
w s
ampl
e si
ze
39
Littl
e de
sire
for a
fast
er s
ervi
ce in
exc
hang
e fo
r mor
e er
ratic
pun
ctua
lity
Pas
seng
ers
do n
ot m
ind
how
long
a s
ervi
ce ta
kes,
pro
vidi
ng it
arr
ives
at t
he s
ched
uled
tim
e A
llow
s th
em to
pla
n/sc
hedu
le
Bel
ief t
hat f
aste
r doe
sn’t
mea
n si
gnifi
cant
ly fa
ster
and
is li
kely
to b
e on
ly o
ne/tw
o m
inut
es
quic
ker.
“If th
ey te
ll m
e it’
s go
ing
to ta
ke 1
0 m
inut
es,
I'm h
appy
for i
t to
take
10
min
utes
, don
’t re
ally
ca
re if
it ta
kes
eigh
t.“
Lond
on g
roup
, Com
mut
er, Y
oung
er.
“Yea
h, if
you
r tra
in jo
urne
y ta
kes
norm
ally
20
min
utes
and
they
can
redu
ce it
to 1
0 m
inut
es
but t
hey
coul
d be
10
min
utes
late
, the
n yo
u're
st
ill g
ettin
g fro
m y
our h
ouse
to y
our u
ltim
ate
dest
inat
ion
at th
e sa
me
time.
I’d
rath
er th
ey
wer
e m
ore
accu
rate
.“ Lo
ndon
gro
up, B
usin
ess/
Leis
ure,
Old
er.
Task
3
10%
90%
B
ase:
80
CA
UTI
ON
: The
se fi
gure
s sh
ould
be
used
onl
y as
an
indi
catio
n, d
ue to
the
low
sam
ple
size
40
73%
17
%
10%
“I th
ink
gene
rally
mor
e tra
ins
are
dela
yed
than
th
ey a
re c
ance
lled
or d
isru
pted
so
if yo
u're
taki
ng
an a
vera
ge, I
wou
ld ra
ther
that
they
wer
e m
ore
on
time
and
a co
uple
of c
ance
llatio
ns.”
Lond
on g
roup
, Com
mut
er, Y
oung
er.
The
maj
ority
des
ire m
ore
train
s ar
rivin
g on
tim
e –
refle
cts
belie
f / e
xper
ienc
e th
at d
elay
s oc
cur m
ore
frequ
ently
than
can
cella
tions
Thos
e w
ho o
pt fo
r few
er d
ays
of m
ajor
dis
rupt
ion
argu
e th
at th
is w
ill re
duce
the
num
ber o
f peo
ple
stra
nded
/una
ble
to g
et s
omew
here
, par
ticul
arly
on
long
er jo
urne
ys
Whe
reas
can
cella
tions
/del
ays
may
stil
l allo
w a
rriv
al a
t des
tinat
ion,
just
late
.
“I pu
t num
ber t
hree
aga
in b
ecau
se if
you
can
't ge
t th
ere
at a
ll , i
f you
hav
e a
day
of m
ajor
dis
rupt
ion,
if
you
can'
t get
ther
e at
all
it co
uld
real
ly m
ess
up
a ho
liday
or w
hate
ver ,
whe
reas
bei
ng s
light
ly la
te
isn’
t the
end
of t
he w
orld
.” Lo
ndon
Gro
up, B
usin
ess/
Leis
ure,
Old
er.
Task
4
Bas
e: 8
0 C
AU
TIO
N: T
hese
figu
res
shou
ld b
e us
ed o
nly
as a
n in
dica
tion,
due
to th
e lo
w s
ampl
e si
ze
41
Per
form
ance
mea
sure
men
t tra
de-o
ff ex
erci
ses
42
Lear
ning
abo
ut s
ome
of th
e cr
iteria
aga
inst
whi
ch P
PM
and
CaS
L ar
e m
easu
red
furth
er e
rode
s pa
ssen
gers
’ per
cept
ions
of t
he c
redi
bilit
y of
thes
e m
easu
res
Pas
seng
ers
are,
aga
in, l
eft ‘
outra
ged’
at s
ome
of th
e ta
ctic
s th
at th
e tra
in c
ompa
nies
can
em
ploy
, sp
ecifi
cally
: A
ddin
g ex
tra ti
me
to th
e tim
etab
le b
etw
een
the
penu
ltim
ate
and
last
sto
p C
ance
l tra
ins
by 1
0pm
the
nigh
t bef
ore
they
are
due
to ru
n an
d th
ey d
on’t
coun
t as
canc
elle
d M
iss
out s
tatio
ns to
sav
e tim
e an
d ge
t sub
sequ
ent t
rain
s ru
nnin
g to
sch
edul
e
This
furth
er s
treng
then
s th
e pe
rcep
tion
that
PP
M/C
aSL
does
not
pro
vide
a re
pres
enta
tive
view
of t
he
train
ser
vice
as
too
man
y ‘lo
opho
les’
exi
st.
“I
thin
k th
e fig
ures
they
see
m to
pub
lish
are
certa
inly
pub
lishe
d to
pul
l the
woo
l ove
r our
eye
s. If
they
just
told
us
wha
t was
ha
ppen
ing,
we
coul
d sa
y “Y
eah,
that
’s o
kay,
we
unde
rsta
nd it
’s
a di
fficu
lt jo
b yo
u’ve
got
to d
o, w
e un
ders
tand
ther
e’s
goin
g to
be
del
ays
and
canc
ella
tions
”, w
e un
ders
tand
all
that
, if y
ou ju
st
told
us
the
truth
then
we’
d be
abl
e to
dea
l with
that
.” G
lasg
ow g
roup
, Bus
ines
s/Le
isur
e, O
lder
.
Furth
er le
arni
ng o
n P
PM
/CaS
L
43
28%
22
%
25%
15
%
10%
Not
a c
lear
con
sens
us
Som
e se
e th
is a
s ac
cept
able
pro
vidi
ng th
ey a
rriv
e at
thei
r des
tinat
ion
at th
e sc
hedu
led
time
– ex
pect
atio
ns a
re m
anag
ed
Man
y fe
el it
is u
nacc
epta
ble
Ano
ther
way
that
the
figur
es a
re m
anip
ulat
ed
Not
a fa
ir/ge
nuin
e re
flect
ion
of h
ow th
e se
rvic
e ha
s pe
rform
ed
View
ed a
s m
ore
acce
ptab
le if
add
ition
al ti
me
is a
dded
on
at e
very
sta
tion
the
train
sto
ps a
t E
.g. o
ne/tw
o m
inut
es
But
not
just
bet
wee
n pe
nulti
mat
e an
d en
d st
atio
n.
“It’s
not
fixi
ng th
e pr
oble
m. I
t’s ju
st ly
ing
and
sayi
ng
that
it is
on
time,
but
it’s
not
.” G
lasg
ow g
roup
, Bus
ines
s/Le
isur
e, O
lder
.
“If it
sai
d a
coup
le o
f min
utes
fro
m w
hen
it ac
tual
ly a
rriv
es a
t its
des
tinat
ion,
you
kno
w I
wou
ld
rath
er p
erso
nally
that
than
it
arriv
e tw
o m
inut
es la
te a
nd I
mis
s m
y co
nnec
tion.
” G
lasg
ow g
roup
, Com
mut
er,
You
nger
.
“If a
trai
n sa
id it
add
ed a
n ex
tra m
inut
e on
the
time
betw
een
each
sto
p an
d it
actu
ally
took
that
tim
e,
whe
ther
that
mea
nt it
was
goi
ng s
light
ly s
low
er o
r w
hate
ver,
that
I w
ould
n't h
ave
a pr
oble
m w
ith.”
Lond
on g
roup
, Com
mut
er, Y
oung
er.
Task
1
Bas
e: 8
0 C
AU
TIO
N: T
hese
figu
res
shou
ld b
e us
ed o
nly
as a
n in
dica
tion,
due
to th
e lo
w s
ampl
e si
ze
44
80%
20%
Larg
e m
ajor
ity a
gree
that
an
inte
rval
mea
sure
men
t sys
tem
wou
ld b
e m
ore
appr
opria
te
Per
cept
ion
that
this
wou
ld p
rovi
de a
mor
e ac
cura
te re
flect
ion
of p
unct
ualit
y ex
perie
nced
by
the
maj
ority
of p
asse
nger
s.
“It’s
stil
l goi
ng to
inco
nven
ienc
e th
e fir
st p
eopl
e at
the
plat
form
or s
ome
peop
le o
n on
e tra
in, b
ut fo
r the
maj
ority
it’
s pr
obab
ly g
oing
to im
prov
e th
eir e
xper
ienc
e.”
Lond
on g
roup
, Com
mut
er, Y
oung
er.
Task
2
Bas
e: 8
0 C
AU
TIO
N: T
hese
figu
res
shou
ld b
e us
ed o
nly
as a
n in
dica
tion,
due
to th
e lo
w s
ampl
e si
ze
45
10%
90%
Mix
ed fe
elin
gs e
xist
on
leaf
mul
ch
Som
e re
cogn
ise
that
it is
a g
enui
ne p
robl
em
Oth
ers
stru
ggle
with
the
conc
ept a
nd d
on’t
belie
ve o
ther
cou
ntrie
s ha
ve s
uch
prob
lem
s
Pas
seng
ers
desi
re a
pre
dict
able
ser
vice
rath
er th
an u
ncer
tain
ty a
nd d
o no
t opp
ose
an a
men
ded
timet
able
S
afet
y is
par
amou
nt
Any
am
ende
d tim
etab
le w
ould
nee
d to
be
in p
lace
for a
lim
ited
perio
d i.e
. tw
o/th
ree
mon
ths
With
adv
ance
not
ice
abou
t its
intro
duct
ion
Thos
e ag
ains
t the
idea
arg
ue th
at th
e TO
Cs
rece
ive
enou
gh le
eway
as
it is
W
hils
t a la
ck o
f tru
st re
sults
in p
asse
nger
s be
lievi
ng th
at T
OC
s w
ould
intro
duce
the
timet
able
s to
o ea
rly o
r whe
n le
af m
ulch
isn'
t a p
robl
em.
“Peo
ple
just
nee
d to
pla
n th
eir
jour
ney,
peo
ple
acce
pt
diffe
rent
tim
es o
f yea
r tha
t yo
u m
ight
not
get
the
sam
e le
vel o
f ser
vice
, it’s
kno
win
g in
adv
ance
so
you
can
take
ot
her a
ctio
n to
get
ther
e.“
Gla
sgow
gro
up, C
omm
uter
, Y
oung
er.
Task
3
Bas
e: 8
0 C
AU
TIO
N: T
hese
figu
res
shou
ld b
e us
ed o
nly
as a
n in
dica
tion,
due
to th
e lo
w s
ampl
e si
ze
46
40%
60
%
Pas
seng
ers
divi
ded
on th
e in
trodu
ctio
n of
an
amen
ded
timet
able
Saf
ety
is p
aram
ount
A
n am
ende
d tim
etab
le is
like
ly to
be
mor
e pu
nctu
al
But
, any
am
ende
d tim
etab
le
shou
ld b
e co
mm
unic
ated
wel
l in
adva
nce
TOC
s sh
ould
at l
east
atte
mpt
to d
eliv
er th
e fu
ll tim
etab
le
Rat
her t
han
adm
ittin
g de
feat
st
raig
htaw
ay
P
asse
nger
s’ p
erce
ptio
ns o
f ext
rem
e w
eath
er li
kely
to b
e si
gnifi
cant
ly d
iffer
ent
from
TO
Cs’
P
asse
nger
s st
rand
ed a
t non
-sto
ppin
g st
atio
ns
Supp
ort a
n am
ende
d tim
etab
le
Opp
ose
an a
men
ded
timet
able
Per
form
ance
sho
uld
be m
easu
red
agai
nst t
he a
men
ded
timet
able
, pro
vidi
ng p
asse
nger
s ar
e gi
ven
notic
e of
this
and
it is
not
intro
duce
d at
ver
y sh
ort n
otic
e.
“I th
ink
that
they
sho
uld
atte
mpt
to d
o it.
In m
y ex
perie
nce
ther
e’s
like
liter
ally
a c
oupl
e of
sn
owfla
kes
and
that
’s it
the
who
le n
etw
ork
goes
off
. To
be h
ones
t the
y w
ould
brin
g in
an
amen
ded
timet
able
too
soon
, the
y’d
say
right
that
’s
it.”
Gla
sgow
gro
up, C
omm
uter
, Y
oung
er.
Task
4
Bas
e: 8
0 C
AU
TIO
N: T
hese
figu
res
shou
ld b
e us
ed o
nly
as a
n in
dica
tion,
due
to th
e lo
w s
ampl
e si
ze
47
66%
30
%
4%
1%
0%
Alm
ost u
nani
mou
s ag
reem
ent t
hat t
his
shou
ld n
ot b
e al
low
ed to
hap
pen
Mor
e ac
cept
able
if th
e cu
t off
poin
t was
ear
lier t
han
10pm
and
pas
seng
ers
wer
e gi
ven
mor
e w
arni
ng e
.g. 2
4 ho
urs.
“I co
uld
unde
rsta
nd if
they
had
a
plat
form
whe
re th
ey c
ould
info
rm
ever
yone
“thi
s tra
in’s
can
celle
d” a
nd
you'
re g
iven
ple
nty
of n
otic
e bu
t if t
hey
didn
’t re
ally
pro
vide
not
ice,
I’d
stro
ngly
di
sagr
ee th
at th
is is
acc
epta
ble.
” Lo
ndon
gro
up, C
omm
uter
, You
nger
.
“It d
epen
ds w
hat t
ime
the
train
s ar
e th
e
next
day
ent
irely
bec
ause
you
can
get
a
six
o’ c
lock
trai
n an
d yo
u're
sle
epin
g at
10
pm, I
thin
k it
has
to b
e 12
hou
rs’
notic
e or
som
ethi
ng li
ke th
at.”
Gla
sgow
gro
up, C
omm
uter
, You
nger
.
Task
5
Bas
e: 8
0 C
AU
TIO
N: T
hese
figu
res
shou
ld b
e us
ed o
nly
as a
n in
dica
tion,
due
to th
e lo
w s
ampl
e si
ze
48
19%
81%
The
maj
ority
feel
that
all
pass
enge
rs s
houl
d be
trea
ted
equa
lly, r
egar
dles
s of
tim
e of
trav
el
Unf
air t
o le
ave
pass
enge
rs s
trand
ed
And
una
ble
to g
et h
ome
Mos
t pas
seng
ers
can
reca
ll a
time
whe
n th
ey h
ave
been
stra
nded
or a
trai
n ha
s no
t st
oppe
d at
thei
r sta
tion
The
only
exc
eptio
n is
Lon
don
Som
e pa
ssen
gers
feel
that
Lon
don
is b
ette
r equ
ippe
d to
cop
e if
they
can
celle
d tra
in
serv
ices
for a
sho
rt pe
riod
Alte
rnat
ive
rout
es/fo
rms
of tr
ansp
ort a
vaila
ble
‘Dis
aste
r day
s’ s
houl
d be
mea
sure
d se
para
tely
in p
unct
ualit
y st
atis
tics
And
mea
sure
d ag
ains
t how
wel
l the
‘dis
aste
r’ w
as d
ealt
with
.
“I th
ink
it’s
the
mos
t fai
r, ev
eryo
ne’s
goi
ng to
get
de
laye
d an
d th
at’s
just
the
way
it g
oes.
You
all
get
dela
yed
toge
ther
.” Lo
ndon
gro
up, B
usin
ess/
Leis
ure,
Old
er.
Task
6
Bas
e: 8
0 C
AU
TIO
N: T
hese
figu
res
shou
ld b
e us
ed o
nly
as a
n in
dica
tion,
due
to th
e lo
w s
ampl
e si
ze
49
Sum
mar
y
50
Per
form
ance
mea
sure
s an
d tru
st
Pas
seng
ers’
vie
ws
on p
erfo
rman
ce ta
rget
s an
d m
easu
rem
ent a
re in
form
ed b
y th
e w
ider
con
text
with
in
whi
ch th
ey a
sses
s th
e ra
ilway
s. T
here
are
thre
e ke
y is
sues
:
Lim
ited
trus
t - la
ck o
f tru
st is
a re
curr
ing
them
e in
pas
seng
ers’
ass
essm
ent o
f rai
lway
per
form
ance
Th
is s
tudy
, in
line
with
muc
h pr
evio
us re
sear
ch, h
ighl
ight
s a
larg
e de
gree
of c
ynic
ism
abo
ut th
e m
otiv
es a
nd
beha
viou
r of t
he ra
ilway
s W
hile
this
cyn
icis
m c
an o
ften
appe
ar il
l-inf
orm
ed, i
t ref
lect
s bo
th a
(not
unr
easo
nabl
e) la
ck o
f und
erst
andi
ng
abou
t how
the
railw
ay o
pera
tes
and
the
very
lim
ited
emot
iona
l eng
agem
ent t
hat c
onsu
mer
s ha
ve w
ith th
e ra
ilway
s
As
such
, the
re is
a te
nden
cy to
sus
pect
the
wor
st ra
ther
than
giv
e th
e be
nefit
of t
he d
oubt
– a
n im
pres
sion
that
is
rein
forc
ed w
hen
pass
enge
r lea
rn m
ore
abou
t the
det
ail t
hat s
its b
ehin
d P
PM
and
CaS
L
Lack
of c
ompe
titio
n - w
hile
con
sum
ers
are
ofte
n sc
eptic
al a
bout
the
clai
ms
mad
e by
‘big
bus
ines
s’,
this
is c
ompo
unde
d in
the
case
of t
he ra
ilway
s by
a b
elie
f (w
heth
er ri
ght o
r wro
ng) t
hat t
he in
dust
ry is
no
t sub
ject
to th
e sa
me
leve
l of c
ompe
titiv
e m
arke
t dis
cipl
ine
as o
ther
sec
tors
Th
is h
as th
e co
ntra
dict
ory
effe
ct o
f inc
reas
ing
the
perc
eive
d im
porta
nce
of th
e ra
ilway
s ha
ving
stri
ngen
t pe
rform
ance
targ
ets,
whi
le a
lso
rein
forc
ing
pass
enge
rs’ s
uspi
cion
s th
at th
e ra
ilway
s w
ill d
o no
thin
g m
ore
than
the
bare
min
imum
to re
ach
such
targ
ets
Lack
of t
rans
pare
ncy
- for
the
mos
t par
t, pa
ssen
gers
do
not k
now
wha
t tar
gets
are
set
, by
who
m th
ey
are
set o
r how
they
are
mon
itore
d. F
urth
erm
ore,
eve
n w
hen
they
are
info
rmed
abo
ut ta
rget
s, th
ey
stru
ggle
to s
ee a
link
bet
wee
n th
ese
mea
sure
s an
d th
eir o
wn
jour
ney
expe
rienc
e or
how
thes
e m
easu
res
cont
ribut
e to
ser
vice
impr
ovem
ent m
ore
gene
rally
Li
nkin
g pe
rform
ance
mor
e ex
plic
itly
to c
ompe
nsat
ion
and
/ or f
ines
is s
een
by s
ome
as a
bet
ter w
ay o
f ref
lect
ing
pass
enge
rs’ e
xper
ienc
es b
ut s
ome
feel
suc
h co
sts
will
ulti
mat
ely
be b
orne
by
the
pass
enge
r.
51
Pas
seng
ers
valu
e P
PM
/CaS
L in
prin
cipl
e, b
ut q
uest
ion
how
th
ese
mea
sure
s ar
e im
plem
ente
d in
pra
ctic
e P
asse
nger
s’ k
ey p
riorit
y is
for r
elia
bilit
y an
d a
such
ther
e is
bro
ad a
gree
men
t tha
t bot
h pu
nctu
ality
(esp
ecia
lly) b
ut a
lso
canc
ella
tions
sho
uld
be s
tring
ently
mea
sure
d
As
such
, bot
h P
PM
and
CaS
L ar
e se
en a
s re
leva
nt a
nd u
sefu
l in
prin
cipl
e H
owev
er, i
n pr
actic
e, th
e ov
eral
l tar
gets
are
see
n as
rela
tivel
y le
nien
t (pa
rticu
larly
PP
M)
The
curr
ent m
easu
res
are
also
con
side
red
to c
onta
in to
o m
any
‘loop
hole
s’ –
this
rein
forc
es p
erce
ptio
ns o
f len
ienc
y an
d un
derm
ines
trus
t in
the
who
le m
easu
rem
ent p
roce
ss
Pas
seng
ers’
resp
onse
to th
e ab
ove
is to
sug
gest
that
mor
e –
and
mor
e de
taile
d –
info
rmat
ion
shou
ld b
e pr
ovid
ed. S
peci
fical
ly:
Mor
e de
tail
Dep
artu
re/a
rriv
al ti
mes
sho
uld
be m
easu
red
for e
very
trai
n st
oppi
ng a
t eve
ry s
tatio
n E
very
ser
vice
and
jour
ney
mea
sure
d In
form
atio
n on
cau
se o
f del
ay/c
ance
llatio
n/di
srup
tion
Sep
arat
e de
finiti
ons
for ‘
sign
ifica
ntly
late
’ on
shor
t and
long
dis
tanc
e tra
ins
Som
e su
ppor
t wei
ghtin
g by
pas
seng
er n
umbe
rs (a
lthou
gh th
e ov
eral
l con
sens
us w
as fo
r m
easu
rem
ent p
er tr
ain)
Less
leni
ent
Am
ount
of ‘
leew
ay’ d
eter
min
ing
‘on-
time’
bas
ed o
n jo
urne
y le
ngth
of t
rain
TO
Cs
not a
llow
ed to
add
ext
ra ti
me
to th
e tim
etab
le b
etw
een
penu
ltim
ate
and
term
inat
ing
stat
ion
Sco
tRai
l to
be m
easu
red
agai
nst C
aSL
Can
cella
tions
and
‘dis
aste
r day
s’ a
lway
s re
cord
ed a
nd m
easu
red
Tran
spar
ency
M
ore
com
mun
icat
ion
that
thes
e m
easu
res
exis
t M
ore
com
mun
icat
ion
abou
t how
thes
e fig
ures
are
cal
cula
ted
and
mon
itore
d P
rovi
ding
som
e of
the
addi
tiona
l det
ail d
escr
ibed
abo
ve m
ight
hel
p to
reas
sure
pas
seng
ers
abou
t the
rele
vanc
e an
d in
tegr
ity o
f PP
M a
nd C
aSL
How
ever
, it h
as to
be
ackn
owle
dged
that
sug
gest
ions
suc
h as
thes
e ar
e, in
par
t, a
func
tion
of p
asse
nger
s ha
ving
bee
n as
ked
to lo
ok a
t per
form
ance
mea
sure
s in
a s
usta
ined
and
det
aile
d w
ay in
a re
sear
ch s
ettin
g. In
real
ity, c
onsu
mer
s ar
e se
ldom
will
ing
to e
ngag
e at
suc
h a
leve
l of d
etai
l, al
thou
gh th
ey b
elie
ve a
cces
s to
suc
h de
tail
shou
ld b
e av
aila
ble
if re
quire
d.
52
Issu
es to
con
side
r
Pass
enge
rs w
ant t
rain
pun
ctua
lity
and
canc
ella
tions
to b
e m
easu
red
Giv
en th
is, t
here
may
be
mer
it in
pro
vidi
ng a
dditi
onal
det
ail a
bout
targ
ets
and,
inde
ed, i
n m
akin
g ta
rget
s m
ore
strin
gent
But
in a
dditi
on, c
onsi
dera
tion
shou
ld a
lso
be g
iven
to c
omm
unic
atin
g ta
rget
s in
a w
ay th
at h
elps
add
ress
the
bigg
er,
cont
extu
al is
sues
– p
artic
ular
ly th
at o
f tru
st. T
he fo
llow
ing
mig
ht th
eref
ore
be c
onsi
dere
d:
Obj
ectiv
ity a
nd in
depe
nden
ce
Mor
e in
form
atio
n ab
out w
ho s
ets
and
mon
itors
the
targ
ets,
as
wel
l as
reas
sura
nces
that
they
are
man
aged
in
depe
nden
tly a
nd im
parti
ally
with
pas
seng
ers’
bes
t int
eres
ts ‘a
t hea
rt’
Con
sequ
ence
s an
d in
cent
ives
C
omm
unic
atio
n of
why
it is
nec
essa
ry a
nd im
porta
nt fo
r TO
Cs
and
Net
wor
k R
ail t
o m
eet (
or e
ven
exce
ed)
targ
ets
and
the
cons
eque
nces
of n
on-p
erfo
rman
ce
Asp
iratio
nal a
nd p
rogr
essi
ve
Pos
ition
ing
targ
ets
as th
e ‘m
inim
um a
ccep
tabl
e le
vel’
of p
erfo
rman
ce, g
ivin
g an
ass
uran
ce th
at th
e in
dust
ry
striv
es to
exc
eed
thes
e ta
rget
s an
d pr
ovid
ing
evid
ence
that
it d
oes
so o
n oc
casi
ons”
* C
omm
unic
atin
g im
prov
emen
t ove
r tim
e an
d sh
owin
g th
at ta
rget
s w
ork
in s
uppo
rting
con
tinua
l ser
vice
im
prov
emen
t S
impl
icity
C
onsi
dera
tion
mig
ht b
e gi
ven
to re
-fram
ing
targ
ets
away
from
pur
ely
stat
istic
al in
form
atio
n (a
lthou
gh s
till v
ery
muc
h un
derp
inne
d by
‘har
d’ d
ata)
tow
ards
mak
ing
a sm
all n
umbe
r of s
impl
e bu
t def
initi
ve p
rom
ises
, cas
t in
‘con
sum
er’ l
angu
age
and
with
a c
lear
indi
catio
n of
the
cons
eque
nces
of f
ailu
re.
In th
is c
onte
xt, t
arge
ts m
ight
be
bet
ter c
omm
unic
ated
thro
ugh
Cus
tom
er R
epor
ts, w
hich
app
ear t
o ha
ve s
igni
fican
t pot
entia
l for
bui
ldin
g di
alog
ue w
ith p
asse
nger
s*
For e
xam
ple,
‘We
will
onl
y ev
er c
ance
l a tr
ain
for t
he fo
llow
ing
reas
ons.
..’, ‘
If w
e m
ake
you
mor
e th
an X
m
inut
es la
te w
e w
ill re
fund
Y p
er c
ent o
f you
r far
e’ e
tc.
*http
://w
ww
.tran
spor
tfocu
s.or
g.uk
/new
s/ar
ticle
s/tr
ain-
com
pany
-tran
spar
ency
-wel
com
ed-b
y-pa
ssen
gers
-201
5
53
Transport Focus Links between train punctuality and passenger satisfaction: Journeys across the Greater Anglia franchise in 2012, 2013 & 2014
54
Executive summary Transport Focus is the independent transport user watchdog charged with representing the interests of: Britain’s rail passengers; bus, coach and tram passengers in England but outside London; and users of the strategic road network in England. Amongst other objectives, Transport Focus seeks to understand the needs and expectations of rail passengers and to secure tangible and measurable improvements for them. To support these objectives Transport Focus commissions and publishes the twice-yearly National Rail Passenger Survey (NRPS), which is the benchmark measure of changes in passengers’ attitudes towards all elements of rail travel in the country.
This report, commissioned by Transport Focus in a joint project with the Office of Rail and Road (ORR), provides the results of a study examining passenger satisfaction, as measured by NRPS, alongside rail industry train punctuality data from 2012 to 2014. The principal aim of the study is to further understand the relationships between passenger satisfaction with punctuality and actual train punctuality.
The analysis matched train punctuality data with individual respondents from the survey, establishing how late the train was at the location where the passenger alighted. This allows a direct comparison to be made between the lateness the passenger experienced and their resulting satisfaction with punctuality.
The report focusses on the Greater Anglia (GA) franchise in order to provide comparability with a 2009 report prepared by CDL (now part of GHD), as well as to inform Transport Focus ahead of the East Anglia franchise competition later in 2015.
The document has been prepared solely based on data from:
NRPS records and corresponding reports, supplied by Transport Focus; and
Train punctuality and timetable data, supplied by Abellio Greater Anglia (AGA).
In discussion with Transport Focus it was agreed that:
Due to potential complications resulting from weekend engineering works, NRPS respondents travelling at weekends would be excluded from the analysis;
Train punctuality data used would be that for weekdays during the NRPS survey periods only; and
Cancelled services would be excluded from the study.
Findings from the analysis Impact of lateness on satisfaction with punctuality
The analysis shows the relationship between passenger satisfaction with punctuality as measured by the NRPS and actual lateness experienced by the passenger concerned. Do passengers notice an increase in lateness? How significant is the impact on their satisfaction?
Across all passengers surveyed in the study period:
82 per cent express satisfaction with punctuality when their service arrives early or on time; and
For every minute of lateness, satisfaction with punctuality decreases by three percentage points.
The characteristics of a passenger’s journey can influence how they perceive lateness. Typically, those passengers who travel more frequently are less satisfied and are more sensitive to worsening punctuality. The following graphs illustrate this for different journey purposes and
55
travel frequency. Given that these characteristics are clearly linked – commuters travel most frequently – the graphs have similarities.
For journey purpose (as shown in the figure below):
Commuters are less tolerant of lateness than business and leisure travellers; only 74.9 per cent of commuters are satisfied with punctuality when their train is on time;
94 per cent of leisure travellers are satisfied when their train is on time;
For every minute of lateness, commuters’ satisfaction with punctuality declines by five percentage points; this suggests that those passengers who travel most frequently are influenced by previous punctuality experiences when responding to the survey;
Leisure travellers’ satisfaction with punctuality decreases by just one percentage point for every minute of lateness; and
The rate of decline for leisure travellers increases at the eight minute mark, indicating that leisure passengers have heightened awareness of lateness from this point.
For travel frequency (as shown in the figure below):
Passengers travelling more frequently express a lower satisfaction rate than those travelling less frequently even when their service arrives early or on time; their reaction to increased lateness is almost identical to that of commuters; and
Passengers travelling between one and eight times per month respond to lateness in the same way as passengers travelling less than once a month. This indicates that the satisfaction rate of less frequent travellers is defined by their journey purpose rather than their frequency of travel.
0%
20%
40%
60%
80%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15
Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
Changes in satisfaction with punctuality with an increase in minutes of lateness (by journey purpose)
Leisure Commuters Business
56
Extending this to ticket type, the trend continues. Respondents using weekly, monthly and annual season tickets are less satisfied with punctuality, even when on time. When experiencing delays, their satisfaction with punctuality also declines at a faster rate than those on full and reduced tickets (typically bought by less frequent travellers).
Changes to passenger lateness and satisfaction over time
When a comparison between satisfaction and lateness is considered year by year, conclusions may be drawn about whether passengers’ satisfaction with punctuality follows the trend of the punctuality actually experienced. Most notably (as shown in the figure below):
Of the passengers surveyed, the proportion of passengers arriving within five minutes of their scheduled time increased marginally from 89 per cent (2012) to 91 per cent (2014);
The proportion of passengers arriving on time or early decreased in 2013, but increased again in 2014; and
The response of passengers in terms of their satisfaction with punctuality follows a similar trend, with a dip seen in 2013, followed by a recovery in 2014.
Note that the graph below is only representative of the respondents who were successfully mapped to punctuality data.
Impact of lateness on overall satisfaction
Many years of NRPS results has enabled Transport Focus to determine that satisfaction with punctuality is the key driver for overall satisfaction. As a result, it is expected that the trend
0%
20%
40%
60%
80%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
Changes in satisfaction with punctuality with an increase in minutes of lateness (by frequency of travel)
Three or more times a week One to eight times per month Less than once a month
89% 88% 91%
63%56% 60%
73% 72% 75%
0%
20%
40%
60%
80%
100%
2012 2013 2014
Prop
ortio
n of
pas
seng
ers
Year
NRPS matched passenger lateness vs passenger satisfaction with punctuality
Proportion of passengers arriving within five minutes of scheduled time (RT5)Proportion of passengers arriving early or on time (RT)Satisfaction with punctuality
57
between satisfaction and increased lateness to apply to overall satisfaction as well. While other factors clearly influence a passenger’s overall satisfaction, the graph below shows that lateness does play a key part. The key findings are (as shown in the figure below):
Only 83 per cent expressed overall satisfaction with their service when arriving early or on time; and
For each minute of lateness, the overall satisfaction reduces by 1.5 percentage points.
This does not prove a relationship, but indicates that passenger’s overall responses are influenced by punctuality.
This is supported when considering how passengers’ satisfaction with punctuality and overall satisfaction has changed over the three year study period. Both satisfaction with punctuality and overall satisfaction in 2014 improved in comparison with 2012, as shown in the figure below. Note that the data for this graph is taken from the respective NRPS reports and represents the full survey, not just the matched data.
Impact of satisfaction with punctuality on responses to other NRPS questions
The analysis carried out also considers the impact that passenger satisfaction with punctuality has on responses to other NRPS questions, e.g. satisfaction with train cleanliness and crowding. There are three categories of relationship:
Satisfaction drives satisfaction (satisfaction with punctuality drives positive responses to other questions);
Dissatisfaction drives dissatisfaction (dissatisfaction with punctuality drives negative responses to other questions); and
0%10%20%30%40%50%60%70%80%90%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15
Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
Changes in overall satisfaction with an increase in minutes of lateness
74%78% 80%
70%75% 76%
0%
20%
40%
60%
80%
100%
2012 2013 2014
Prop
ortio
n of
pas
seng
ers
satis
fied
Year
NRPS respondents satisfaction with punctuality and overall satisfaction
Overall satisfaction Satisfaction with punctuality
58
No significant relationship.
The table below allocates NRPS questions to each category, with the strongest relationship at the top of each column. For example, satisfaction with punctuality drives satisfaction with scheduled journey time in a positive direction. Conversely, dissatisfaction with punctuality means passengers are more likely to be dissatisfied with the value for money of their ticket. Note that the relationship with overall satisfaction, discussed earlier, has also been included in this table. The analysis suggests that satisfaction with punctuality drives overall satisfaction, but dissatisfaction with punctuality does not have the same impact on overall dissatisfaction.
Satisfaction with punctuality drives satisfaction
Dissatisfaction with punctuality drives dissatisfaction
No significant relationship
Overall satisfaction
Scheduled journey time
Value for money of ticket
Sufficient space to sit and stand
Ease of getting on and off the train
Train cleanliness
Comparing train and passenger lateness
The analysis reported up to this point is based on the lateness the passenger experiences. It is important to understand that this can be significantly different from the train lateness that is widely reported by the rail industry in the form of the Public Performance Measure (PPM) and ’right-time’ arrivals. The graph below highlights this by considering only services travelling away from London. The reason these services have been chosen is because the destination of the passenger and of the train is likely to be different. The graph shows there is a one minute lag between the two lines (as shown in the figure below). This means:
If a train from London is reported as being on time, the average passenger will be one minute late; and
On services departing London, 69 per cent of trains arrive at their ultimate destinations on time or early, compared with 56 per cent of passengers arriving at their station on time or early.
0%
5%
10%
15%
20%
25%
30%
-5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 101112131415161718192021222324252627282930
Prop
ortio
n of
pas
seng
ers/
trai
n
Lateness (minutes)
Lateness of train at passenger destination vs train destination (departures from London)
Passenger Train
59
Passenger satisfaction insight
During the analysis, some useful insight was also gained into how passenger satisfaction varies by passenger characteristic.
Satisfaction by gender: In general female respondents show higher levels of satisfaction – both with punctuality and overall – than male respondents; the difference ranging from two to seven percentage points over the past three years.
Satisfaction by age: Older age groups are considerably more satisfied than younger age groups (as shown below).
Satisfaction by route: All routes have generally been improving with respect to satisfaction with punctuality. None more so than Stansted Express which has satisfaction levels 13 percentage points higher than any other route in 2014, although this may be related to changes in the composition of the Stansted Express building block within NRPS.
0%10%20%30%40%50%60%70%80%90%
100%
16-25 26-34 35-44 45-54 55-64 65+
Prop
ortio
n of
pas
seng
ers
satis
fied
Age groups
Satisfaction with punctuality by age
2012 2013 2014
60
Table of contents 1. Introduction................................................................................................................................
1.1 Background .....................................................................................................................
1.2 Report structure ..............................................................................................................
2. Approach ...................................................................................................................................
2.1 Source data .....................................................................................................................
2.2 NRPS route definitions....................................................................................................
2.3 Mapping NRPS respondents to delay data ....................................................................
3. Passenger satisfaction observations ........................................................................................
3.1 Application of weighting – gender and age .....................................................................
3.2 Satisfaction by gender and age ......................................................................................
3.3 Satisfaction by time of day ..............................................................................................
3.4 Satisfaction by route .......................................................................................................
3.5 Satisfaction by journey purpose ......................................................................................
4. Influence of satisfaction with punctuality ...................................................................................
4.1 Influence of punctuality on overall satisfaction ...............................................................
4.2 Influence of punctuality on other NRPS criteria ..............................................................
5. Train punctuality observations ..................................................................................................
5.1 Overview .........................................................................................................................
6. Passenger punctuality analysis .................................................................................................
6.1 Overview .........................................................................................................................
6.2 Punctuality by time of day ...............................................................................................
6.3 Punctuality by route ........................................................................................................
6.4 Passenger punctuality vs train punctuality .....................................................................
7. Relationship between satisfaction with punctuality and actual punctuality ...............................
7.1 Satisfaction with punctuality by lateness ........................................................................
8. Relationship between overall satisfaction and actual punctuality .............................................
8.1 Overall satisfaction by length of delay and gender .........................................................
8.2 Overall satisfaction by length of delay and journey purpose ..........................................
9. Impact of remapping .................................................................................................................
Appendix 1 – Passenger punctuality by time of day and route ...........................................................
63
63
63
64
64
65
67
68
68
69
70
71
72
74
74
77
81
81
83
83
83
84
85
87
87
96
96
97
98
99
61
Table index Table 1 Date ranges and sample size for NRPS datasets ..................................................................
Table 2 Sample size of Bugle data......................................................................................................
Table 3 GA route categorisation and definition ...................................................................................
Table 4 Adjustments made on origin/destination stations for WA remapping ....................................
Table 5 Combined NRPS and Bugle sample size for each year ........................................................
Table 6 Influence of NRPS factors on overall satisfaction ..................................................................
Table 7 Ticket type groupings .............................................................................................................
Figure index Figure 1 NRPS and Bugle data matching process ..............................................................................
Figure 2 Unweighted NRPS respondent proportion based on gender and age .................................
Figure 3 Weighted NRPS respondent proportion based on gender and age .....................................
Figure 4 Overall satisfaction by gender ...............................................................................................
Figure 5 Satisfaction with punctuality by gender .................................................................................
Figure 6 Overall satisfaction by age ....................................................................................................
Figure 7 Satisfaction with punctuality by age ......................................................................................
Figure 8 Satisfaction with punctuality by time of day ..........................................................................
Figure 9 Weighted respondent proportions by route ...........................................................................
Figure 10 Overall satisfaction by route ................................................................................................
Figure 11 Satisfaction with punctuality by route ..................................................................................
Figure 12 Overall satisfaction by journey purpose ..............................................................................
Figure 13 Satisfaction with punctuality by journey purpose ................................................................
Figure 14 Overall satisfaction and satisfaction with punctuality ..........................................................
Figure 15 Approach to determining influence of satisfaction with punctuality ....................................
Figure 16 Example of a matrix that indicates no relationship between satisfaction with punctuality and other NRPS criteria ................................................................................
Figure 17 Example of a matrix that indicates a strong relationship between satisfaction with punctuality and other NRPS criteria ................................................................................
Figure 18 Comparison of satisfaction with punctuality and overall satisfaction ..................................
Figure 19 Comparison of satisfaction with punctuality and journey time ............................................
Figure 20 Comparison of satisfaction with punctuality and value for money ......................................
Figure 21 Comparison of satisfaction with punctuality and crowding .................................................
Figure 22 Comparison of satisfaction with punctuality and ease of being able to get on and off the train ......................................................................................................................
64
65
65
66
67
74
93
67
68
68
69
69
70
70
71
71
72
72
73
73
75
75
76
77
77
78
79
79
80
62
Figure 23 Comparison of satisfaction with punctuality and train cleanliness ......................................
Figure 24 Train punctuality by year .....................................................................................................
Figure 25 Trains arriving early or on time by departure hour ..............................................................
Figure 26 Trains arriving early or on time by route .............................................................................
Figure 27 Passenger punctuality by year ............................................................................................
Figure 28 Passengers arriving early or on time by departure hour .....................................................
Figure 29 Passengers arriving early or on time by route ....................................................................
Figure 30 Distribution of passenger lateness by route ........................................................................
Figure 31 Train lateness vs passenger lateness .................................................................................
Figure 32 Train lateness vs passenger lateness for services departing London ................................
Figure 33 Comparison of satisfaction with punctuality and actual passenger punctuality ..................
Figure 34 Satisfaction with punctuality vs passenger lateness ...........................................................
Figure 35 Satisfaction with punctuality vs passenger lateness by gender ..........................................
Figure 36 Satisfaction with punctuality vs passenger lateness by age (1) .........................................
Figure 37 Satisfaction with punctuality vs passenger lateness by age (2) .........................................
Figure 38 Satisfaction with punctuality vs passenger lateness by journey purpose ...........................
Figure 39 Satisfaction with punctuality vs passenger lateness by route (1) .......................................
Figure 40 Satisfaction with punctuality vs passenger lateness by route (2) .......................................
Figure 41 Satisfaction with punctuality vs passenger lateness by ticket type (1) ...............................
Figure 42 Satisfaction with punctuality vs passenger lateness by ticket type (2) ...............................
Figure 43 Satisfaction with punctuality vs passenger lateness by frequency of travel .......................
Figure 44 Overall satisfaction vs passenger lateness .........................................................................
Figure 45 Overall satisfaction vs passenger lateness by gender ........................................................
Figure 46 Overall satisfaction vs passenger lateness by journey purpose .........................................
Figure 47 Overall satisfaction - route remapping comparison ............................................................
Figure 48 Satisfaction with punctuality - route remapping comparison ..............................................
Figure 49 Passengers arriving early or on time by hour and route (1) ................................................
Figure 50 Passengers arriving early or on time by hour and route (2) ................................................
80
81
82
82
83
84
84
85
85
86
87
88
88
89
90
90
92
92
94
94
95
96
97
97
98
98
99
99
63
1. Introduction 1.1 Background
Transport Focus is the independent transport user watchdog charged with representing the interests of Britain’s rail passengers; bus, coach and tram passengers in England but outside London; and users of the strategic road network in England. Amongst other objectives, Transport Focus seeks to understand the needs and expectations of rail passengers and to secure tangible and measurable improvements for them. To support these objectives, Transport Focus commissions and publishes the twice-yearly National Rail Passenger Survey (NRPS), which is the benchmark measure of changes in passengers’ attitudes towards all elements of rail travel in the country.
Evidence from a wide range of research, including that of Transport Focus, has highlighted that punctuality and reliability of train services is one of the key measures of each Train Operating Company’s (TOC’s) NRPS scores. As a result, Transport Focus commissioned this study to further examine the relationship between NRPS results and the punctuality of train services.
The report focusses on the Greater Anglia (GA) franchise in order to provide comparability with a 2009 report prepared by CDL (now part of GHD), as well as to inform Transport Focus ahead of the East Anglia franchise competition later in 2015. The GA franchise is currently operated by Abellio Greater Anglia (AGA).
1.2 Report structure
This report provides insight into satisfaction, train punctuality and passenger punctuality before bringing the elements together. Section 2 details the approach and assumptions made during the data analysis.
The initial focus of the report is on satisfaction among passengers using the GA franchise, with Section 3 presenting NRPS results for satisfaction with punctuality and overall satisfaction. Section 4 considers how these two measures of satisfaction are related, before considering how satisfaction with punctuality influences other NRPS criteria.
The focus then turns to lateness, with Section 5 analysing the punctuality of terminating AGA trains (restricted to those within the NRPS survey dates). Section 6 considers the lateness of services at the passenger’s destination, which is often different from that of the train. Train and passenger punctuality are directly compared in this section.
Section 7 brings Sections 3, 5 and 6 together by mapping NRPS respondents against train punctuality data to analyse how passengers’ satisfaction with punctuality varies as lateness increases. Section 8 repeats the same process, but for overall satisfaction.
Finally, Section 9 gives a summary of the impact of remapping services to London Overground and Crossrail, to enable Transport Focus to understand how passenger satisfaction may change after remapping.
64
2. Approach 2.1 Source data
For this study, three main sources of data have been used, each of which are explained in more detail in the following sections. These are:
NRPS records1;
Bugle data (train punctuality data, explained in Section 2.1.2)2; and
MOIRA timetable data (explained in Section 2.1.3)3.
The datasets used in this study have been obtained either from publically-available sources or directly from Transport Focus or AGA.
2.1.1 NRPS records
NRPS survey results contain the individual responses from each passenger surveyed, including details of the train service they travelled on. From this dataset it is possible to determine the purpose of their journey, the frequency of travel, personal characteristics, as well as scores for satisfaction of a number of attributes relating to their journey. It is important to emphasise that the satisfaction is intended to relate to the single journey they have undertaken and not be influenced by previous experiences.
The NRPS records date back to 1999 and are published on a twice-yearly basis; each survey period is known as a ‘wave’. In order to establish a suitable sample, the analysis takes into account the NRPS results from the last six waves, covering 2012, 2013 and 2014. Table 1 summarises details of each wave.
Year Wave NRPS Survey Start date
NRPS Survey End date
NRPS sample size
Proportion of Weekday Responses
Spring 2012 26 28-Jan-12 30-Mar-12 2,454 86 per cent
Autumn 2012 27 01-Sep-12 12-Nov-12 2,156 87 per cent
Spring 2013 28 12-Jan-13 24-Mar-13 2,267 87 per cent
Autumn 2013 29 02-Sep-13 11-Nov-13 2,226 86 per cent
Spring 2014 30 02-Feb-14 13-Apr-14 2,313 86 per cent
Autumn 2014 31 02-Sep-14 11-Nov-14 2,226 91 per cent
Table 1 Date ranges and sample size for NRPS datasets
Due to potential complications resulting from weekend engineering works, NRPS respondents travelling at weekends would be excluded from the analysis.
Weighting of survey results
In order to obtain a statistically robust assessment of passenger satisfaction, the NRPS survey obtains results from the whole GA franchise, including different sizes of stations. The results are also obtained at different times of day, and day of week so that the views of a mix of commuters, business and leisure travellers are represented within the published results.
In order that the NRPS results correctly represent the views of all passengers travelling, the individual responses are given a weighting to reflect the significance of the individual response.
1 Received from Transport Focus on 5th February 2015 2 Received from AGA on 31st March 2015 3 Received from AGA on 3rd March 2015
65
These weightings are used throughout the analysis, except where stated otherwise. A demonstration of the impact of weighting is given in Section 3.1.
2.1.2 Bugle data
DATASYS BUGLE is a rail performance management software tool used widely by the UK rail industry. TOCs use this software to monitor, manage and improve train performance by capturing and analysing train delays and lateness data. The data produced by this software is widely known as Bugle data.
AGA provided the Bugle data for days specific to the NRPS survey. The lateness at every stop for every train service run during these periods has been captured. All cancelled services have been excluded from the Bugle dataset, as they are difficult to quantify in the context of this study. The sample size for each of the waves is as detailed in Table 2.
Period Bugle sample size (number of stops)
Spring 2012 620,837
Autumn 2012 737,410
Spring 2013 726,516
Autumn 2013 741,760
Spring 2014 728,879
Autumn 2014 745,060
Table 2 Sample size of Bugle data
2.1.3 MOIRA timetable data
MOIRA is a software system designed to predict how changes to timetables will affect passenger revenue. For this study, however, the model is used for the source data contained within it, rather than its ability as a revenue forecasting model. The MOIRA data has been used to obtain the specific timetable information on all train services operated by AGA during the sample date ranges described above. This was supplied by AGA.
2.2 NRPS route definitions
The AGA sample for NRPS is organised into routes (or ‘building blocks’) so that variation in satisfaction within the franchise can be understood. Table 3, below, identifies the routes within AGA that were used as part of the analysis, as defined in Transport Focus’s rail passenger satisfaction ‘at a glance’ report.
GA Routes Definition
Intercity London to Norwich journeys, plus a few shorter workings
Mainline Journeys on outer suburban Great Eastern services including London to Ipswich, plus branches to Harwich, Clacton, Walton, Sudbury, Southminster and Braintree. Also includes journeys on London to Southend Victoria service and London Liverpool Street to Chelmsford and Colchester
Metro Journeys on London Liverpool Street to Ilford, Gidea Park and Shenfield
West Anglia Outer Journeys on West Anglia routes London - Hertford East, London to Cambridge, London to King's Lynn, Cambridge to
66
King's Lynn, Cambridge, Bishop’s Stortford and Hertford East via Tottenham Hale
Stansted Express Journeys on Stansted Express, on Greater Anglia trains which start or end at Stansted Airport, where the passenger has an origin or destination of the airport4
West Anglia Inner Journeys on West Anglia routes London to Enfield Town, London to Chingford, London to Cheshunt and Romford to Upminster.
Rural Journeys on Ipswich to Felixstowe, Lowestoft, Cambridge and Peterborough rail lines, plus Norwich to Lowestoft, Yarmouth, Sheringham and Cambridge lines
Table 3 GA route categorisation and definition
Due to the proposed remapping of West Anglia (WA) services to London Overground and Great Eastern Metro services to Crossrail and London Overground, the latest round of NRPS results (wave 31) has undergone some remapping. This has resulted in some changes to routes allocated to previous NRPS responses as detailed below.
2.2.1 Creation of West Anglia Inner for remapping to London Overground
Historically, within NRPS, West Anglia has been a route on its own. In wave 31 (Autumn 2014), it has been split into West Anglia Inner and West Anglia Outer, as shown in Table 3. This enables the West Anglia Inner sample to be remapped to London Overground in 2015. For this analysis, to enable consistency, all respondents from previous waves have been allocated to Inner and Outer as appropriate.
West Anglia Outer origin/destination West Anglia Inner origin/destination
King’s Lynn Cheshunt
Ely Chingford
Cambridge Enfield Town
Bishop’s Stortford London Fields
Broxbourne
Hertford East
Tottenham Hale
Table 4 Adjustments made on origin/destination stations for WA remapping
2.2.2 Redefining Metro for remapping to Crossrail and London Overground
Historically, the Metro building block has contained Liverpool Street to Shenfield, Liverpool Street to Southend Victoria, Wickford to Southminster and Romford to Upminster trains. In wave 31 (Autumn 2014), the Southend and Southminster trains have been reallocated to the Mainline building block and Romford to Upminster trains to the West Anglia Inner building block. All
4 The composition of the Stansted Express building block has changed in Wave 31 (Autumn 31). Prior to this, the passenger journey was not taken into account, only the origin and destination of the train. No re-categorisation has taken place – in this study – to reflect this change.
67
respondents in previous waves have been re-categorised to reflect these changes. This enables the Metro route to be remapped to Crossrail in 2015, and the Romford to Upminster services to be remapped to London Overground with the West Anglia Inner services.
2.3 Mapping NRPS respondents to delay data
In order to understand how passengers perceive punctuality, it is necessary to map NRPS respondents to Bugle lateness information. Figure 1 outlines the process undertaken as part of this study.
Figure 1 NRPS and Bugle data matching process
During the mapping process, every NRPS respondent’s journey is matched to a MOIRA service from the respective timetable. As part of a separate process, Bugle services are also mapped to MOIRA services. The MOIRA timetables bring the two datasets together, identifying the lateness that each individual passenger experienced on their journey. The outcome from this matching process is the minutes of lateness – or delay – on arrival at the stations where passengers alight.
Matching the NRPS responses with Bugle data through MOIRA timetables maximises the proportion of successful matches, resulting in a success rate of 94 per cent of 2012 NRPS respondents, 94 per cent of 2013 NRPS respondents and 92 per cent of 2014 NRPS respondents. This results in the sample size, as shown in Table 5, for each of the three years, totalling 10,849 respondents. Note that the six waves have been combined into three years, rather than considered individually, because of the common trend for autumn survey results to be more positive than spring survey results.
Year Sample size
2012 3,714
2013 3,561
2014 3,574
Table 5 Combined NRPS and Bugle sample size for each year
68
3. Passenger satisfaction observations This section gives an insight into passengers’ responses with respect to satisfaction with punctuality and overall satisfaction.
3.1 Application of weighting – gender and age
Figure 2 outlines the proportion of NRPS survey respondents based on gender and age for the last three years. This is simply based on the number of respondents in each category.
Figure 2 Unweighted NRPS respondent proportion based on gender and age
To illustrate the explanation in Section 2.1.1, if the weighting for each response in NRPS is used, the proportions change slightly. Figure 3 shows the weighted proportion of NRPS respondents based on gender and age, and is more reflective of AGA’s passengers. This illustrates that Transport Focus had surveyed too few male passengers, and thus applied a higher weighting to the responses received from males; likewise for 26-34, 35-44, 45-54 and 65-69 year olds.
Figure 3 Weighted NRPS respondent proportion based on gender and age
The unweighted results show that 53.7 per cent of survey respondents for the past six waves are female. 23.9 per cent of respondents are between the age of 45 and 54 with respondents over the age of 81 forming less than one per cent of survey respondents.
46.3%
53.7%
Male Female
0%
5%
10%
15%
20%
25%
30%
Age groups
49%51%
Male Female
0%
5%
10%
15%
20%
25%
30%
Age groups
69
3.2 Satisfaction by gender and age
This section outlines overall satisfaction and satisfaction with punctuality based on the NRPS respondents’ gender and age. In terms of both overall satisfaction (Figure 4) and satisfaction with punctuality (Figure 5), female respondents are more satisfied with punctuality than male respondents. Between 2012 and 2014, overall satisfaction among female respondents remained static whereas male respondents became more satisfied, reducing the gap between the two.
Figure 4 Overall satisfaction by gender
The difference between male and female satisfaction with punctuality is less pronounced than for overall satisfaction; the gap reduced to just two percentage points in 2014.
Figure 5 Satisfaction with punctuality by gender
Figure 6 outlines the variations in overall satisfaction by different age categories. For the three years of study, respondents in age categories 26-34, 35-44 and 45-54 are the least satisfied age groups. Respondents in age categories 65+ are generally more satisfied in terms of overall satisfaction.
83% 80%83%
72%75% 76%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2012 2013 2014
Prop
ortio
n of
pas
seng
ers
satis
fied
Year
Female Male
79%76% 77%
72% 73% 75%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2012 2013 2014
Prop
ortio
n of
pas
seng
ers
satis
fied
Year
Female Male
70
Figure 6 Overall satisfaction by age
Figure 7 shows how satisfaction with punctuality varies according to the defined age group categories. The trend of increasing satisfaction with age is more pronounced for satisfaction with punctuality with each age range taking a distinct step up. The only exception is in 2014, where 16-25 year olds are more satisfied than 26-34 year olds. This does not impact significantly on the observation.
Figure 7 Satisfaction with punctuality by age
3.3 Satisfaction by time of day
This section considers whether time of travel has an impact on satisfaction with punctuality. The results displayed in Figure 85 are based on NRPS respondents’ journey departure time. Using departure time makes definition of ‘peak’ more difficult; AM peak is usually defined by the time of arrival at the train’s destination. The term ‘peak’ in this report refers to passengers departing between 0700 and 1000 and between 1600 and 1900.
Figure 8 shows there is a notable decline in satisfaction during the peak hours. Section 3.5 discusses the difference in satisfaction between commuters and other travellers, stating that commuters are generally less satisfied, which could be the reason for the decline. In order to provide clarity, the train and passenger punctuality needs to be considered. Sections 5.1.1 and 6.2, respectively, look at this.
5 Due to significantly lower number of respondents using services before 0600 and after 2000, the outcome displayed is limited to hours between 0600 and 2000.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
16-25 26-34 35-44 45-54 55-64 65+
Prop
ortio
n of
pas
seng
ers
satis
fied
Age groups
2012 2013 2014
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
16-25 26-34 35-44 45-54 55-64 65+
Prop
ortio
n of
pas
seng
ers
satis
fied
Age groups
2012 2013 2014
71
Figure 8 Satisfaction with punctuality by time of day
3.4 Satisfaction by route
In order to better understand the franchise, the NRPS data were categorised according to the routes described in Table 3. For context, Figure 9 shows the distribution of NRPS respondents based on the routes they travel on; over 60 per cent of respondents travel on GE routes. To understand the extent of the remapping, 29 per cent of demand is to be transferred to either London Overground or Crossrail in 2015.
Figure 9 Weighted respondent proportions by route
Figure 10 shows that, in comparison with 2012, overall satisfaction in 2014 increased on all routes, with the exception of Rural. Respondents using Stansted Express services have the highest level of overall satisfaction. This may, however, relate to changes in composition of the Stansted Express building block within NRPS.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
06 07 08 09 10 11 12 13 14 15 16 17 18 19 20
Prop
ortio
n of
pas
seng
ers
satis
fied
Departure hour
2012 2013 2014
9.2%
36.4%
4.8%
16.6%
14.5%
3.7%
14.7%
Intercity
Mainline
Rural
West Anglia Outer
West Anglia Inner
Stansted Express
Metro
72
Figure 10 Overall satisfaction by route
Similarly, Figure 11 displays how satisfaction with punctuality across the GA routes changed over the past three years. The results show that respondents using Stansted Express, Mainline, West Anglia Outer, Rural and Metro are more satisfied in 2014 than in 2012. Notably, Stansted Express has seen the greatest rate of improvement resulting in a level of satisfaction some 13 percentage points higher than any other routes.
Figure 11 Satisfaction with punctuality by route
3.5 Satisfaction by journey purpose
NRPS respondents can select commute, business or leisure as their journey purpose. The lowest overall satisfaction of the three is recorded by commuters with an average of just 73 per cent satisfied across the last three years, compared with 89 per cent satisfaction amongst leisure travellers. Figure 12 illustrates this, showing a higher rate of overall satisfaction in 2014 than in 2012 across all journey purposes.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Intercity StanstedExpress
Mainline West AngliaOuter
West AngliaInner
Metro Rural
Prop
ortio
n of
pas
seng
ers
satis
fied
Routes
2012 2013 2014
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Intercity StanstedExpress
Mainline West AngliaOuter
West AngliaInner
Metro Rural
Prop
ortio
n of
pas
seng
ers
satis
fied
Routes
2012 2013 2014
73
Figure 12 Overall satisfaction by journey purpose
A repeat of the above analysis for satisfaction with punctuality shows that the commuter impact is more pronounced, with 68 per cent of commuters satisfied in comparison with 90 per cent of leisure travellers. When considering satisfaction by time of day in parallel (Figure 8), there is a clear link between poor peak satisfaction and poor commuter satisfaction. The cause of this will become clearer when considering the punctuality data in latter sections.
Figure 13 Satisfaction with punctuality by journey purpose
0%
20%
40%
60%
80%
100%
2012 2013 2014
Prop
ortio
n of
pas
seng
ers
satis
fied
Year
Business Leisure Commute
0%
20%
40%
60%
80%
100%
2012 2013 2014
Prop
ortio
n of
pas
seng
ers
satis
fied
Year
Business Leisure Commute
74
4. Influence of satisfaction with punctuality As part of each NRPS publication, Transport Focus publishes a supporting document6 known as the Multivariate Analysis Report. It outlines the factors that are the biggest drivers of overall satisfaction. The results are summarised in Table 6.
4.1 Influence of punctuality on overall satisfaction
Over the history of the NRPS, the results show that passengers are most likely to be satisfied with their journey if they are satisfied with punctuality/reliability. However, the statistics show a reduction in strength of the relationship between punctuality and overall satisfaction for the Greater Anglia franchise. In Spring 2012, punctuality/reliability contributed 43 per cent as a driver of overall satisfaction. This declined to 31 per cent in Spring 2014. Other factors, such as cleanliness of the inside of the train, are having an increasing impact on overall satisfaction. This is common on routes where punctuality has improved.
Multivariate Analysis Factors
Spring 2012 (Wave 26)
Autumn 2012 (Wave 27)
Spring 2013 (Wave 28)
Autumn 2013 (Wave 29)
Spring 2014 (Wave 30)
Autumn 2014 (Wave 31)
Punctuality/reliability (i.e. the train arriving/departing on time)
42.8% 36.4% 34.2% 36.7% 31.0% 26.0%
The cleanliness of the inside of the train
11.5% 13.2% 19.3% 19.3% 23.1% 21.4%
The length of time the journey was scheduled to take (speed)
6.3% 11.8% 10.7% 11.2% 10.6% 9.2%
The ease of being able to get on and off the train
4.3% 1.8% 5.2% 3.5% 6.8% 12.3%
Sufficient room for all the passengers to sit/stand
4.7% 3.0% 3.2% 3.8% 3.8% 6.3%
Other (5 per cent or less per factor (based on wave 31 results)
30.4% 48.0% 27.4% 25.5% 24.7% 24.8%
Total Variance 100% 100% 100% 100% 100% 100%
Table 6 Influence of NRPS factors on overall satisfaction
Nevertheless, it is clear that satisfaction with punctuality is still the main driver. Therefore a strong relationship between overall satisfaction and satisfaction with punctuality is still expected to be evident in this study.
Figure 14 compares overall satisfaction with satisfaction with punctuality across AGA. As expected, with increased satisfaction with punctuality, overall satisfaction also increases. This does not prove a relationship, but suggests the presence of one. Note that the values in this graph are sourced from the NRPS report, representing the full sample of passengers surveyed.
6 Downloaded from http://www.transportfocus.org.uk/research/rail-research
75
Figure 14 Overall satisfaction and satisfaction with punctuality
To further investigate this relationship, and to draw a conclusion regarding the relationship between satisfaction with punctuality and other NRPS criteria (Section 4.2) the following approach has been applied to compare the responses to two different NRPS questions.
In responding to NRPS survey questions, respondents have five options available to them (excluding the option to ignore the question), namely:
Very satisfied;
Fairly satisfied;
Neither satisfied nor dissatisfied;
Fairly dissatisfied; and
Very dissatisfied.
Figure 15 illustrates that each response option is assessed in isolation. This means, for all passengers who answered that they are very satisfied with punctuality, the proportion of these passengers who are very satisfied with the second criteria is calculated, along with the proportion who are fairly satisfied and so on.
Figure 15 Approach to determining influence of satisfaction with punctuality
The outcome of the analysis is displayed in matrices using a coloured format where any relationship between punctuality and the selected NRPS criteria is defined based on the following.
74% 78% 80%70%
75% 76%
0%10%20%30%40%50%60%70%80%90%
100%
2012 2013 2014
Prop
ortio
n of
pas
seng
ers
satis
fied
Year
Overall satisfaction Satisfaction with punctuality
76
As an example, Figure 16 is an illustration of a scenario where there is no relationship between satisfaction with punctuality and the other criteria. Note, in this and all matrices, each column adds up to 100 per cent, emphasising that the group of respondents who answered ‘very satisfied’ with punctuality are considered in isolation from those who answered ‘fairly satisfied’ with punctuality. In this example, the split of satisfaction with the second NRPS criteria is identical regardless of the response to satisfaction with punctuality. As a result, regardless of whether a passenger is very satisfied or very dissatisfied with punctuality, they have a 45 per cent chance of being very satisfied with the second criteria. This results in the colours being consistent row by row.
Figure 16 Example of a matrix that indicates no relationship between satisfaction with punctuality and other NRPS criteria
Figure 17, conversely, illustrates an example of a strong relationship. In this example, if a passenger is very satisfied with punctuality, they are more likely to be very satisfied with the other criteria. As a result, if a passenger is very satisfied with punctuality, there is a 75 per cent chance they are very satisfied with the second criteria. However, if they are very dissatisfied with punctuality, there is zero per cent chance of them being very satisfied with the second criteria. This results in a green diagonal across the matrix.
Very Weak Very StrongStrongNeutralWeak
Very Dissatisfied
Fairly Dissatisfied
NeitherFairly
SatisfiedVery
Satisfied
1 2 3 4 5
Ver
y D
issa
tisfie
d
1 5% 5% 5% 5% 5%
Fai
rly
Dis
satis
fied
2 10% 10% 10% 10% 10%
Nei
ther
3 15% 15% 15% 15% 15%
Fai
rly
Sat
isfie
d
4 25% 25% 25% 25% 25%
Ver
y S
atis
fied
5 45% 45% 45% 45% 45%
Satisfaction with Punctuality
NR
PS
Crit
eria
77
Figure 17 Example of a matrix that indicates a strong relationship between satisfaction with punctuality and other NRPS criteria
Before considering, other NRPS criteria, this approach is applied to Overall Satisfaction to further understand the relationship. Figure 18 provides a comparison between the responses to satisfaction with punctuality and overall satisfaction in the AGA NRPS results for 2012, 2013 and 2014. While the green diagonal is not as obvious as the example above, it is still present, particularly relating to the higher levels of satisfaction. It shows that when passengers are satisfied with punctuality they are more likely to be satisfied overall. However, at low levels of satisfaction with punctuality, the overall satisfaction is spread. This indicates that satisfaction with punctuality drives overall satisfaction, strengthening the view that the two are closely linked. The same cannot be said for dissatisfaction; a passenger dissatisfied with punctuality is not necessarily dissatisfied overall.
Figure 18 Comparison of satisfaction with punctuality and overall satisfaction
4.2 Influence of punctuality on other NRPS criteria
As demonstrated in Table 6, passengers’ satisfaction with punctuality influences their satisfaction with other factors measured by NRPS. This has been tested through the production of a series of the matrices described in Section 4.1 looking at the relationship with:
Scheduled journey time;
Very Dissatisfied
Fairly Dissatisfied
NeitherFairly
SatisfiedVery
Satisfied
1 2 3 4 5
Ver
y D
issa
tisfie
d
1 75% 15% 8% 2% 0%
Fai
rly
Dis
satis
fied
2 15% 60% 15% 8% 2%
Nei
ther
3 8% 15% 54% 15% 8%
Fai
rly
Sat
isfie
d
4 2% 8% 15% 60% 15%
Ver
y S
atis
fied
5 0% 2% 8% 15% 75%
NR
PS
Crit
eria
Satisfaction with Punctuality
Very Dissatisfied
Fairly Dissatisfied
NeitherFairly
SatisfiedVery Satisfied
1 2 3 4 5
Ver
y D
issa
tisfie
d
1 19% 3% 1% 1% 0%
Fai
rly
Dis
satis
fied
2 32% 19% 8% 3% 1%
Nei
ther
3 24% 31% 34% 13% 3%
Fai
rly
Sat
isfie
d
4 25% 42% 50% 68% 39%
Ver
y S
atis
fied
5 2% 4% 6% 16% 57%
Satisfaction with Punctuality
Ove
rall
Sat
isfa
ctio
n
78
Value for money;
Sufficient space to sit and stand;
Ease of being able to get on and off the train; and
Cleanliness of the inside of the train.
4.2.1 Influence of punctuality on satisfaction with scheduled journey time
Figure 19 considers how passenger satisfaction with punctuality impacts on satisfaction with the scheduled journey time. The green diagonal axis indicates a reasonable correlation between the two. This says either:
a. Passengers cannot distinguish between scheduled journey time and delay; or
b. Passenger’s view of scheduled journey time is influenced by their views of punctuality.
Given that the highest values are skewed towards the right hand corner of the matrix, it indicates that, for this measure, satisfaction drives satisfaction. As a result, if a passenger is satisfied with punctuality they are also more likely to be satisfied with scheduled journey time.
Figure 19 Comparison of satisfaction with punctuality and journey time
4.2.2 Influence of punctuality on satisfaction with value for money
The relationship with value for money is slightly different (Figure 20). The green diagonal axis is more definitive towards the low satisfaction end of the matrix, meaning passengers are very unlikely to be satisfied with value for money if they are dissatisfied with punctuality. For this, the conclusion drawn is that dissatisfaction with punctuality drives dissatisfaction with value for money.
Very Dissatisfied
Fairly Dissatisfied
NeitherFairly
SatisfiedVery Satisfied
1 2 3 4 5
Ver
y D
issa
tisfie
d
1 26% 4% 3% 1% 0%
Fai
rly
Dis
satis
fied
2 15% 19% 10% 4% 2%
Nei
ther
3 26% 27% 38% 8% 3%
Fai
rly
Sat
isfie
d
4 27% 42% 43% 73% 17%
Ver
y S
atis
fied
5 6% 9% 6% 14% 78%
The
leng
th o
f tim
e th
e jo
urne
y w
as s
ched
uled
to
take
(s
peed
)
Satisfaction with Punctuality
79
Figure 20 Comparison of satisfaction with punctuality and value for money
4.2.3 Influence of punctuality on satisfaction with available space
Considering passengers’ responses to the question about sufficient space to sit or stand (Figure 21) shows passengers are unlikely to be dissatisfied with space if the train is on time. This may be because a high proportion of these trains are not crowded. It may also be because passengers who board crowded trains tolerate having to stand if the train takes its scheduled time, but become less satisfied if it is late. This is reflected by an increased tendency to be dissatisfied with availability of space when satisfaction with punctuality decreases.
Figure 21 Comparison of satisfaction with punctuality and crowding
4.2.4 Influence of punctuality on satisfaction with ease of getting on and off the train
With respect to ease of being able to get on and off the train, the strength of influence from satisfaction with punctuality is insignificant. In all columns, the level of response is not significantly different to be able to determine a relationship (Figure 22).
Very Dissatisfied
Fairly Dissatisfied
NeitherFairly
SatisfiedVery Satisfied
1 2 3 4 5
Ver
y D
issa
tisfie
d
1 63% 42% 32% 18% 7%
Fai
rly
Dis
satis
fied
2 18% 30% 31% 27% 16%
Nei
ther
3 9% 16% 23% 24% 19%
Fai
rly
Sat
isfie
d
4 6% 11% 10% 26% 31%
Ver
y S
atis
fied
5 3% 1% 3% 5% 27%The
val
ue f
or m
oney
for
the
pric
e of
you
r tic
ket
Satisfaction with Punctuality
Very Dissatisfied
Fairly Dissatisfied
NeitherFairly
SatisfiedVery Satisfied
1 2 3 4 5
Ver
y D
issa
tisfie
d
1 31% 20% 13% 7% 2%
Fai
rly
Dis
satis
fied
2 16% 16% 14% 10% 4%
Nei
ther
3 16% 19% 29% 19% 8%
Fai
rly
Sat
isfie
d
4 25% 35% 33% 48% 38%
Ver
y S
atis
fied
5 13% 11% 11% 16% 47%
Satisfaction with Punctuality
Suf
ficie
nt r
oom
for
all
the
pass
enge
rs t
o si
t/st
and
80
Figure 22 Comparison of satisfaction with punctuality and ease of being able to get on and off the train
4.2.5 Influence of punctuality on satisfaction with train cleanliness
Comparing satisfaction with punctuality and cleanliness of the inside of the train, there is also no significant relationship. Cleanliness of the inside of the train is noted as the second biggest driver of overall satisfaction in Table 6. In order for it to be a key driver alongside punctuality, it is reasonable to expect the two to have differing characteristics. In all columns, the proportion of people who are satisfied is reasonably consistent in comparison to other criteria (Figure 23).
Figure 23 Comparison of satisfaction with punctuality and train cleanliness
Very Dissatisfied
Fairly Dissatisfied
NeitherFairly
SatisfiedVery Satisfied
1 2 3 4 5
Ver
y D
issa
tisfie
d
1 11% 4% 3% 2% 1%
Fai
rly
Dis
satis
fied
2 10% 8% 6% 4% 2%
Nei
ther
3 23% 22% 32% 17% 7%
Fai
rly
Sat
isfie
d
4 40% 51% 49% 59% 42%
Ver
y S
atis
fied
5 16% 14% 10% 18% 48%
Satisfaction with Punctuality
The
eas
e of
bei
ng a
ble
to g
et o
n an
d of
f th
e tr
ain
Very Dissatisfied
Fairly Dissatisfied
NeitherFairly
SatisfiedVery Satisfied
1 2 3 4 5
Ver
y D
issa
tisfie
d
1 17% 10% 6% 4% 2%
Fai
rly
Dis
satis
fied
2 23% 21% 16% 13% 7%
Nei
ther
3 21% 27% 34% 21% 12%
Fai
rly
Sat
isfie
d
4 32% 36% 38% 52% 48%
Ver
y S
atis
fied
5 8% 6% 4% 9% 31%
Satisfaction with Punctuality
The
cle
anlin
ess
of t
he in
side
of
the
trai
n
81
5. Train punctuality observations This section looks at how AGA’s train punctuality at terminus has changed over the past three years. This includes analysis of punctuality by route and by time of day, limited to the NRPS survey dates. This section takes into account all the train services that were not subject to cancellation.
5.1 Overview
Within the NRPS survey dates for the past three years, the proportion of trains arriving early or on time (‘right-time’) has decreased from 70 per cent in 2012 to 68 per cent in 2014, while the proportion within five minutes of scheduled time (RT5) has remained consistent at around 91%. There is, on average, a 23 percentage points difference between the proportion of ‘right-time’ and the proportion of RT5. This is comparable with the latest information published by Network Rail showing a 91 per cent Moving Annual Average (MAA). Figure 24 outlines the variations in ‘right-time’ and RT5 services across AGA over the three years of the study.
Figure 24 Train punctuality by year
5.1.1 Punctuality by time of day
This section considers how the proportion of services arriving at their destination on time varies by time of day. As shown in Figure 257, the proportion is generally lower in peak departure hours compared with those departing in off peak hours. This conclusion supports the analysis in Section 3.3 on passenger satisfaction, where satisfaction is generally lower in peak hours.
7 Due to significantly lower number of respondents using services before 0600 and after 2000, the outcome displayed below is limited to hours between 0600 and 2000.
70% 68% 68%
92% 91% 92%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2012 2013 2014
Prop
ortio
n of
trai
ns
Year
Arriving early or on time (RT) Arriving within five minutes of scheduled arrival time (RT5)
82
Figure 25 Trains arriving early or on time by departure hour
5.1.2 Punctuality by route
Figure 26 outlines the proportion of services arriving early or on time at their destination across the franchise, illustrating an improvement in punctuality across West Anglia, including Stansted Express. The Stansted Express results reflect the large improvement in passenger satisfaction seen on this route in Figure 10, which may be because of changes to the definition of the NRPS building block. However, Figure 10 also shows improvements in satisfaction on Great Eastern routes, where train performance worsened. The only route that shows worsening train performance and worsening passenger satisfaction is Rural.
Figure 26 Trains arriving early or on time by route
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Prop
ortio
n of
trai
ns
Departure hour
2012 2013 2014
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Intercity StanstedExpress
Mainline West AngliaOuter
West AngliaInner
Metro Rural
Prop
ortio
n of
trai
ns
Routes
2012 2013 2014
83
6. Passenger punctuality analysis
This section looks at how the punctuality of the train service at intermediate stations impacts on the punctuality experienced by the passenger. While train punctuality is measured at train destination, a high proportion of passengers do not alight there. This section, therefore, considers the punctuality of passengers when they arrive at their destination.
6.1 Overview
Figure 27 outlines the variations in the proportion of passengers arriving early or on time and those arriving within five minutes of the scheduled arrival time across the franchise for the three years of the study. These follow a similar trend to those for train destination with a decrease in the level of ‘right-time’ arrivals. The proportion of respondents arriving early or on time declined to 60 per cent in 2014 in comparison with 63 per cent in 2012. However, the proportion of respondents arriving within five minutes of their scheduled arrival time (RT5) improved by two percentage points to 91 per cent.
Figure 27 Passenger punctuality by year
6.2 Punctuality by time of day
Sections 3.3 and 5.1.1 showed that both satisfaction with punctuality and actual punctuality of trains are at their lowest in peak hours. Figure 28 is more difficult to draw conclusions from as the sample of passengers is considerably smaller than the sample of trains. On average, for 2014, 57 per cent of passengers arrived on time in the AM peak, 58 per cent in the PM peak and 64 per cent in the off-peak. Therefore, at a high level, punctuality in the peaks is lower for passengers as well as for trains, and explains passenger satisfaction with punctuality being lower in these periods.
89% 88%91%
63%56%
60%
0%
20%
40%
60%
80%
100%
2012 2013 2014
Prop
ortio
n of
pas
seng
ers
Year
Arriving within five minutes of scheduled time (RT5) Arriving early or on time (RT)
Due to significantly lower number of respondents using services before 0600 and after 2000, analysis by time of
day is limited to hours between 0600 and 2000.
84
Figure 28 Passengers arriving early or on time by departure hour
6.3 Punctuality by route
When considering passenger punctuality by route, the results are very similar to those seen in Section 5.1.2. Improvements in ‘right-time’ arrivals have been seen on West Anglia routes; most notably on Stansted Express. The improvements seen on Stansted Express are more significant from a passenger perspective than they are from a train perspective, and are more aligned with the scale of improved satisfaction seen in Section 3.4, which, as noted, may be because of changes in how the Stansted Express building block is defined within NRPS. The worsening punctuality seen on Great Eastern routes is not reflected in passenger satisfaction with punctuality and would require further consideration.
Figure 29 Passengers arriving early or on time by route
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
06 07 08 09 10 11 12 13 14 15 16 17 18 19 20
Prop
ortio
n of
pas
seng
ers
Departure hour
2012 2013 2014
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Intercity StanstedExpress
Mainline West AngliaOuter
West AngliaInner
Metro Rural
Prop
ortio
n of
pas
seng
ers
Route
2012 2013 2014
85
Considering the full breakdown of lateness at route level indicates that, on most routes, the proportion of respondents arriving significantly late is notably low, reflecting the PPM levels of around 90 per cent. Exceptions to this are Mainline and Intercity routes which see nearly five per cent of services arrive more than 15 minutes late.
Figure 30 Distribution of passenger lateness by route
6.3.1 Punctuality by time of day and route
Due to small sample size in some time bands, the trends seen by time of day at route level are not presented here. The charts have been retained in the Appendix for reference.
6.4 Passenger punctuality vs train punctuality
The analysis in Section 6 is based on passenger lateness. It is important to note that this can be significantly different from train lateness, whether Public Performance Measure (PPM) or ‘right-time’. When comparing the recorded train punctuality with actual passenger experience – as seen in Figure 31 – there is a distinct gap between the two. This demonstrates that, even though a train arrives on time, some passengers do not.
Figure 31 Train lateness vs passenger lateness
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Intercity StanstedExpress
Mainline West AngliaOuter
West AngliaInner
Metro Rural
Prop
ortio
n of
pas
seng
ers
Routes
RT or Early 1-4 minutes late 5-14 minutes late
15-29 minutes late 30-59 minutes late 60+ minutes late
0%
5%
10%
15%
20%
25%
30%
-5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 171819 20 21 22 23 24 25 2627 28 29 30Prop
ortio
n pa
ssen
gers
/trai
ns
Lateness (minutes)
Passenger Train
86
Figure 32 further emphasises this by considering only services travelling away from London. These trains are selected because the destination of the passenger and the train is likely to be different. The graph shows a bigger lag, of over one minute, between the two lines. This means:
If a train from London is reported as being on time, the average passenger is one minute late; and
On train services departing London, 69 per cent arrive at the ultimate destination on time or early, compared with 56 per cent of passengers.
Figure 32 Train lateness vs passenger lateness for services departing London
0%
5%
10%
15%
20%
25%
30%
-5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 101112131415161718192021222324252627282930
Prop
ortio
n of
pas
seng
ers/
trai
ns
Lateness (minutes)
Passenger Train
87
7. Relationship between satisfaction with punctuality and actual punctuality The previous sections have considered satisfaction with punctuality and actual punctuality in isolation. This section brings these together to establish how the two are related. The first point to note is that passenger satisfaction with punctuality follows a similar trend for both ‘right-time’ and arrivals up to five minutes late. As Figure 33 illustrates, all measures increased since 2013.
Figure 33 Comparison of satisfaction with punctuality and actual passenger punctuality
7.1 Satisfaction with punctuality by lateness
Using the approach defined in Section 2.3, a dataset was produced with each respondent mapped to the actual level of delay they experienced on the day of the survey. This dataset enables consideration of how satisfaction changes as the passenger lateness – or delay experienced – increases. Note that weighting is not applied as part of this process, so any differences between graphs in this section and those in previous sections should be attributed to weighting of responses.
In the similar study published in 2009, titled “Examining the links between Train Performance Measures and Customer Satisfaction”, graphs similar to Figure 348 were produced demonstrating three defining factors of a passenger’s perception of punctuality.
1. The intercept on the vertical axis, which determines how satisfied a passenger is with punctuality, even when the train is on time or early. In Figure 34, below, the value is 83.8 per cent.
2. The gradient, which defines by how much satisfaction reduces for each minute of additional delay. In Figure 34 the value is just over three percentage points.
3. The tipping point, at which the curve takes a notable increase in gradient. This is the point at which passengers’ sensitivity to delay heightens. Figure 34 does not have a tipping point, implying that passengers perceive delay from the minute they experience it and satisfaction falls consistently as delay increases.
These findings are important to reiterate.
8 Due to significantly lower number of respondents experiencing delays in excess of 20 minutes, analysis on these were proved to be not accurate. In this section, results associated with lateness are limited to 20 minutes of delay.
89% 88% 91%
63%56%
60%
73% 72% 75%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2012 2013 2014
Prop
ortio
n of
pas
seng
ers
satis
fied
Year
Proportion of passengers arriving within five minutes of scheduled time (RT5)
Proportion of passengers arriving early or on time (RT)
Satisfaction with punctuality
88
1. Even when passengers arrive at their destination on time or early, not all are satisfied with punctuality; in fact close to 18 per cent of passengers surveyed are not satisfied, even when on time.
2. With an increase in lateness, satisfaction decreases linearly; Figure 34 shows this is three percentage points for every additional minute of lateness.
3. Even at high levels of delay, some passengers are still satisfied; 51 per cent of passengers are satisfied with delays between 12 and 15 minutes.
Figure 34 Satisfaction with punctuality vs passenger lateness
7.1.1 By gender
Figure 35 shows how satisfaction with punctuality varies among female and male respondents. Section 3.2 showed that female respondents are usually more satisfied across the NRPS surveys and this is no different. As expected, the analysis in this section shows a close and consistent pattern of satisfaction decline with an increase in lateness experienced, with the gradient of both being at around three percentage points per minute, with no tipping point being evident.
Figure 35 Satisfaction with punctuality vs passenger lateness by gender
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15
Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
Female Male
89
7.1.2 By age
The NRPS groups respondents into multiple age categories. However, to maintain sufficient sample for analysis, some amalgamation of groups has taken place, resulting in the following age ranges.
16-25 26-34 35-44 45-54 55-64 65+
As outlined in Figure 36 and Figure 37, satisfaction with punctuality across the age groups is relatively consistent in terms of decline as lateness increases. The sample size makes inferences difficult, but there appears to be a steeper gradient for 26-34 year olds, indicating a higher sensitivity to increasing lateness. While Section 3.2 shows younger people to be generally less satisfied with punctuality, they are more forgiving of delays in excess of nine minutes.
Figure 36 Satisfaction with punctuality vs passenger lateness by age (1)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15
Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
16-25 26-34 35-44
The following observations are noted on sample size:
The number of respondents in age category 16-25 who experienced delays in excess of seven
minutes is significantly lower (an average of three respondents per minute of lateness in
comparision with an average of 49 respondents per minute of lateness overall) which results in an
inaccurate representation of satisfaction level; and
The number of respondents in age category 26-34 who experienced delays in excess of eight
minutes is significantly lower (an average of three respondents per minute of lateness in
comparision with an average of 72 respondents per minute of lateness overall) which results in an
inaccurate representation of satisfaction level.
The following observations are noted on sample size:
The number of respondents in age category 65+ who experienced five minutes and eight minutes of
delays is significantly lower (an average of four respondents per minute of lateness in comparision
with an average of 48 respondents per minute of lateness overall) which results in an inaccurate
representation of satisfaction level; and
The number of respondents in age category 55-64 who experienced delays between 12-15 minutes is
significantly lower (an average four respondents per minute of lateness in comparision with an
average of 51 respondents per minute of lateness overall) which results in an inaccurate
representation of satisfaction level.
90
Figure 37 Satisfaction with punctuality vs passenger lateness by age (2)
7.1.3 By journey purpose
An additional parameter that influences satisfaction is the journey purpose (e.g. commuter, business or leisure). Figure 38 outlines the satisfaction with punctuality by journey purpose, as given on the NRPS response.
Figure 38 Satisfaction with punctuality vs passenger lateness by journey purpose
There are clear differences between the three journey purposes with the following intercept, gradient and tipping point outlined for each.
Journey Purpose Intercept Gradient Tipping Point
Commuter 74.9 per cent 5 percentage points None
Business 91.0 per cent 2 percentage points 5 minutes
Leisure 94.0 per cent 1 percentage point 8 minutes
Overall 83.8 per cent 3 percentage points None
Intercept
This shows that, even when a service arrives on time (i.e. ‘right-time’ or early), satisfaction with punctuality is just 74.9 per cent for commuters, 91 per cent for business travellers and 94 per cent for leisure travellers.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15
Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
45-54 55-64 65+
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15
Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
Leisure Commuters Business
91
The ‘shortfall’ below 100 per cent could provide some indication of the influence that previous journeys has on satisfaction with punctuality ‘today’. The finding that leisure travellers are more influenced by today’s travel than commuters is consistent with this, given the average frequency of travel for these groups. The National Rail Travel Survey (NRTS) suggests that 77 per cent of commuters travel five-days per week, whereas 70 per cent of leisure passengers travel less than once every month (or have not travelled before).
This could indicate that commuters are not answering the NRPS survey based solely on the experiences of that day. Instead, they may be using the survey as an opportunity to express views on past performance, consciously or otherwise.
Gradient
While commuters are less happy than leisure and business passengers when trains are on time, they are also more sensitive to delay, with each additional minute of delay costing five percentage points in satisfaction. For leisure passengers it is one percentage point per minute of delay and for business travellers, two percentage points per minute.
Tipping point
Where it has previously been difficult to identify a tipping point, it becomes more evident when looking by journey purpose. While satisfaction falls from the first minute of lateness, it is evident that, for leisure passengers, the gradient increases at around eight minutes delay, and around five minutes for business passengers.
7.1.4 By route
The analysis in this section is subject to significantly lower NRPS sample size as delay increases. However, some inferences can still be drawn from Figure 39 and Figure 40. Observations include:
Respondents using Intercity services are not particularly sensitive to delay; 78 per cent of passengers are satisfied with delays of up to 15 minutes; and
All other, generally shorter distance, routes have very similar (more aggressive) characteristics regarding the rate of decrease in satisfaction; this could be due to the passenger mix on these routes, or the comparably short journey times meaning that delay is a higher proportion of overall journey time.
The following observations are noted regarding sample size:
The number of respondents using services in Rural route who experienced delays in excess of five
minutes is significantly lower (an average of three respondents per minute of lateness in comparision
with an average of 36 respondents per minute of lateness overall) which results in an inaccurate
representation of satisfaction level; and
The number of respondents using services in Metro route who experienced delays between six and
ten minutes is significantly lower (an average of eight respondents per minute of lateness in
comparision with an average of 50 respondents per minute of lateness overall) which results in an
inaccurate representation of satisfaction level.
92
Figure 39 Satisfaction with punctuality vs passenger lateness by route (1)
Figure 40 Satisfaction with punctuality vs passenger lateness by route (2)
7.1.5 By ticket type
A parameter that is also known to affect passenger satisfaction is ticket type. This is largely because passengers’ journey purpose and ticket type is correlated with frequency of travel. The analysis by ticket type takes into account the majority of ticket types available to passengers. However, due to sample size, they have been grouped as summarised in Table 7. Some ticket types have been excluded from the analysis due to insufficient sample size; they are also detailed in the table below.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15
Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
Intercity Metro Mainline Rural
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15
Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
West Anglia Outer West Anglia Inner Stansted Express
The following observations are noted regarding sample size:
The number of respondents using services in West Anglia Inner route who experience delays between
five to seven minutes and in excess of 15 minutes are significantly lower (an average of five
respondents per minute of lateness in comparision with an average of 50 respondents per minute of
lateness overall) which results in an inaccurate representation of satisfaction level; and
The number of respondents using services in West Anglia Outer route who experience delays
between seven to ten minutes is significantly lower (an average of 14 respondents per minute of
lateness in comparision with an average of 78 respondents per minute of lateness overall) which
results in an inaccurate representation of satisfaction level.
93
Ticket Type Ticket Type Grouping for Analysis
Anytime single/return (includes standard single/return) Anytime single/return
Anytime day single/return
Off-peak/super off-peak single/return Off-peak/super off-peak single/return
Off-peak/super off-peak day single/return
Advance Advance
Day travelcard Day travelcard
Weekly or monthly season ticket Weekly or monthly season ticket
Annual season ticket Annual season ticket
Special promotion ticket
Excluded Rail staff pass/Privilege ticket/Police concession
Free travel pass (e.g. Freedom pass)
Table 7 Ticket type groupings
Figure 41 outlines how satisfaction with punctuality varies based on increased lateness for respondents in possession of anytime single/return, off-peak/super off-peak single/return, advance and day travelcard tickets. Figure 42 covers this for season tickets users.
It is clear that passengers in possession of season tickets are generally less satisfied with punctuality. These passengers reflect commuter characteristics, while advance ticket holders more closely resemble leisure travel.
The following observations are noted regarding sample size:
The number of respondents using Day Travelcard who experienced delays between six to eight
minutes and in excess of 15 minutes are significantly lower (an average of five respondents per
minute of lateness in comparision with an average of 29 respondents per minute of lateness overall)
which results in an inaccurate representation of satisfaction level; and
The number of respondents using Advance tickets who experienced delays between 9 to 12 minutes
is significantly lower (an average of 14 respondents per minute of lateness in comparision with an
average of 45 respondents per minute of lateness overall) which results in an inaccurate
representation of satisfaction level.
94
Figure 41 Satisfaction with punctuality vs passenger lateness by ticket type (1)
Figure 42 Satisfaction with punctuality vs passenger lateness by ticket type (2)
7.1.6 By frequency of travel
In order to maintain a representative sample across the categories of journey frequency listed in the NRPS, they are grouped together into:
Three or more times a week;
One to eight times per month; and
Less that once a month.
Figure 43 illustrates how satisfaction with punctuality changes with increased lateness. Again, given that there is a natural correlation between frequency of travel and journey purpose, the patterns are as expected, with only 73 per cent of passengers travelling three or more times a week being satisfied with punctuality. As has been seen for commuters, this reduces by
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15
Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
Anytime single/return Off-peak/super off-peak single/return
Advance Day travelcard
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15
Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
Weekly or monthly season ticket Annual season ticket
95
approximately five percentage points for every minute of lateness.
Figure 43 Satisfaction with punctuality vs passenger lateness by frequency of travel
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15
Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
Three or more times a week One to eight times per month Less than once a month
96
8. Relationship between overall satisfaction and actual punctuality As outlined in Section 7, there is a clear decline in satisfaction with punctuality as lateness increases. This section looks at the relationship between passenger lateness and overall satisfaction. If punctuality does have a substantial bearing on overall satisfaction, similar relationships should be apparent in this section.
Before beginning the analysis, it is important to note that some of the relationships shown below may be circumstantial. Overall satisfaction can be affected by many factors. As a result of this, only a selection of the analysis from Section 7 is repeated.
Figure 44 shows a clear reduction in overall satisfaction with increased lateness. In this graph overall satisfaction starts to reduce after two minutes of lateness, falling by 14 percentage points by six minutes lateness, and then stabilising until nine minutes of lateness, at which point there is an obvious tipping point. While the intercept with the vertical axis is at approximately the same level as for satisfaction with punctuality, the rate of decline is less aggressive.
Figure 44 Overall satisfaction vs passenger lateness
8.1 Overall satisfaction by length of delay and gender
Following a similar pattern to satisfaction with punctuality by gender (Figure 35), the overall satisfaction also declines with increased lateness (Figure 45). However, the rate of decline is lower than that for satisfaction with punctuality. Both male and female respondents react to increased lateness in the same way, with overall satisfaction reducing at approximately the same rate.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15
Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
97
Figure 45 Overall satisfaction vs passenger lateness by gender
8.2 Overall satisfaction by length of delay and journey purpose
For journey purpose, overall satisfaction trends are consistent with those seen for satisfaction with punctuality. In the case of overall satisfaction (Figure 46), the rate of decline as lateness increases is not as aggressive as for satisfaction with punctuality (Figure 38). However, this is consistent across all journey purposes, allowing little inference to be drawn about the strength of relationship between overall satisfaction and satisfaction with punctuality by journey purpose.
Figure 46 Overall satisfaction vs passenger lateness by journey purpose
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
Female Male
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0 1 2 3 4 5 6 7 8 9 10-11 12-15
Prop
ortio
n of
pas
seng
ers
satis
fied
Lateness (minutes)
Leisure Commuters Business
98
9. Impact of remapping While, historically, Metro and West Anglia routes had lower levels of satisfaction with punctuality and overall satisfaction, our analysis suggests that they have improved sufficiently to close the gap on other routes. As a result, the remapping of these services will not result in a significant difference between the NRPS results for Abellio Greater Anglia before and after 31 May 2015. Figure 47 outlines the overall satisfaction for each year, considering the franchise as it has been, and then with the remapped routes excluded. In 2014, had Metro and West Anglia Inner routes been excluded, overall satisfaction would have been unchanged in comparison with 2013, while satisfaction with punctuality would have been one percentage point higher.
Figure 47 Overall satisfaction - route remapping comparison
Figure 48 shows satisfaction with punctuality for 2012, 2013 and 2014 considering the franchise as a whole and with the exclusion of remapped routes
Figure 48 Satisfaction with punctuality - route remapping comparison
70%75% 76%
81%77% 77%
0%
20%
40%
60%
80%
100%
2012 2013 2014Prop
ortio
n of
pas
seng
ers
satis
fied
Year
All AGA Remapped excluded
74%78% 80%81% 79% 80%
0%
20%
40%
60%
80%
100%
2012 2013 2014
Prop
ortio
n of
pas
seng
ers
satis
fied
Year
All AGA Remapped excluded
99
Appendix 1 – Passenger punctuality by time of day and route Figure 49 and Figure 50 show respondents arriving to their destination early or on time on GA routes on the Great Eastern and West Anglia routes.
Figure 49 Passengers arriving early or on time by hour and route (1)
Figure 50 Passengers arriving early or on time by hour and route (2)
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
06 07 08 09 10 11 12 13 14 15 16 17 18 19 20
Prop
ortio
n of
pas
seng
ers
Departure Hour
Intercity Metro Mainline Rural
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
06 07 08 09 10 11 12 13 14 15 16 17 18 19 20
Prop
ortio
n of
pas
seng
ers
Departure Hour
West Anglia Outer West Anglia Inner Stansted Express
Contact Transport Focus
Any enquiries regarding this research should be addressed to:
Mike HewitsonHead of Policy and IssuesTransport Focust 0300 123 0830e [email protected] www.transportfocus.org.uk
Fleetbank House 2-6 Salisbury Square London EC4Y 8JX
Transport Focus is the operatingname of the Passengers’ Council
Published in November 2015© 2015 Transport Focus