development and validation of the driving fatigue scale · development and validation of the...
TRANSCRIPT
Development and validation of the driving fatigue scale
Paula A . D esm ond*, G erald M atthew s** and P eter A . H ancock*
*H um an F ac to rs R esearch L ab ora to ry , U n iv ersity o f M in n e so ta , U S A , **U n iv ersity o f
D undee, D epartm ent o f Psychology, D undee, Scotland.
D riv in g p e rfo rm an ce m ay be se n sitiv e to sta te ch an g es in d u ced by d ru g s an d a lco h o l,
inc lu d in g fa tigue and de-aro u sa l. M ascord , W alls and S ta rm er (1995) found th a t d riv ers
rep o rted inc reased b o redom and fa tig u e and w ere less w illin g to d riv e fo llo w in g a lcohol
in take. T hese find ings are co nsisten t w ith stud ies o f d riv ing fa tigue such as F u lle r’s (1984)
study w hich found that truck drivers at the end o f their shift experienced increased feelings o f
drow siness, lack o f energy and the urge to stop driving. Several scales have been developed to
assess sym ptom s o f fatigue. Saito , K ogi and K ash iw agi (1970) d eveloped a fa tigue scale to
m easure three types o f fa tigue sym ptom s. T he firs t sym ptom is asso c ia ted w ith d row siness
and du llness, the second concerns m otivational aspects, atten tion and concen tra tion , and the
th ird sym ptom re la tes to physical sym ptom s. T he psychom etric bases o f ex is ting scales are
o ften suspect, and th e ir app licab ility to d riv ing is uncertain . S tud ies o f d riv ing fatigue, have
em ployed un id im ensional fa tigue scales w hich m ake the d icho tom ous d istin c tio n betw een
a le r tn e s s an d fa tig u e (e .g . M a c k ie , O ’H an lo n & M c A u le y , 1 974). H o w e v e r , a
m ultid im ensional m easure o f fatigue is clearly necessary if a com prehensive understand ing o f
the dynam ics o f the subjective experience o f fatigue are to be understood.
T he p resen t study em ploys a m u ltid im ensional fa tigue scale in th ree stud ies o f sim ula ted
d riv ing perform ance. In all three studies, d rivers perfo rm ed both a fatigu ing drive, in the first
p a rt o f w hich a ch aracter-de tec tion task w as p erfo rm ed in add ition to d riv ing , and a con tro l
d rive in w hich no ad d itional secondary task w as perform ed . D rivers co m p le ted the fa tigue
scale and m easu res o f m ood , cogn itive and m o tiv a tio n a l co m p o n en ts o f sub jec tiv e sta tes
before and after both drives. T he study had th ree m ain aim s. T he firs t a im w as to determ ine
w hether fatigue is characterized by a single d im ension only. T he second aim w as to exam ine
the relationsh ip betw een fatigue and general sta te d im ensions. T he th ird aim w as to determ ine
w hether driv ing fatigue re la tes only to d river stress tra its , o r to personality in general as w ell.
H ence , the fa tigue scales w ere co rre la ted w ith the general tra its , such as ex trav ersio n and
n eu ro tic ism , as w ell as d riv in g -sp ec ific tra its a sso c ia ted w ith v u ln e rab ility to stress and
fatigue.
- 62 1 -
M E T H O D
T hree stud ies w ere cond u c ted to exam ine the e ffec ts o f task -in d u ced fa tigue on su b jec tive
state and d riv ing p erfo rm ance using the A ston D riv er S im ula to r. S u b jec ts p erfo rm ed bo th a
con tro l and fa tigue d rive on separa te occasions. T he p ap er p resen ts d a ta ob ta in ed fro m the
fatigue condition only. In this drive, subjects perfo rm ed a dem an d in g ch a rac te r de tec tion task
in ad d itio n to d riv in g . T he fa tig u e in d u c tio n p ro ced u re re q u ire d su b je c ts to re a c t to
in fo rm ation p resen ted on road-signs w hilst d riv ing a t a constan t speed. E ach sign p resen ted a
sequence o f seven characters as the d riv er approached it e .g . C U 4 K P IA . F o r each sequence,
sub jects w ere requ ired to look fo r odd o r even num bers and w ere in stru c ted to p ress a button
se t in to the stee rin g w h ee l w hen they d e tec ted the ‘ta rg e t’ nu m b er. A c o lo re d sta r w as
p re sen ted b e fo re each seq u en ce w hich se rved to p rim e su b je c ts to lo o k fo r e ith e r o d d
(denoted by a red star) o r even (denoted by a green star) num bers. A to tal o f 528 signs w ere
p re sen ted to sub jec ts. In all th ree s tud ies, su b jec ts p erfo rm ed th ree p rac tice ru n s p rio r to
p e rfo rm in g the fa tig u e and co n tro l d rives. F o llo w in g p ra c tic e , su b je c ts c o m p le te d the
subjective sta te m easures.
Subjects
80 drivers (40 m en and 40 w om en) w ere run in S tud ies 1 and 2, and 9 6 d riv ers (48 m en and
48 w om en) w ere run in S tudy 3. M ean age o f d rivers in S tudy 1, 2 and 3 w ere 21 .6 (range:
18-30), 20 .58 (range: 18-30) and 20.67 (range 18-30) respectively . T he sub jec t sam ples w ere
com bined to give a total sam ple o f 256 subjects.
Subjective m easures
M easures o f fa tig u e , m o o d and o th er sub jec tiv e sta te m easu res w ere ad m in iste red in the
stud ies. B ased on a concep tual rev iew o f the lite ra tu re , a 24 -item fa tig u e sca le w as used.
M ood w as assessed w ith the U W IS T M ood A djective C heck lis t (U M A C L : M atthew s, Jones
and C ham berlain , 1990) w hich assesses b ipo lar m ood d im ensions o f tense arousal, energetic
a ro u sa l , an d h e d o n ic to n e (p le a sa n tn e s s o f m o o d ) , to g e th e r w ith an a d d it io n a l
anger/frustration scale. A m odifica tion o f Sarason, Sarason, K eefe, H ayes & S h ea rin ’s (1986)
C ognitive In terference Q uestionnaire (CIQ ) w as used to assess in trud ing though ts. T he scale
co m p rises 1 0 item s re la tin g to task -re lev an t in te rfe ren ce and 1 1 item s re la tin g to task-
irre lev an t personal item s. T w o valida ted scales w ere used to a ssess p e rce iv ed co n tro l and
m o tivation (M atthew s, C am pbell, Joyner, H uggins, F alconer and G illilan d , 1996). Sub jects
a lso co m p le ted the E y sen ck Personality Q uestionnaire (E y sen ck and E y sen ck , 1975), the
D riv ing B ehavior Inven tory (D BI: G ulian , M atthew s, G lendon , D av ies and D ebney , 1989),
and a 10-item F a tig u e P ro n e n ess scale . T he E P Q m easu res th ree m a jo r d im en sio n s o f
p ersonality : E x traversion (E), N euro tic ism (N ) and P sych o tic ism . T h e D B I is a v a lid a ted ,
questionnaire m easu re o f pred isposition to various d im ensions o f d riv er stress. In th is study ,
- 6 2 2 -
the A g gression , D islike o f D riv ing , A lertness scales w ere o f in terest. T he F atigue P roneness
scale consists o f ten item s requ iring subjects to ra te how th e ir fee ling and behav iors change as
a consequence o f p ro longed driving.
P rocedure
S ub jects co m p le ted the D B I, E P Q and F a tig u e P ro n en ess sca les p r io r to p e rfo rm in g the
fa tigue and con tro l d rives. Sub jec ts p erfo rm ed th ree p rac tice ru n s b efo re the m ain d rives.
F o llo w in g th e p rac tice ru n s, sub jec ts com p le ted the fa tigue sta te sca les, m ood , m o tivation ,
cognitive in terference and perceived con tro l scales. Sub jects then com pleted the m ain drives.
The contro l d rives in S tudy 1, 2 and 3 lasted 13, 19 and 40 m inu tes respec tive ly . T he duration
o f the fa tigue d riv es in S tudy 1, 2 and 3 w ere 33, 38 and 40 m inu tes respec tive ly . F o llow ing
the m ain d rive, subjects com pleted all the sub jective sta te m easures.
R E SU L T S
A prin c ip a l co m p o n en t analys is o f the 24 fa tigue scale item s w as co n d u c ted , fo llo w ed by
ob lique ro ta tio n , u sing the d irec t ob lim in crite rion . F o u r substan tial fac to rs w ere ob ta ined :
v isual fatigue, boredom , m alaise and m uscu la r fa tigue (see T ab le 1 fo r scale item s).
T ab le 1 : F atigu e sca le item s
B oredom V isual fatigue M alaise M uscu lar fatigue
B ored F lickering in eyes Flave a headache F e e l t i r e d in th e w hole body
W ould ra th er be do ing som eth ing else
F ee lin g o f h eav in ess in the eyes
H e a r i n g a b i l i t y reduced
H av ing trem ors in the lim bs
F ed up w ith the task E y es feel strained H um m ing in ears F eel stiff in the legs
A pathetic V ision is b lurred F eel sick o r nauseous U n a b le to s tra ig h ten up in postu re
D o n ’t w an t to do the task ev er again
R o a d a p p e a r s to ‘sw im ’
F eel stom ach pains
D o n ’t c a r e w h a t happens next
U n a w a re o f o b je c ts o f f the road
F eel ill
F i n d t h e t a s k m onotonous
D o n ’t w a n t to th in k about the task
A s expected , v isua l fa tigue w as defined by item s re la ting to v isual d iscom fort. B oredom was
associated w ith fee lings o f apathy , reduced in terest in the task and an aversion to fu rther task
effort. M ala ise item s re lated to physical sym ptom s o f illness and reduced aud ito ry sensitiv ity .
- 6 2 3 -
M uscular fatigue w as defined by item s related to bodily d iscom fort both in postu re and lim bs.
B asic scale p roperties are given in T ab le 2: m ean, standard dev ia tion , and a lp h a coeffic ient.
A lpha coeffic ien ts fo r v isua l fatigue, boredom and m alaise scales w ere sa tisfactorily h igh and
ind ica te adequate in ternal consistencies o f the scales. T he fo u r item m u scu la r fa tigue scale
show ed a low er alpha than the o ther scales, and so has rather w eak in ternal consistency. Table
2 show s the resu lts o f the paired t-tests calcu lated fo r p re- and post-d riv e fa tigue scales. T he
resu lts ind ica te that the scales w ere sensitive to the fa tigue m an ipu la tion : boredom , v isual
fatigue, m uscu la r fatigue and m alaise increased significantly fo llow ing the drive.
T ab le 2 : M ean s, stan d ard d ev ia tio n s an d a lp h a co e ffic ie n ts for p re- an d p ost-ta sk
fatigue scales
Pre-drive Post-drive
Scale M ean Standard
deviation
a M ean Standard
deviation
a t
B oredom 13.45 8.43 0.89 21.75 9.74 0.92 17 4 9 ***
V isual 7.33 6.19 0 . 8 6 13.76 7.12 0.89 17 16***
M alaise 2.92 4.08 0.75 5.17 5.81 0.82 9 0 8 ***
M uscu lar 5.12 3.85 0.59 9.74 4.53 0.70 19 56***
***p < . 0 0 1
T able 3 g ives the in tercorre la tions o f pre- and post d rive fa tigue-scale scores. T he test-re test
corre la tions are g iven on the lead ing d iagonal in the table. T he co rre la tions are sa tisfacto rily
h igh and suggest that ind iv idual d ifferences w ere stable o v e r tim e. T est-re tes t rs w ere less
than the alphas repo rted in T able 2, bu t w ere still substantial, im p ly ing m odera te stab ility o f
ind iv idual d ifferences over tim e. Pre- and post-d rive in ter-sca le co rre la tions are g iven above
and below the lead ing d iagonal, respec tive ly . C o rre la tions w ere substan tial and suggest the
presence o f a general fatigue factor.
T able 3 : In tercorrelations betw een pre- and post-drive fa tigue scales
Boredom V isual
fatigue
M alaise M uscu lar
fatigue
B oredom 0.65** 0.48** 0.38** 0.52**
V isual fatigue 0.32** 0.60** 0.51** 0 .6 6 **
M alaise 0.35** 0.47** 0.73** 0.62**
M uscular fatigue 0.49** 0.55** 0.58** 0.60**
**p < . 0 1
Table 4 gives the corre la tions betw een post-d rive fatigue scales and the sta te scale scores. T he
corre la tions show that there is a substantial overlap betw een constructs. T he fa tigue scales all
- 6 2 4 -
show sim ilar re la tionsh ips w ith energetic arousal, hedonic tone and anger/fru stra tion . T ense
arousal re lates to v isual fatigue, m alaise and m uscu la r fatigue but no t boredom . T he pattern o f
corre la tions betw een the fa tigue scales are also sim ilar fo r ta sk -re lev an t and task -irre levan t
cognitive in terference, m otivation , and perceived control.
T able 4 : C orrelations betw een post-drive fatigue sca les and general state m easures
B oredom V isual
fatigue
M alaise M uscu lar
fatigue
E nergetic arousal -0.57** -0.42** -0.34** -0.50**
T ense arousal 0.08 0.31** 0.23** 0.33**
H edonic tone -0.57** -0.44** -0.46** -0.51**
A nger/frustration 0.51** 0.39** 0.39** 0.44**
T ask-re levan t in terference 0 .2 1 ** 0.32** 0.25** 0.34**
T ask-irrelevan t in terference 0.44** 0.28** 0.36** 0.41**
P erceived contro l -0.29** -0.34** -0.27** -0.34**
M otivation -0.63** -0.17** -0.27** -0.26**
**p < . 0 1
C orrelations w ere calcu lated betw een the fatigue scales and E PQ , D B I and Fatigue P roneness
scales in o rder to de term ine w hether fa tigue re la tes to general personality tra its o r to driv ing-
specific traits (see T able 5).
T ab le 5 : C o rre la tion s betw een fa tigu e sca les an d E P Q , D B I and F a tig u e P ron en ess
scales.
B oredom V isual fatigue M alaise M uscu lar fatigue
Pre Post Pre Post Pre Post P re Post
E -0.04 0 . 0 1 0 . 0 0 0.05 0 . 0 2 0 . 0 2 -0 .04 -0.03
N 0 . 0 0 0.06 0.13* 0 .2 2 ** 0.08 0.13* 0 . 1 2 * 0.17**
P 0.08 0:09 -0.05 -0 . 0 0 -0.07 0 . 0 1 -0.09 -0.06
A G G R ESS 0 .2 2 ** 0 .2 0 ** 0.07 0.08 0.07 0.05 0 .2 1 ** 0.13*
D ISLIK E 0 . 1 2 * 0.14* 0 .2 1 ** 0.19** 0.17** 0.15** 0.16** 0 .2 1 **
A LER T -0.17** -0.15* -0.06 -0.06 -0 .2 1 ** 0.16* -0 .2 2 ** -0.08
F A TIG U E 0.17** 0.26** 0 . 1 2 0.25** 0.13* 0 .2 2 ** 0.16** 0.24**
*p<.05, **p<.01
N w as re lated to post-d riv e v isua l fatigue, m alaise , and m uscu la r fatigue. F atigue P roneness
w as the strongest single p red ic to r o f post-d rive fatigue as predicted.
- 6 2 5 -
C O N C L U S I O N S
The data suggest that the fo u r driv ing fatigue scales m ay p rov ide a re liab le and valid basis fo r
charac te riz in g sta te ch anges induced by the dem an d s o f d riv ing , and , perh ap s, by ex ternal
agen ts such as d rugs. T he d im en sio n s d isc rim in a ted re sem b le tho se o b ta in ed in p rev io u s
q u e s tio n n a ire s tu d ie s , b u t th ey h av e a so u n d p sy c h o m e tric b as is . T h ey a re in te rn a lly
consisten t, sensitive to an experim enta l m anipu la tion o f fa tigue, and m ean ing fu lly re la ted to
re lev an t personality d im ensions. A lthough the study w as one o f sim ula ted d riv ing , p o st-d rive
fa tig u e re la ted to su b je c ts ’ rep o rts o f fa tig u e reac tio n s to rea l d riv in g , m easu red by the
F atigue P roneness scale. T he co rre la tions betw een the fa tig u e sca les and the sta te m easu res
show that the re is co n sid erab le overlap b etw een su b jec tiv e fa tigue and v ario u s a sp ec ts o f
stress, and care m ay be req u ired in d is tin g u ish in g fa tig u e an d stre ss e ffe c ts in em p irica l
studies.
R E F E R E N C E S
E ysenck , H . J. & E ysenck , S. B. G. (1975). M anual o f the E ysenck P ersonality Q uestionnaire. London: H odder & S toughton.
Fuller, R. G . C. (1984). P ro longed driv ing in convoy: T he truck d r iv e r’s experience . A cciden t A nalysis & P revention , 16, 371-382.
G ulian , E ., M atthew s, G ., G lendon, A. I., D avies, D . R ., & D ebney , L. M . (1989). D im ensions o f driver stress. E rgonom ics, 32, 585-602.
M ackie , R. R ., O ’H an lo n , J. F . & M cA uley , M . E. (1974). A S tudy o f H ea t, N o ise and V ibration in R ela tion to D river Perform ance and P hysio log ica l S tatus. R eport 1735, H um an Factors R esearch Incorporated , Santa B arbara R esearch Park, G ole ta , C alifornia .
M ascord , D. J., W alls, J. & S tarm er, G . A . (1995). Fatigue and alcoho l: in terac tive e ffec ts on hum an perform ance in driv ing-related tasks. In L. H artley (Ed.), F atigue and D riv ing : D riv ing Im pairm ent, Fatigue and D riv ing S im ulation ( p p .). T ay lor & F rancis.
M atthew s, G ., Jones, D. M ., & C h am berla in , A. G. (1990). R e fin in g the m easu rem en t o f m ood: T he U W IS T M ood A djective C hecklist. B ritish Journal o f P sycho logy , 81, 17-42.
M atthew s, G ., C am pbell, S., Joyner, L ., G illiland , K ., H ugg ins, J ., & F alconer, S. V alidation o f a com prehensive stress sta te questionnaire: T ow ards a sta te ‘B ig T h re e ’? P ap er p resen ted to the E ighth E uropean C onference on Personality , G hent, B e lg ium , Ju ly 1996.
Saito, Y ., K ogi, K. & K ash iw agi, S. (1970). F actors u nderly ing sub jec tive fee ling o f fatigue. Journal o f S cience o f L abour, 46, 205-224.
- 6 2 6 -