Vol. 8, Part II, pp. 69-78, 1969 . Life Sciences Printed in Great Britain .
Pergamon Press
TIME ESTIMATION RELATED TO PERSONALITY, COGNITIVE SPEED AND SCHIZOPHRENIA
Anne Broadhurst Department of Psychiatry, University of Birmingham
(Received 3 October 1968; in final form 6 November 1968) It is now established that schizophrenics suffer disorders of subjective time (1,2) and accuracy of time perception has recently been suggested (3) as an index of reality testing in which these patients also show deficit .
Time
perception in normals has been much investigated, too, in connection with e .g . phenomena of attention (4) and arousal (5), diurnal variability (6), hypnosis (7) and personality (8) .
Unfortunately, although Clausen's (9) attempt to
find merits and demerits of the standard methods (verbal estimates, reproduction and operant estimates) is exemplary, the confusion of varying methodologies in time-estimation studies (10) has left many questions yet unsolved .
This
paper reports on time estimation (derived from time filled by
motor tasks) and its personality and cognitive relationships which have as yet received little attention .
Opportunity is taken to explore the relation
ship between time estimation and the cognitive slowness in schizophrenia which several authors have convincingly demonstrated (11,12,13) . Method Subjects*
The normal subjects were 24 male nurses of All Saints' Hospital,
Birmingham, who volunteered for experimental work within their duty hours . Their mean age was 33 .5 + (S .D .) 9 .06 years and none had been known to suffer from psychiatric disorder . Long-stay male chronic schizophrenic patients from the same hospital, who had been taken off all drugs at least one week prior to this investigation,
69
70
TIME ESTIMATION
Vol . 8, No . 2
constituted the pool of 43 subjects from which the experimental . subjects were drawn .
All subjects attemnted the tasks, but 19 were discarded because
they failed to follow the inrtructionç or .,aye incorrlete records (particularly in the motor tasks not reported on in detail here) . remaining 24 schizophrenics had a mean age of 39 .6 + l? .7 significantly different (by t-test) from the normal's age .
The
years .+hich is not Their mean
duration of present admission to hospital was 11 .7 + 9.2 years which gives an indication (slightly biassed towards apparent health by fre,uent and recent readmissions in a few cases) of chronicity of the disorder . Schizophrenia was diagnosed by the consultant in char,e of the ratient and in no case was the diagnosis a matter of doubt or disaireement between psychiatrists . Procedure .
In the course of an investigation of motor ,erforrrance (tr he
reported elsewhere) subjects were asked to tap re,)eatedly with ::. metal dylus on a metal plate for a period of five minutes and attemnted a standard pursuit-rotor tracking task
(following an eccentric soot rotating at I speed
of 60 r .p .m .) for a period of five minutes . another almost immediately.
The work periods followed one
Half of the subjects, randomly aspirned, in each
group were given the tapping; task first, half had the ìnin uit-rotor task . first.
The tasks,
chosen for their simplicity and for their difference in
learning involvement (tapping is an already overlemrned skill,
r»rsuit rotor
success requires practice to learn), were demonctratrd rnd subjects were asked to do the same but given no indication of the iiiration of the work period . Immediately after the signal to stop work on each tas:; the subjects were asked "How long have you been ta ming/following the snot?"
Answers ,.ere
recorded and converted to minutes and decimal fractions of minutes .
Time
estinu.,tion was therefore measured by the method of verbal estim= , tes rath-r than by reproduction .
The former method, de=-ite well-known disadvantages (o),
Vol. 8, No. 2
TIME ESTIMATION
71
e .g . of rounding of responses, was adopted as being more suited to long periods, to filled periods and to the deteriorated patient group in question . On the same day or the following day all subjects attempted the Eysenck Personality Inventory (EPI), Form A (î4), giving scores on extraversion (E), Neuroticism (N) and a Lie scale (L) .
Schizophrenics were encouraged to
cooperate in this--if necessary having the questions read out to them but avoiding interference with or manipulation of the subjects' responses . Similarly, an intelligence test--the Progressive Matrices Test (15)--was administered individually to all subjects .
Administration was standard for
all normal subjects with an additional measure taken of speed of cognitive functioning .
As each problem solution was written (by the subject) the
experimenter unostentatiously noted times on a 60" stop-watch .
At the finish,
calculations of solution times were made and adjusted to eliminate writing time and time to turn pages .
By excluding incorrect solutions (whose time
confounds continuance and error (16) ) and by using only solutions to Matrices Set A (at a difficulty level which enabled all subjects to solve correctly at least seven problems) an -iverage solution time in seconds was round for each subject .
Administration A.nd scoring for schizophrenics was
identical tc normals' as far as possible .
Results were obtained from some
apathetic or resistant :ubjects after assistance was given by page turning and by recording the solutions spoken, or indicated by pointing . Results and Discussion Table 1 displays the results without differentiating between those s ;ah.;ects of each group who were administered the tapping task prior to the pursuit-rotor task and vice versa , since analysis of variance of the time estimation scores showed no significant effect of order of administration . Likewise, this analysis showed no significant effect of the difference between the two tasks--tapping and pursuit rotor--when estimates are made of the amount of time spent upon each of them . Results from normal subjects for intelligence, E, N and L scores are
72
TIME ESTIMATION
Vol . 8, No. 2
TABLE 1 Results from Time Estimation, Cognitive and Personality Measures, Schizophrenics n = 24
Nmrmals n = 24
t
P
N.S .
M
S .D .
M
S .D .
39 "58
12 .09
33 "54
9.96
7. .926
Estimate of 5 minutes tapping (minutes)
7.76
4.40
4.55
2 .14
3.380 <0 .01
Estimate of 5 minutes pursuit rotor (minutes)
6.88
4.38
3.44
2 .10
3.534 t0 .01
Intelligence quotient (Matrices)
83 .21
10 .40
98 .21
12 .91
Cognitive speed Average, in seconds (Matrices, Set A)
1311
8.92
11 .99
5 .73
0.511
N.S .
Neuroticism score (E .P .I .--N)
11 .16
6.52
7 .16
5.21
1 .849
N.S .
Extraversion score (E .P .I .--E)
11 .29
2 .79
11 .58
3.16
0.318
N.S .
5 .12
2.15
2 .75
1 .67
4 .189 ,C0.001
11 .75
9 .16
Age (years)
Lie score (E.P .I .--L) Length of hospitalization (years)
.4824C0 4 .001
consistent with the normative'data . It can be seen from Table 1 that, while normals underestimate slightly, schizophrenics overestimate the expired time interval .
The difference
between the two groups gives the major significant finding of the analysis of variance
(7! = 14 .45, d .f ., 1,44, PC0 .01), in the direction of greater
estimates of elapsed time at both tasks by schizophrenics than by nornals . Phis finding from the analysis of variance based on the tasks combined is confirmed in separate analyses by .1-test of the estimates on the two tasks (Table 1) . The significant differences between the two groups in intelligence and
Vol. 8, No. a
TIME ESTIMATION
FPI Lie score (Table 1) is also noteworthy . consequences of institutionalisation .
7s
Both of these findings way be
No relationship can be found, however,
between time estimation and intelligence (see below) so that this failure of group matching is of small consequence for this experiment .
However, the
high Lie score found among schizophrenics throws doubt on the reliability of other personality measures in this group . rroduct moment correlations were calculated between time estimatione and other measures within the normal and schizophrenic groups separately . are rresented in Table 2 .
They
Although it could be argued that, for example,
extraversion is a dimension of personality unrelated to--possibly orthogonal to--the disorder schizophrenia, the results from the two groups, normal and schizophrenic, are here differentiated because of the doubtful reliability of personality and cognitive measures obtained from such chronic and deteriorated patients . The question must be asked ; on what does the ability to give correct verbal estimates of elapsed time depend?
It may be that biochemical evidence
relating to the 'biolorical cle)ck' (17) will eventually explain the comparative concordance of subjective time passage and geophysical time passage in normals and the discordrnce of these two in echizonhrenia . this evidence we must consider the available psychological data .
Lacking
It is clear
that estimation of elapsed time is not a function of the subject's age (i .e . his total life experience of elapsed time) or intelligence since correlations found between these variables and estimations (Table 2) are uniformly non-significant .
On the other hand, the significant correlations
within both normal and schizophrenic grc,ups between estimates of time at two rather different tasks suggest a considerable degree of individual consistency in time estimation .
These confirmations of earlier findings
(18,19) serve to show that our samples exhibit characteristics of groups previously investigated and other results-from this source can be assumed to have some generality also .
74
TIME ESTIMATION
Vol . 8, No. 2
TABLE 2 Product moment Correlations between Time Estimations, Cognitive and Personality Measures . Schizophrenics Time Estimation
Normals
n = 24
a = 24
Age vs .
pursuit rotor tapping
0 .22
0.36
0 .22 0 .20
Intelligence vs .
pursuit rotor tapping
-0 .02 0 .21
-0 .01 0 .22
Neuroticism Vs .
pursuit rotor tapping
0.36
-0 .32
Extraversion vs .
pursuit rotor tapping
0 .34 0 .12
0.39'
Cognitive speed vs .
pursuit rotor tapping
0 .12 0 .07
0 .45' 0 .41'
0 .580*
0.69"'
Tapping estimation vs . pursuit rotor estimation Hospitalization vs .
pursuit rotor tapping
Hospitalization vs . lie score
0 .41'
-0.37 0 .06
0 .14 0 .34 0 .51**
' significant at the 0 .05 level significant at the 0 .01 level "' significant at the 0 .001 level
Among schizophrenics there is no relationship found here }-etween accuracy of estimate and degree of illness measured by length of hospitalization . However, the high Lie score (EPI) arl the significant correlation between duration of hospitalization and the Lie score in the group confines the general finding (20) of raised L in psychiatric ~-rcups and su ; -gests that L may be, not only a function of disorder, but also of degree of disorder insofar as duration of stay in hospital can also be reRare.ed as a measure of severity of disorder . Psychotic patients frequently yield high scores on measures of N-neuroticism (21)--the only personality measure relating- directly to
Vol . 8, No. 2
TIME ESTIMATION
psychiatric disorder that has been assessed here .
75
In this connection we find
that neuroticism correlates positively with the time estimations by schizophrenics--in one case significantly (P40 .05) and in the other bordering on significance--which means that the higher the neuroticism of the individual, the greater the estimate of elapsed time .
Though results in the
normal group are statistically non-significant, it is interesting that the trend there is distinctly in the other direction, the greater the stabilit; of the individual, the greater the time estimate .
There is perhaps greater
reliance to be placed on this finding than on the positive correlation of neuroticism and time estimation in schizophrenia (with high L score) but the failure to reach significance renders interpretation unnecessary. Turning to the EPI extraversion score (E), it is clear from Eysenck'e (22) typological postulate that extraverts will overestimate elapsed time while introverts may underestimate it .
Hence a positive correlation is
anticipated between extraversion and time estimations .
Once again, it must
be remembered that the EPI scores from the chronic schizophrenic subjects are of doubtful reliability.
(High L scores, as found here and characteristic of
institutionalised patients, contributed to stimulate the development of the more appropriate Eysenck-Withers Personality Inventory (23)--too late for inclusion in this experiment) .
Nevertheless a positive correlation was
uniformly found, but only in the case of normals estimating pursuit-rotor time does this reach significance .
The task difference for correlations with
extraversion--greater correlations for pursuit rotor than for tapping for both groups--suggests that the learning task has a different inhibitory potential than the simple répetitive motor tapping and that this type of inhibition more markedly affects the extravert causing him to extend his time estimates on the job . Cognitive speed has not previously been assessed in relation to subjective time though Orme (18) comments anecdotally that "large verbal time
76
Vol . 8, No. 2
TIME ESTIMATION
estimators tend to be . . . quick" .
The measure here taken (average time to
solve Progressive Matrices problems, Set A) is not the most satisfactory measure of intellectual speed (16) although we have avoided the obvious pitfalls of confounding speed and power .
It gives us a greater (slower), but
not significantly greater, score for schizophrenics than for normals, consistent with, though not resoundingly confirming previous findings
(13) .
pi ng the findings which relate cognitive speed to time estimation (Table 2) we see consistently positive correlations--contrary to Orme's observation.
The person whose cognitive functioning makes him take longer to
solve a problem is the same person who reports that elapsed time is greater. Knowledge of cognitive slowness in schizophrenia (12 V24), and of time overestimation in schizophrenia would predicate an association (correlation) between time distortion and cognitive slowness in schizophrenia.
The
observation that this correlation is significant among the normals,
(not so
among the schizophrenics), indicates that we are here dealing with a phenomenon of wider generality, not limited to the disorder of schizophrenia . In fact this group difference now cuts across and overshadows the task difference mentioned above.
Clearly, cognitive speed is worthy of closer
study in relation to time estimations in normal and abnormal groups . Conclusion Verbal estimates of filled (motor task) periods of five minutes are unrelated to age, intelligence or the task .
Schizophrenics overestimate
significantly (as c(impared with normals who underestimate slightly), and since Falk and Bindra (25) found stress brings about time overestimation, this may be attributed to the intrusion of internal stimuli causing subjective stress . The prediction of overestimation with extraversion is supported, particulaf in normals who give the more reliable personality scores .
The
relationship between time estimation and neuroticism is unclear, but, normal subjects show a marked link between large verbal time estimation and slowness of problem solving .
Vol . 8, No . 2
TM ESTIMATION
77
summary Twenty-four normal and twenty-four chronic schizophrenic male subjects estimated duration of elapsed time at two simple motor tasks .
Personality
and cognitive correlates of these estimates were calculated .
Cognitive
speed was related to time estimation only in the normal group.
As predicted
from Eysenck's (22) work, there is a positive relationship between time estimation and extraversion . Acknowledgements The assistance of Dr . N .W. Imlah, Medical Director at All Saints' Hospital, Birmintham is gratefully acknowledged .
With his active
encouragement, his staff provided volunteer normal subjects and facilitated the work with patients .
Thanks are particularly due to T.A . Blythin and M.J .
Clarke whose administrative assistance was invaluable . The assistance of D.W . Fullmr in statistical analyses is also gratefully acknowledged . References
83, 530 (1936) "
P. Schilder, J. nerv . ment . Dis.
2.
J.E . Orme, Acts psychd. 20,
285 (1962) .
3.
F.T . Melgesl C.E . Fougerousse, Jr ., J. Psychiat . Res. 4,
4.
M. Sturt, Brit . J. Psychol .
5.
R .C .H . Aitken and J.L . Gedye, Brit . J. Psychol .
6.
D.H . Thor, Percept. mot . Skills ,
7.
S. Fogel and A. Hoffer, J. clin . exp. Psychopathol .
8.
J.E . Orme, Acta Psychol. 22,
430 (1964) .
9.
J. Clausen, J. exn . Psychol .
40, 756 (1950) .
127 (1966) .
D, 382 (1923) . 22, 253 (1968) .
U, 451 (1962) .
10 .
f. Wallace and A.I . Rabin, Psychol . Bull .
11 .
Harriet Babcock, Dementia Precox . Co ., New York (193
,
24 (1962) .
2Z, 213 (1960) .
A psychological study.
Science Press
14-B . Shapiro and Elizabeth H. Nelson, Brit . J. mad . Psychol.
(1955) .
28, 239
78
13 .
TM ESTIMATION
Vol. 8, No. 2
H .J . Eysenck, G .W . Granger and J .C . Brengelmann, Perceptual processes and mental illness . Maudsle Monographs . No .2 . Chapman 8 . Hall, London
(1957)-
14 .
H .J . Eysenck and Sybil B .G . Eysenck, Manual of the E senck Personality Inventory . IIniver . London Press, London 19
15 .
J .C . Raven, Guide to the Standard Pro . essive 11atrices Sets A BCD E . IT .K . Lewis, London (1960) .
16 .
W .D . Furneaux, Handbook of abnormal behaviour p .
17.
G .R . !Iawkes, R .J .T . Joy and W .D . Evnns, J . Psychol .
18 .
J .E . Orme, Brit . J . Psychiat . 112,
19.
A .I . Rabin, J . clin . Psychol .
20.
Sybil B .G . Eysenck and I? .J . Eysenck, Life Sciences , `,
21 .
H .J . Eysenck, Handbook of abnormal behaviour p . 1 . Pitman, Lonc'on (1960) .
22 .
H .J . Eysenck, The dynamics of anxiety and hysteria . Paul, London (1957) "
23 .
Sybil B .G . Eysenck, Manual of the E senck-Withers Personnlity Tnventor . for Subnormal . subjects 50-80 I .Q . . UnivP^ . London Press, London 19
24 .
Anne Broadhurst, J . rent . Sci . 104, 1123 (1Y;P) .
25 .
J .L . Falk and D . Bindra, J . exp . Psychol .
(1950) .
167 .
and
Pitman, London
22, 183 (1962) .
37 (1966) .
13, 88 (1957) . 343 (19`3) .
iioutlete & Eegan
47, 279 (19`:4) .