The relationship between emotional intelligence and
James D.A. Parker
*, Graeme J. Taylor
, R. Michael Bagby
Department of Psychology, Trent University, Peterborough, Ontario, Canada, K9J 7B8
Department of Psychiatry, University of Toronto and Mount Sinai Hospital, Toronto, Canada
Department of Psychiatry, University of Toronto, and Centre for Addiction and Mental Health, Clarke Division,
Received 24 June 1999; received in revised form 18 November 1999; accepted 24 January 2000
In this study, the empirical association between the apparently similar constructs of emotional intelli-
gence and alexithymia was examined using latent variable analysis in a large community sample of adults
(N=734). The Twenty-Item Toronto Alexithymia Scale (TAS-20) and the BarOn Emotional Quotient
Inventory (EQ-i) were used to assess alexithymia and emotional intelligence. Results revealed that although
the constructs are independent, they overlap considerably and are strongly and inversely related. #2000
Elsevier Science Ltd. All rights reserved.
Keywords: Alexithymia; Emotional intelligence; Structural equation modelling
Introduced by Salovey and Mayer (1989/1990) a decade ago, the construct of emotional intel-
ligence is generating considerable public interest, mostly due to the popularization of the concept
by Goleman (1995, 1998). A conceptually similar construct is alexithymia. Although less well
known in the popular press, the alexithymia construct was introduced almost 30 years ago
(Nemiah and Sifneos, 1970; Nemiah, Freyberger & Sifneos, 1976) and has generated a far greater
amount of empirical research than has the emotional intelligence construct (Taylor, Bagby &
Parker, 1997). A comparison of the de®nitions of emotional intelligence and alexithymia suggests
that the two constructs are closely related.
0191-8869/00/$ - see front matter #2000 Elsevier Science Ltd. All rights reserved.
Personality and Individual Differences 30 (2001) 107±115
* Corresponding author. Tel.: +1-705-748-1283; fax: +1-705-748-1580.
E-mail address: firstname.lastname@example.org (J.D.A. Parker).
Derived from the broader construct of social intelligence, emotional intelligence was de®ned
originally by Salovey and Mayer (1989/1990) as ``the ability to monitor one's own and others'
feelings and emotions, to discriminate among them and to use this information to guide one's
thinking and actions'' (p. 189). This de®nition encompasses two subtypes of personal intelligence
described by Gardner (1983) Ð intrapersonal intelligence (the ability to access one's own feeling
life) and interpersonal intelligence (the ability to read the moods, intentions, and desires of oth-
ers). Bar-On (1996, 1997), who has worked extensively on developing a comprehensive inventory
for assessing emotional intelligence since 1980, employs a much broader de®nition of the con-
struct, which includes adaptive capacities and abilities to control impulses and cope with stress, as
well as intrapersonal and interpersonal intelligence.
Whereas the emotional intelligence construct emerged from an integration of an array of
research ®ndings on how people appraise, communicate, and use emotion (Salovey & Mayer,
1989/1990), the alexithymia construct, formulated by Nemiah et al. (1976), arose primarily from
a cluster of cognitive and aective characteristics that had been observed in clinical situa-
tions, initially among patients with so-called `classical' psychosomatic diseases (Nemiah &
Sifneos, 1970; Ruesch, 1948), and later among patients with substance use disorders (Krystal
& Raskin, 1970), posttraumatic stress disorders (Krystal, 1968), and eating disorders (Bruch,
The salient features of the alexithymia construct include diculties in identifying and describ-
ing subjective feelings, a limited imaginal capacity, and an externally oriented style of thinking
(Taylor, Bagby & Parker, 1991, 1997). Although diculty in monitoring the feelings and emo-
tions of other people is not part of the de®nition, empirical studies have shown that individuals
with high degrees of alexithymia experience diculties in accurately identifying emotions in the
facial expressions of others (Lane et al., 1996; Parker, Taylor & Bagby, 1993b); and clinicians
report that alexithymic individuals manifest a limited capacity for empathizing with the emo-
tional states of others (Krystal, 1979; McDougall, 1989; Taylor, 1987). In addition, there is
empirical evidence that alexithymia is associated with diculties in discriminating among dier-
ent emotional states (Bagby, Parker, Taylor & Acklin, 1993), and with a limited ability to think
about and use emotions to cope with stressful situations (Parker, Taylor & Bagby, 1998; Schaer,
Although Salovey and Mayer (1989/1990; Salovey, Hsee, and Mayer, 1993) acknowledge an
overlap of the emotional intelligence and alexithymia constructs, they have made little attempt to
evaluate empirically the relationships between the two constructs. One possible explanation for
this is that these investigators have yet to introduce a standardised method for assessing emo-
tional intelligence. Salovey and Mayer's (1989/1990) de®nition of the construct was oper-
ationalized recently, however, by Schutte et al. (1998), who developed and validated a 33-item
self-report scale. In a mixed, but rather small, university student and community sample
(N25), this scale correlated strongly and negatively (rÿ0:65) with the 26-item Toronto
Alexithymia Scale (TAS; Taylor, Ryan & Bagby, 1985).
In a more recent study, Davies, Stankov and Roberts (1998) used the three factor scales of the
Twenty-Item Toronto Alexithymia Scale (TAS-20; Bagby, Parker & Taylor, 1994a) to assess the
appraisal and expression of emotions in the self and the recognition of emotions in others. In a
second-order factor analysis, which included a number of dierent scales related to emotional
intelligence, the TAS-20 factor scales loaded signi®cantly on factors pertaining to `emotional
108 J.D.A. Parker et al. / Personality and Individual Dierences 30 (2001) 107±115
clarity' and/or `emotional awareness'. In addition, one of the TAS-20 factors (externally-oriented
thinking) correlated negatively with Mehrabian and Epstein's (1970) questionnaire measure of
The aim of the present study was to examine further the relationship between the alexithymia
construct, as measured by the TAS-20, and Bar-On's (1996, 1997) model of emotional intelli-
gence, as measured by the EQ-i, using latent variable analyses. We predicted that the TAS-20
total score and the scores for each of its three factors would be independent of, but strongly and
inversely associated with, the total score of the EQ-i.
The sample was comprised of 734 adults (329 men and 405 women) residing in several small
cities and towns in central Ontario, Canada. The participants were invited to take part in the
study by means of advertisements posted in the local community. The mean age of the sample
was 32.53 years (SD 13:98). The sample had a mean level of education of 13.58 years
(SD=1.78). Eighty-six percent of the participants identi®ed themselves as White, 4% as Black,
4% as Asian, 2% as Native American, and 4% did not indicate their race. All participants were
informed that they would be asked to complete two questionnaires for a research study on emo-
tions and personality. The questionnaires were completed by the research participants in their
own homes under the supervision of one of several research assistants.
The 20-Item Toronto alexithymia scale (TAS-20) is a widely used self-report measure of alex-
ithymia with a ®ve-point Likert rating format (Bagby et al., 1994a; Taylor et al., 1997). The TAS-
20 consists of three factors: diculty identifying feelings (Factor 1), diculty describing feelings
(Factor 2), and externally-oriented thinking (Factor 3) (Bagby et al., 1994; Parker, Bagby, Tay-
lor, Endler & Schmitz, 1993a; Taylor et al., 1997). The TAS-20 has demonstrated convergent and
discriminant validity, and scores show high agreement with observer ratings of alexithymia
(Bagby, Taylor & Parker, 1994).
The BarOn Emotional Quotient Inventory (EQ-i; Bar-On, 1997) is a 133-item inventory with
13 subscales that cluster into four second-order factors Ð intrapersonal (emotional self-aware-
ness, assertiveness, self-regard, self-actualization, independence), interpersonal (empathy, rela-
tionship skills, social responsibility), adaptability (problem solving, reality testing, ¯exibility), and
stress management (stress tolerance, impulse control) Ð and two subscales that assess general
mood (happiness and optimism). The EQ-i also contains two validity subscales (positive impres-
sion and negative impression). As the purpose of this study was to examine only the core features
of emotional intelligence, only the four second-order factors that assess the central facets in Bar-
On's model of emotional intelligence were used in the analyses. The reliability and construct
validity of the inventory have been demonstrated with samples from several dierent countries
(Bar-On, 1997). Con®rmatory factor analysis of the EQ-i has indicated that the four second-order
J.D.A. Parker et al. / Personality and Individual Dierences 30 (2001) 107±115 109
factors are highly correlated and best represented as a unitary construct. As such, all model test-
ing in the present study used a single scale score to assess and represent the emotional intelligence
Means and standard deviations for the TAS-20 and its three factors and the EQ-i and its four
second-order scales are presented in Table 1. The men scored higher than the women on the TAS-
20 total scale [t7323:99, p<0:001], the diculty describing feelings factor [t7324:15,
p<0:001], and the externally-oriented thinking factor [t7324:91, p<0:001]. Women scored
higher than men on the total EQ-i [t7323:52, p<0:001] and on the Intrapersonal scale of the
EQ-i [t73212:18, p<0:001]. Although the dierences between men and women on these
scales were statistically signi®cant, this signi®cance is likely attributable to the large sample size.
As a further check on combining male and female data together, the covariance matrices (four
EQ- i scales and the 20 items from the TAS-20) for men and women were compared using Sta-
tistica 5.1 (StatSoft, 1995). The criteria for equivalence was a Steiger±Lind RMSEA Index
(RMSEA; Steiger & Browne, 1984) less than 0.10, a Population Gamma Index (PGI; Tanaka &
Huba, 1985) greater than 0.90, and an Adjusted Population Index (APGI; StatSoft, 1995) greater
than 0.80. The RMSEA was 0.031, the PGI was 0.988, and the APGI was 0.975. The magnitude
of the three indexes suggests that the covariance matrices were virtually identical for men and
women. The internal reliabilities (Cronbach's alpha) for the measures are also displayed in Table
1. All these coecients are in the acceptable range.
The relationships between the TAS-20 and its three factors and the EQ-i and its four second-
order factors were examined using Pearson product moment correlations. As shown in Table 2,
Means, standard deviations, and internal reliability coecients for the TAS-20 and the EQ-i for men (N329) and
Scale Men Women
Mean SD Alpha Mean SD Alpha
DIF 13.66 5.84 0.85 13.50 5.42 0.80
DDS 13.12 4.34 0.76 11.79 4.28 0.76
EOT 20.37 5.07 0.66 18.41 4.91 0.67
Total score 47.15 11.98 0.85 43.70 11.21 0.83
Intrapersonal 149.02 20.53 0.93 146.55 20.41 0.93
Interpersonal 109.86 14.83 0.86 121.61 11.20 0.82
Adaptability 92.17 12.99 0.86 93.67 11.98 0.85
Stress management 64.23 10.31 0.84 63.63 9.95 0.83
Total score 413.28 49.43 0.90 425.46 43.80 0.88
TAS-20 is the Twenty-Item Toronto Alexithymia Scale; DIF=diculty identifying feelings; DDF=diculty
describing feelings; EOT=externally oriented thinking; EQ-i is the BarOn Emotionl Quotient Inventory.
110 J.D.A. Parker et al. / Personality and Individual Dierences 30 (2001) 107±115
the TAS-20 and its three factors correlated signi®cantly with the total score and the four factors
of the EQ-i. Lower levels of emotional intelligence and its four components were associated with
higher levels of alexithymia and its salient facets. The magnitude of these correlations are uni-
formly moderate to large in magnitude, suggesting a signi®cant amount of overlap.
In order to determine if alexithymia and emotional intelligence are independent constructs, we
tested a series of two-factor models with alexithymia hypothesized as one latent factor and emo-
tional intelligence hypothesized as the other latent factor. As some investigators (e.g., Deary,
Scott & Wilson, 1997) have suggested that the three TAS-20 factors, as a whole, may not best
represent the alexithymia construct, we tested separate two-factor models for each of the three
facets of the alexithymia construct (using congruent items from the TAS-20), as well as for total
alexithymia (using the three subscales from the TAS-20). Emotional intelligence was estimated
using the four second-order factor scales from the EQ-i (intrapersonal, interpersonal, stress
management, and adaptability).
The various models were tested using con®rmatory factor analysis with Statistica 5.1 (StatSoft,
1995). Following the recommendation of Cole (1987), and Marsh, Balla and McDonald (1988),
the goodness-of-®t of the various models was evaluated using multiple criteria: the goodness-of-®t
index (GFI; Jo
Èreskog & Sorbom, 1986), the adjusted GFI (AGFI; Jo
Èreskog & Sorbom, 1986), the
Non-Normed Fit Index (NNFI; Bentler, 1990), the Comparative Fit Index (CFI; Bentler, 1990),
and the Root Mean Square Residual (RMS; Bentler, 1990). Multiple criteria were used because
each index has dierent strengths and weaknesses in assessing the goodness-of-®t between a
hypothetical model and the actual data (Cole, 1987). Based on recommendations of Anderson
and Gerbing (1984), Bentler (1992), Cole (1987), and Marsh et al. (1988), the following criteria
were used to indicate the goodness-of-®t of the model to the data: GFI>0.850; AGFI>0.800,
NNFI>0.900, CFI>0.900, and RMS<0.10.
The ®rst model examined the relationship between emotional intelligence (estimated by the
intrapersonal, interpersonal, adaptability, and stress management scales from the EQ-i) and the
latent alexithymia variable of `diculty identifying feelings' (estimated from items 1, 3, 6, 7, 9, 13,
and 14 on the TAS-20). The GFI (0.919), AGFI (0.875), NNFI (0.900), CFI (0.911), and RMS
(0.052) all met the criteria standards for adequacy of ®t. The parameter estimate between the
emotional intelligence and diculty identifying feelings factors was ÿ0.78 (p<0:001). Given the
magnitude of this parameter estimate, we also examined whether the measured variables were
better represented by a one-factor model. The one- and two-factor models could then be com-
Correlations between the TAS-20 and EQ-i (N734)
Identify feelings Describe feelings Externally oriented thinking Total score
Intrapersonal ÿ0.58 ÿ0.62 ÿ0.35 ÿ0.66
Interpersonal ÿ0.38 ÿ0.47 ÿ0.41 ÿ0.54
Adaptability ÿ0.59 ÿ0.46 ÿ0.37 ÿ0.62
Stress management ÿ0.54 ÿ0.33 ÿ0.20 ÿ0.47
Total score ÿ0.64 ÿ0.61 ÿ0.42 ÿ0.72
All correlations are signi®cant (p<0:01).
J.D.A. Parker et al. / Personality and Individual Dierences 30 (2001) 107±115 111
pared by testing the dierence in the goodness-of-®t chi-squares (Breckler, 1990). The two-factor
solution was found to oer a better ®t to the data than the one-factor model [chi-square
difference 282:43 (df 1), p<0:001].
The second model examined the relationship between emotional intelligence (estimated by the
same four EQ-i scales used in the previous analysis) and the latent alexithymia variable of `di-
culty describing feelings' (estimated from items 2, 4, 11, 12, and 17 on the TAS-20). The GFI
(0.917), AGFI (0.857), NNFI (0.900), CFI (0.905), and RMS (0.053) all met the criteria standards
for adequacy of ®t. The parameter estimate between the emotional intelligence and diculty
describing feelings factors was ÿ0.70 (p<0:001). Given the magnitude of this parameter esti-
mate, one- and two-factor models for the measured variables were compared by testing the dif-
ference in the goodness-of-®t chi-squares. The two-factor solution was found to oer a better ®t
to the data than the one-factor model [chi-square difference 239:84 (df 1), p<0:001].
The third model examined the relationship between the latent variable of emotional intelligence
(estimated by the four EQ-i scales used in the previous analyses) and the latent alexithymia vari-
able of `externally-oriented thinking' (estimated from items 5, 8, 10, 15, 16, 18, 19, and 20 on the
TAS-20). The GFI (0.908), AGFI (0.865), NNFI (0.887), CFI (0.896), and RMS (0.066) all met
the criteria standards for adequacy of ®t. The parameter estimate between the emotional intelli-
gence and externally-oriented thinking factors was ÿ0.55 (p<0:001). One- and two-factor mod-
els for the measured variables were also compared by testing the dierence in the goodness-of-®t
chi-squares. The two-factor solution was found to oer a better ®t to the data than the one-factor
model [chi-square difference 337:77 (df 1), p<0:001].
The fourth model examined the relationship between the latent variable of alexithymia (esti-
mated using the three TAS-20 subscales of `diculty identifying feelings', `diculty describing
feelings', and `externally-oriented thinking') and emotional intelligence (estimated by the four
EQ-i scales used in the previous models). The GFI (0.967), AGFI (0.912), NNFI (0.956), CFI
(0.934), and RMS (0.067) all met the criteria standards for adequacy of ®t. The parameter esti-
mate between alexithymia and emotional intelligence was ÿ0.94 (p<0:001). Given the very high
parameter estimate between alexithymia and emotional intelligence, one- and two-factor models
for the measured variables were compared by testing the dierence in the goodness-of-®t chi-
squares. Consistent with the previous analyses, the two-factor model was found to oer a better
®t to the data than the one-factor model [chi-square difference 47:09 (df 1), p<0:001].
The results of this study con®rm our prediction that alexithymia and emotional intelligence are
independent but strongly correlated (inversely) constructs. While the parameter estimates and the
Pearson correlations examining the associations between the latent and nonlatent factors of the
TAS-20 and the total EQ-i score were strong, for each of the latent models tested a two-factor
solution was superior to a one-factor solution. The ®nding that each of the TAS-20 factors pro-
duced moderate to high parameter estimates with the EQ-i factor suggests that the various facets
of the alexithymia construct are related to emotional intelligence.
The ®nding of a strong negative correlation between the TAS-20 and the EQ-i is consistent
with Schutte et al.'s (1998) ®nding of a negative correlation between the TAS and their 33-item
112 J.D.A. Parker et al. / Personality and Individual Dierences 30 (2001) 107±115
emotional intelligence scale, and provides further evidence that alexithymia and emotional intel-
ligence are inverse but strongly overlapping constructs. Although Davies et al. (1998) questioned
the reliability of the TAS-20 and whether there is an adequate methodology for assessing the
emotional intelligence construct, the TAS-20 and EQ-i total scales and factor scales demonstrated
acceptable levels of internal consistency in this sample, as they have done in previous studies with
a variety of non-clinical populations (Bar-On, 1997; Bagby et al., 1994a; Parker et al., 1993a).
Moreover, Davies et al.'s (1998) conclusion about the reliability of the TAS-20 is compromised as
these investigators failed to reverse the scoring on the negatively keyed items on the scale (L.
Stankov, personal communication, November 16, 1999).
The total TAS-20 and its three factors correlated signi®cantly and negatively with the intrapersonal
and interpersonal factors of the EQ-i, which encompass the two personal intelligences described by
Gardner (1983) and the two central facets of emotional intelligence in Salovey and Mayer's (1989/
1990) de®nition (i.e., emotional self-awareness and empathy). These ®ndings are consistent with
clinical reports that high alexithymic individuals manifest a limited capacity for empathizing with the
emotional states of others (Krystal, 1979; McDougall, 1989; Taylor, 1987), and with ®ndings from
an earlier study (Taylor et al., 1997) in which all three factors of the TAS-20 correlated negatively
with the aective orientation scale (Booth-Butter®eld & Booth-Butter®eld, 1990); the latter scale
assesses the extent to which individuals are aware of and use emotions to guide communication.
In the present study, signi®cant negative correlations were obtained between the TAS-20 and
its three factors and the adaptability and stress management factors of the EQ-i. These ®ndings
are consistent with results from previous studies in which alexithymia was found to be associated
with maladaptive defense and coping styles (Parker et al., 1998), vulnerability to stress (Bagby et
al., 1994b), and psychiatric disorders and somatic illnesses that involve problems in the modula-
tion of distressing aects (Taylor et al., 1997).
Given that previous research has found alexithymia to be associated with both illness beha-
viour (Lumley, Tomakowsky & Torosian, 1997) and increased mortality from all causes (Kau-
hanen, Kaplan, Cohen, Julkunen & Salonen, 1996), the ®ndings from the present study raise the
possibility that high emotional intelligence might be a protective factor for mental and physical
health. Future studies using longitudinal and prospective designs might evaluate this possibility.
Meanwhile, the ®ndings from the present study should alert clinicians and researchers to recog-
nize that the presence of alexithymic characteristics in their patients suggest low emotional intel-
ligence, and that highly alexithymic individuals not only lack the capacity to use emotions to
guide their behaviour, but are also intolerant of stress and have limited adaptive resources.
Although alexithymia and emotional intelligence are inverse but strongly overlapping con-
structs, alexithymia is a more narrowly de®ned construct than Bar-On's (1997) broad con-
ceptualization of emotional intelligence. Indeed, as Mayer, Salovey and Caruso (2000, in press)
point out, the EQ-i is based on a model of emotional intelligence, which combines mental abilities
concerned with understanding emotions with a diverse set of personality characteristics that Bar-
On (1997) relates to the potential to succeed in life. For example, as noted earlier, the intra-
personal second-order factor of the EQ-i assesses not only emotional self-awareness, but also
assertiveness, self-regard, self-actualization, and independence. And in addition to assessing the
ability to empathize, the interpersonal factor of the EQ-i assesses relationship skills and social
responsibility. Although the TAS-20 may correlate with many of these characteristics, they are
not part of the de®nition of the alexithymia construct; nor are they included in the models of
J.D.A. Parker et al. / Personality and Individual Dierences 30 (2001) 107±115 113
emotional intelligence proposed by Salovey and Mayer (1989/1990) and Mayer et al. (2000, in
press). Future research might investigate the relationship between alexithymia and models of
emotional intelligence that restrict themselves to mental abilities concerning the awareness and
cognitive processing of emotion.
This study was supported by a research grant to the ®rst author (]410-95-0634) from the Social
Sciences and Humanities Research Council of Canada (SSHRC). The authors would like to
thank Reuven Bar-On for his comments and Dana Reker for her help with collecting the data
used in this study.
Anderson, J. C., & Gerbing, D. W. (1984). The eect of sampling error on convergence, improper solutions, and
goodness-of-®t indices for maximum likelihood con®rmatory factor analysis. Psychometrika, 49, 155±173.
Bagby, R. M., Parker, J. D. A., Taylor, G. J., & Acklin, M. W. (1993). Alexithymia and the ability to distinguish
dierent emotional states. Poster presentation at the Annual Meeting of the American Psychosomatic Society,
Charleston, SC, March, .
Bagby, R. M., Parker, J. D. A., & Taylor, G. J. (1994a). The Twenty-Item Toronto Alexithymia Scale Ð I. Item
selection and cross-validation of the factor structure. Journal of Psychosomatic Research, 38, 23±32.
Bagby, R. M., Taylor, G. J., & Parker, J. D. A. (1994b). The Twenty-Item Toronto Alexithymia Scale Ð II. Con-
vergent discriminant, and concurrent validity. Journal of Psychosomatic Research, 38, 33±40.
Bar-On, R. (1996). The era of the ``EQ'': De®ning and assessing emotional intelligence. The 104th Annual Convention
of the American Psychological Association, Toronto, .
Bar-On, R. (1997). BarOn emotional quotient inventory. Toronto: Multi-Health Systems Technical manual.
Bentler, P. M. (1990). Comparative ®t indexes in structural models. Psychological Bulletin, 107, 238±246.
Bentler, P. M. (1992). On the ®t of models to covariances and methodology to the Bulletin. Psychological Bulletin, 112,
Booth-Butter®eld, M., & Booth-Butter®eld, S. (1990). Conceptualizing aect as information in communication pro-
duction. Human Communication Research, 16, 451±476.
Breckler, S. J. (1990). Applications of co-variance structure modelling in psychology: Cause or concern? Psychological
Bulletin, 107, 260±273.
Bruch, H. (1973). Eating disorders. Obesity, anorexia nervosa, and the person within, Basic BooksNew York.
Cole, D. A. (1987). Utility of con®rmatory factor analysis in test validation research. Journal of Consulting and Clinical
Psychology, 55, 584±594.
Davies, M., Stankov, L., & Roberts, R. D. (1998). Emotional intelligence: In search of an elusive construct. Journal of
Personality and Social Psychology, 75, 989±1015.
Deary, I. J., Scott, S., & Wilson, J. A. (1997). Neuroticism, alexithymia, and medically unexplained symptoms. Per-
sonality and Individual Dierences, 22, 551±564.
Gardner, H. (1983). Frames of mind: The theory of multiple intelligences. New York: Basic Books.
Goleman, D. (1995). Emotional intelligence. New York: Bantam Books.
Goleman, D. (1998). Working with emotional intelligence. New York: Bantam Books.
Èreskog, K. G. & Sorbom, D. (1986). LISREL VI: Analysis of linear relationships by maximum likelihood, instrumental
variables, and least square methods. Mooresville, IN: Scienti®c Software.
Kauhanen, J., Kaplan, G. A., Cohen, R. D., Julkunen, J., & Salonen, J. T. (1996). Alexithymia and risk of death in
middle-aged men. Journal of Psychosomatic Research, 41, 541±549.
114 J.D.A. Parker et al. / Personality and Individual Dierences 30 (2001) 107±115
Krystal, H. (1968). Massive psychic trauma. New York: International Universities Press.
Krystal, H. (1979). Alexithymia and psychotherapy. American Journal of Psychotherapy, 33, 17±31.
Krystal, H. & Raskin, H. (1970). Drug dependence. Detroit: Wayne State University Press.
Lane, R., Sechrest, L., Reidel, R., Weldon, V., Kaszniak, A., & Schwartz, G. (1996). Impaired verbal and nonverbal
emotion recognition in alexithymia. Psychosomatic Medicine, 58, 203±210.
Lumley, M. A., Tomakowsky, J., & Torosian, T. (1997). the relationship of alexithymia to subjective and biomedical
measures of disease. Psychosomatics, 38, 497±502.
Marsh, H. W., Balla, J. R., & McDonald, R. P. (1988). Goodness-of-®t indexes in con®rmatory factor analysis: The
eect of sample size. Psychological Bulletin, 103, 391±410.
Mayer, J. D., Salovey, P., & Caruso, D. (2000). Competing models of emotional intelligence. In R. J. Sternberg,
Handbook of human intelligence (pp. ±). New York: Cambridge University Press.
McDougall, J. (1989). Theatres of the body: A psychoanalytic approach to psychosomatic illness. New York: Norton.
Mehrabian, A., & Epstein, N. (1970). A measure of emotional empathy. Journal of Personality, 40, 525±543.
Nemiah, J. C., Freyberger, H., & Sifneos, P. E. (1976). Alexithymia: A view of the psychosomatic process. In O. W.
Hill, Modern trends in psychosomatic medicineVol. 3 (pp. 430±439). London: Butterworths.
Nemiah, J. C., & Sifneos, P. E. (1970). Aect and fantasy in patients with psychosomatic disorders. In O. W. Hill,
Modern trends in psychosomatic medicineVol. 2 (pp. 26±34). London: Butterworths.
Parker, J. D. A., Bagby, R. M., Taylor, G. J., Endler, N. S., & Schmitz, P. (1993a). Factorial validity of the 20-item
Toronto Alexithymia Scale. European Journal of Personality, 7, 221±232.
Parker, J. D. A., Taylor, G. J., & Bagby, R. M. (1993b). Alexithymia and the recognition of facial expressions of
emotion. Psychotherapy and Psychosomatics, 59, 197±202.
Parker, J. D. A., Taylor, G. J., & Bagby, R. M. (1998). Alexithymia: Relationship with ego defense and coping styles.
Comprehensive Psychiatry, 39, 91±98.
Ruesch, J. (1948). The infantile personality. Psychosomatic Medicine, 10, 134±144.
Salovey, P., Hsee, C. K., & Mayer, J. D. (1993). Emotional intelligence and the self-regulation of aect.
In D. M. Wegner, & J. W. Pennebaker, Handbook of mental control (pp. 258±277). Englewood Clis, NJ: Prentice
Salovey, P., & Mayer, J. D. (1989/1990). Emotional intelligence. Imagination, Cognition, and Personality, 9, 185±211.
Schaer, C. E. (1993). The role of adult attachment in the experience and regulation of aect. Doctoral dissertation, Yale
Schutte, N. S., Malou, J. M., Hall, L. E., Haggerty, D. J., Cooper, J. T., Golden, C. J., & Dornheim, L. (1998).
Development and validation of a measure of emotional intelligence. Personality and Individual Dierences, 25, 167±
StatSoft (1995). Statistica for Windows. Tulsa, OK: StatSoft.
Steiger, J. H., & Browne, M. W. (1984). The comparison of independent correlations between optional linear compo-
sites. Psychometrika, 49, 11±24.
Tanaka, J. S., & Huba, G. J. (1985). A ®t index for covariance structure models under arbitrary GLS estimation.
British Journal of Mathematical and Statistical Psychology, 38, 197±201.
Taylor, G. J. (1987). Psychosomatic medicine and contemporary psychoanalysis. Madison, CT: International Uni-
Taylor, G. J., Bagby, R. M., & Parker, J. D. A. (1991). The alexithymia construct: A potential paradigm for psycho-
somatic medicine. Psychosomatics, 32, 153±164.
Taylor, G. J., Bagby, R. M. & Parker, J. D. A. (1997). Disorders of aect regulation: Alexithymia in medical and psy-
chiatric illness. Cambridge: Cambridge University Press.
Taylor, G. J., Ryan, D., & Bagby, R. M. (1985). Toward the development of a new self-report alexithymia scale. Psy-
chotherapy and Psychosomatics, 44, 191±199.
J.D.A. Parker et al. / Personality and Individual Dierences 30 (2001) 107±115 115