Archives of Clinical Neuropsychology 32 (2017) 1026–1036

The Impact of Executive Functions and Emotional Intelligence on Iowa

Gambling Task Performance: Focus on Right Frontal Lobe Damage
Oksana Zinchenko1,*, Elena Enikolopova2

Downloaded from https://academic.oup.com/acn/article/32/8/1026/3979241 by guest on 24 February 2023

1
Centre for Cognition and Decision Making, National Research University Higher School of Economics, Moscow, Russian Federation
2
Department of Psychology, Lomonosov Moscow State University, Moscow, Russian Federation
*Corresponding author at: Centre for Cognition and Decision Making, National Research University Higher School of Economics, 101000,
Moscow, 3A Krivokolenny Pereulok, Russian Federation. Tel.: +7(495) 916-8900 ext. 22370; fax: +7 (495) 6287931.
E-mail address: ozinchenko@hse.ru (O. Zinchenko).
Editorial Decision 27 June 2017; Accepted 29 June 2017

Abstract
Objective: Decision-making in the Iowa Gambling Task (IGT) has been intensively studied regarding both the “hot” and “cold” compo-
nents. The ventromedial prefrontal cortex is a key region involved in processing somatic marker information, though recent findings suggest
dorsolateral regions are also important. The dorsolateral prefrontal cortex is also known as a substrate of executive functions—the cold com-
ponent of decision-making. However, there is contradictory evidence about the role of executive functions, as well as the hot component of
decision-making—emotional intelligence. Previous findings suggest that patients with right frontal lobe lesions find decision-making more
problematic in IGT. The goal of this study is to replicate previous findings on IGT performance in patients with dorsolateral lesions com-
pared to controls.
Methods: We obtained data from patients with right frontal lobe tumors (n = 12), localized in the dorsolateral prefrontal cortex, and
healthy controls (n = 21) who undertook the IGT, Wisconsin Card Sorting Test (WCST), Mayer–Salovey–Caruso Emotional Intelligence
Test (MSCEIT), and D-KEFS Color-Word Interference Test.
Results: The performance in the IGT, WCST, and EI tests is impaired in the clinical group. At the subgroup level, we found patients had
lower EI scores regarding the ability to use “emotions for thinking facilitation”. However, we found an interaction between the EI scores
regarding the ability “the perception and identification of emotions” and the performance on WCST only in the patient group.
Conclusion: This study raises the possibility of identifying components of EI which could be helpful in understanding the impairment of
patients with right dorsolateral lesions.
Keywords: Iowa Gambling Task; Right frontal lobe tumor; Tumor patients; Executive functions; Lateral prefrontal cortex; Emotional intelligence

Introduction

This work investigates an intriguing and extensively studied process—decision-making. Research on decision-making
under uncertainty reveals important characteristics of human behavior as a holistic act and distinguishes quantitative specifics
of human behavior from animal behavior. If decision-making is evaluated through the prism of making the right choice, of
picking one option among several, then this choice could be explained as a result of calculations of gains and losses (the
“cold” component of decision-making) with emotional reactions having an affect on the desirability of each option (the “hot”
component of decision-making). While cold decision-making is related to cognitive and rational processes, hot decision-
making is associated with emotional and affective processes (Brand, 1985; Damasio, 1994; Shafir, Simonson, & Tversky,
1993). Hot components, such as affective and emotional reactions, help humans to act in situations requiring immediate
action, while cold components such as the rational determination of risks and the benefits of each possible action are more
valuable in stable situations (Séguin, Arseneault & Tremblay, 2007). These two factors or types of thinking are also referred

© The Author 2017. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
doi:10.1093/arclin/acx065 Advance Access publication on 19 July 2017
 O. Zinchenko, E. Enikolopova / Archives of Clinical Neuropsychology 32 (2017); 1026–1036 1027

to as System 1 and System 2 theoretical components, respectively (Kahneman, 2003). Executive functions (EF), such as plan-
ning, cognitive flexibility and working memory are thought to be linked to the “cold” part of decision-making. We could
assume that the “cold” component of decision-making could be measured on tests with emotionally neutral stimuli, such as
the Stroop Test or Wisconsin Card Sorting Test (Roiser et al., 2013). In contrast, the “hot” component of decision-making, as
mentioned before, involves the processing of emotionally significant information, such as gain–loss comparison and reward
anticipation in the Iowa Gambling Task (IGT).
Initially IGT was tested in patients with ventromedial prefrontal lesions who made significantly more disadvantageous de-
cisions compared to controls (Bechara, Damasio & Damasio, 2000a, 2000b; Bechara, Damasio, Damasio & Anderson, 1994).
These patients, who ignore the long-term consequences of their actions, are insensitive to the emotions of others and unable
to learn after making errors, so consequently they show risky behavior. Bechara and colleagues (1994) suggests a “somatic
marker hypothesis” to explain the origins of this behavior: “somatic markers” are the interoceptive signals associated with
subliminal emotions, which help to guide the decision-making process. Physiological changes, experienced for the first time

Downloaded from https://academic.oup.com/acn/article/32/8/1026/3979241 by guest on 24 February 2023

when stimuli are presented, occur in the body and then transform into emotions in the brain. The next time that the subject is
presented with the same stimuli, the information related to past outcomes is retrieved at the subliminal level as emotions.
These subliminal signals are processed in the ventromedial cortex and amydgala, so the lesions in the ventromedial regions
leading to the insensitivity to these internal physiological signals are associated with reward or avoidance. IGT performance
has been intensively studied to shed light on the specific affect of frontal regions, such as the orbitofrontal and dorsolateral re-
gions (Fellows & Farah, 2005; Manes et al., 2002) and their connections to distant regions, such as the cerebellum (Cardoso
et al., 2014) and the parietal cortices (Huettel, Song & McCarthy, 2005; Smith et al., 2009; Vickery & Jiang, 2009). It has
been suggested that ventromedial lesions located in the right hemisphere are linked to higher forms of impairment compared
to those in the left hemisphere due to the dominant role of the right hemisphere in emotional information processing (Bechara
et al., 2000a, 2000b). However, it is still controversial as to whether IGT performance is related to the “hot” component only;
Buelow & Blaine (2015) suggest first and latter IGT trials are both weakly correlated with the “cold” and “hot” trials of the
Columbia Card Task, which also measure risky behavior. It follows that IGT could be used to measure more complex behav-
ior related to both the “hot” and “cold” components of decision-making.
The understanding of the “hot” component of decision-making extends to not only emotions per se, but to general proces-
sing of emotional information. This has led to a line of research investigating the role of stable personality traits such as
impulsivity and risk-seeking, although less attention has been paid to emotional intelligence (Bar-On, Tranel, Denburg &
Bechara, 2003; Demaree, Burns & DeDonno, 2010; Webb, DelDonno & Killgore, 2014). Based on the somatic markers
hypothesis, Bar-On and colleagues (2003) suggest that decision-making impairment in patients with brain lesions in IGT
could be related to abnormal parameters of social and emotional intelligence. Moreover, these findings imply that neural net-
works associated with cognitive intelligence differ from social–emotional networks (Bar-On et al., 2003). However, neuropsy-
chological examinations and psychological findings suggest that emotional and cognitive intelligence are tightly bonded in
accordance with the “Vygotskian intelligence hypothesis” of the unity of the affective component and intelligence (Vygotsky,
1978). Vygotsky argued that human intelligence is the product of the dynamic interaction of cognitive skills and affective
states during ontogenesis (Vygotsky, 1978). It follows that better performance in the human decision-making process can
only occur when the emotional part (hot) and the rational part (cold) act as a system. This raises the question of how emo-
tional intelligence and executive functioning, which can be measured explicitly, have an affect on decision-making under
uncertainty in IGT.
The evidence regarding the possible relationship between EF as the “cold” component and decision-making measured by
IGT is also still controversial; EF impairment does not always influence IGT performance (Bechara, Damasio, Tranel &
Anderson, 1998), while other studies indicate that both hot and cold processes have an affect on different parts of the task
(Guillaume et al., 2009). Some studies (Gansler, Jerram, Vannorsdall & Schretlen, 2011; Hawthorne & Pierce, 2015) show
that IGT involves novel problem-solving and requires attention in healthy participants more than EF per se. No significant
correlations were found between IGT performance and EF measured by neuropsychological tests in stroke patients (Cardoso,
de, Branco, Cotrena & Fonseca, 2015). This implies that further investigation of the emotional components and executive
functions should be performed to determine the key factors of decision-making performance on IGT.
To address the inconsistency in the findings regarding performance in IGT and EF, the current study investigates the rela-
tionship between “hot” decision-making in IGT and the “cold” component in EF, such as inhibitory control and cognitive
flexibility in classic neuropsychological instruments such as the Wisconsin Card Sorting Test (WCST).
To answer the question of whether IGT performance involves a “hot” component of decision-making or not, we evaluate
the parameters of emotional intelligence by the Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT) to perform
statistical analysis of possible interaction effects. If IGT measures the “hot” component of decision-making, we expect to
reveal an effect of emotional intelligence on IGT performance in clinical groups.
1028 O. Zinchenko, E. Enikolopova / Archives of Clinical Neuropsychology 32 (2017); 1026–1036

Materials and Methods

Our study recruited two participant groups consisting of 12 patients with right frontal lobe tumors and 21 healthy controls
(2:3 study design). Participants were recruited from the Moscow Burdenko Institute of Neurosurgery prior to tumor surgery.
All patients suffered from convexital brain tumors in the right hemisphere, mostly localized in the lateral prefrontal cortex
region, and diagnosed by neurological and neuroimaging assessment. Patients were examined 1–2 days after their admission
to hospital, before surgery. The visits were done in Russian language. Data about the site of lesions were collected by review-
ing patient records and confirmed by the neurologist. Participants of the control group were recruited by e-mail and social net-
works from the Lomonosov Moscow State University and related institutions. Participants in the sample were native Russian
speakers, with at least 1 year of higher (post-secondary) education and were at least 18 years of age. Exclusion criteria were:
left-handedness or ambidextrous (screened by the neuropsychological assessment by Luria’s battery and Annett’s question-
naire of functional asymmetry), other neurological diseases (e.g. epilepsy, aneurism), aphasia and agnosia symptoms, psychi-

Downloaded from https://academic.oup.com/acn/article/32/8/1026/3979241 by guest on 24 February 2023

atric disorders and drug abuse (self-reported and retrospective review of patient history).
Table 1 shows the descriptive sociodemographic and clinical data of the clinical and control groups. The clinical group
included individuals aged between 18 and 58 (M(SD) = 34(12.37)), with at least 1 year of education (M(SD) = 4,33(1,23)), 25%
of whom were female (M(SD) = 1.25(0.45)). The control group included individuals aged between 18 and 54 (M(SD) = 29.86
(11.70)) with at least 1 year of university education (M(SD) = 3.95(1.36)), 43% of whom were female (M(SD) = 1.43(0.51)).
All participants signed written informed consent authorized by the local ethical community. Patients were assessed during
a single session lasting 2 hr, during which all data related to sociodemographic status and neuropsychological assessment
were obtained. The methods used are described subsequently:

• Luria’s Neuropsychological battery (Luria, 1973). This battery includes neuropsychological probes to assess a variety
of cognitive functions: attention, perception of different modalities, handedness, memory, and EF. A questionnaire
about sociodemographic characteristics was included.
• IGT (Bechara et al., 1994). A computerized version of IGT was used (PEBL free software, http://pebl.sourceforge.
net/). The participant is instructed to pick up a card from four decks of cards over the course of 100 trials. Two of
these decks are advantageous (C and D) because they consist of cards with small losses and high long-term gains,
while two others (A and B) are disadvantageous due to the probability of receiving a high monetary loss once every
10 trials. Performance in IGT was assessed by a set of different measures: The total score as a parameter of general
productivity is calculated by subtracting the number of disadvantageous deck selections from the number of advanta-
geous selections (C + D) – (A + B) as well as the partial performance of sets of 20 trials (0–20, 20–40, 40–60,
60–80, 80–100). We also calculated the quantity of choices to access preferences for each deck across all four decks.
The partial performance in blocks of 20 trials has been performed in order to examine the ability to learn about
advantageous and disadvantageous decks using somatic markers from early trials (0–20) to final trials (80–100). The
scores for the performance on the IGT for five consecutive blocks of 20 cards are calculated by subtracting the num-
ber of advantageous selections (decks C, D) from the number of disadvantageous selections (decks A, B). The learn-
ing curves have been created to represent the process of learning.
• WCST (Heaton, Chelune, Talley, Kay, & Curtis, 1993) provides information about cognitive reasoning and flexibil-
ity, and the set-shifting of the participant. It also can be used for the assessment of the level of brain damage to the
prefrontal cortex. Raw scores were used in the analysis (percentage errors and perseverative errors).
• D-KEFS Color-Word Interference Test (Delis, Kramer, Kaplan & Holdnack, 2004) is a part of the Delis–Kaplan
Executive Function System Test to measure the ability to inhibit an automatic verbal response, which typically occurs
in patients with prefrontal cortex damage. It is significantly different from previous versions of the Stroop Test
because of the introduction of a new “inhibition and switching” section. T-scores were used in the analysis.
• Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT v.2.0), Russian adaptation (Mayer, Salovey &
Caruso, 2002; Sergienko & Vetrova, 2009). Mayer and Salovey (Mayer & Salovey, 1997; Salovey & Mayer, 1990)
model emotional intelligence as a set of skills to evaluate, express and regulate one’s emotions. The evaluation of
one’s emotions is divided into verbal and non-verbal subcomponents, while the assessment of others’ emotions in-
cludes non-verbal perception and empathy. A later edition of their model (Mayer et al., 2002, 2003) includes four do-
mains which develop in ontogenesis: the perception and identification of emotions (Sections A and E), the use of
emotions for thinking facilitation (B and F), the understanding and analysis of emotions (C and G), and conscious
emotional control and regulation for personal growth and the improvement of interpersonal communication (D and H).
Our participants were asked to complete Sections A, B, E, F, but because of fatigue, complexity of the task (imagine a
situation and evaluate a role of imaginary emotion) and duration of the study not all patients completed Sections B
 O. Zinchenko, E. Enikolopova / Archives of Clinical Neuropsychology 32 (2017); 1026–1036 1029

Table 1. Sociodemographic characteristics of groups

Age Gender Years of Hemisphere Tumor type and localization Group Time when tumor was
(1-male, schooling diagnosed (months
2- female) (university before operation and
education) current study)
Clinical group description
1 32 1 5 Right Diffuse astrocytoma WHO Grade II–III, Dorsolateral 4
convexital access to the lateral prefrontal cortex,
and premotor cortex in the posterior part.
2 27 1 5 Right Anaplastic astrocytoma Ki 67—10% located in Dorsolateral 7
lateral prefrontal–premotor cortex—
involvement of superior frontal gyrus and
posterior part of middle frontal gyrus,

Downloaded from https://academic.oup.com/acn/article/32/8/1026/3979241 by guest on 24 February 2023

parasagittal location.
3 35 1 5 Right Diffuse astrocytoma WHO Grade II, with Dorsolateral 3
convexital access—lateral prefrontal–premotor
cortex, parasagittal location, d = 3.5 cm
4 43 2 4 Right Oligoastrocytoma WHO Grade II, located in Dorsolateral 8
lateral prefrontal–premotor cortex (from hand
representation region to the medial part of
inferior frontal gyrus)
5 26 2 5 Right Diffuse astrocytoma WHO Grade II, with Dorsolateral 5
convexital access—lateral prefrontal–premotor
cortex, d = 6 cm, with medial part adjoining the
right lateral ventriculus (frontal lobe)
6 49 1 5 Right Oligoastrocytoma IМ Ki 67, Grade III, with Dorsolateral 8
involvement of lateral prefrontal cortex,
premotor cortex (leg representation region) in
the posterior part
7 18 1 1 Right Astrocytoma with involvement of lateral prefrontal Dorsolateral 4
cortex (cortico-subcortical locatization)
8 21 2 4 Right Diffuse astrocytoma WHO Grade II, convexital Dorsolateral 6
fronto-parasagittal localization (lateral
prefrontal–premotor cortex), d = 3 cm
9 20 1 3 Right Gemangioblastoma with surrounding glial Dorsolateral 9
hyperplasia with convexital access (lateral
prefrontal–premotor cortex, hand representation)
10 40 1 5 Right Meningioma, adjoining to fronto-parasagittal Dorsolateral 3
region (premotor and lateral prefrontal cortex),
8 × 7 × 6 cm
11 39 1 5 Right Diffuse astrocytoma WHO Grade II–III, with Dorsolateral 3
involvement of lateral prefrontal cortex
12 58 1 5 Right Glioblastoma (Grade IV), located in frontal cortex Dorsolateral 3
with involvement of premotor cortex and lateral
prefrontal cortex
M(SD) 34(12.37) 1.25(0.45) 4.33(1.23)
Control group description
1 21 1 3 — — Control —
2 29 1 5 — — Control —
3 54 1 5 — — Control —
4 21 2 3 — — Control —
5 24 2 5 — — Control —
6 20 1 2 — — Control —
7 22 1 4 — — Control —
8 51 1 5 — — Control —
9 46 1 5 — — Control —
10 24 1 4 — — Control —
11 21 2 3 — — Control —
12 22 2 4 — — Control —
13 30 1 5 — — Control —
14 18 2 1 — — Control —

(continued on next page)

1030 O. Zinchenko, E. Enikolopova / Archives of Clinical Neuropsychology 32 (2017); 1026–1036

Table 1. (continued)
Age Gender Years of Hemisphere Tumor type and localization Group Time when tumor was
(1-male, schooling diagnosed (months
2- female) (university before operation and
education) current study)
15 22 2 3 — — Control —
16 18 2 1 — — Control —
17 31 1 5 — — Control —
18 48 2 5 — — Control —
19 45 2 5 — — Control —
20 27 1 5 — — Control —
21 33 1 5 — — Control —
M(SD) 29.86(11.70) 1.43(0.51) 3.95(1.36)

Downloaded from https://academic.oup.com/acn/article/32/8/1026/3979241 by guest on 24 February 2023

and F, respectively (6 of 12 patients). We asked only part of the control group to complete Sections B and F to support
the 2:3 design and ratio of participants in the groups. To perform it, after the collection of clinical data we randomly
assigned the subjects in the control group the B and F sections of the MSCEIT v.2.0 test.

Statistical Analyses

The normality of data distribution within groups was assessed with Shapiro–Wilk criteria because of small sample size. In
spite of the small sample size, the data was normally distributed. All variables included in the analysis were derived from nor-
mally distributed populations (all ps > .05).
Parameters related to IGT performance (total score, performance per block and deck preferences) were analyzed through
MANOVA with Shidak post-hoc test for multiple comparisons to access between group differences. Significance was con-
sidered at α = 0.05. The relationships between behavioral measures were analyzed with linear regression analysis. To evaluate
the effect of EI on IGT performance and WCST, a sum of scores from the “perception and identification of emotions” section
(A and E) was used as an independent variable, while total score, performance per block, preference per deck in IGT and
“percentage errors” and “perseverative errors” in WCST were used as the dependent variables. The Pearson correlation was
used while correlation analysis was also implemented. SPPS 22 has been used to analyze the data.

Results

The whole set of data included in the analysis (group descriptive statistics, F, p-values, effect size) is presented in Table 2.

Descriptive Statistics

Statistical analyses did not identify significant differences in sociodemographic characteristics between groups (age:
F(1,31) = 0.920, p = .345; education: F(1,31) = 0.641, p = .430; gender: F(1,31) = 1.021, p = .320). The control and clinical
groups did not differ in regards to clinical variables (no comparisons were significant at p < .05).
We also perform descriptive statistics for the subgroup analysis to control there are no differences in subsets of controls
and patients ((age: F(1,18) = 3.074, p = .097, gender: F(1,18) = 0.900, p = .355, years of education (university): F(1,18) =
2.021, p = .173)). The control and clinical groups did not differ in regards to clinical variables (no comparisons were signifi-
cant at p < .05).

Iowa Gambling Task

The differences in decision-making performance in IGT were firstly assessed by total scores using the formula (C + D) –
(A + B). The performance per block for general decision-making performance was calculated for each 20 trials in all three
groups across the task using the formula (C + D)–(A + B). Also a parameter of preference per deck was calculated for each
deck separately. The MANOVA test was applied with Shidak post-hoc correction (see Table 2 for whole analysis).
 O. Zinchenko, E. Enikolopova / Archives of Clinical Neuropsychology 32 (2017); 1026–1036 1031

Table 2. MANOVA analyses.

F p-Value Partial n-squared Group (1-clinical, 2-control) Descriptive statistics

M SD
Wisconsin Card Sorting Test
Percentage errors 9.476 .004 0.234 1 28.9133 14.97841
2 16.9238 7.49359
Perseverative errors 7.614 .010 0.197 1 18.3550 10.21497
2 11.0171 5.12921
Not perseverative errors 3.779 .061 0.109 1 10.3850 8.56953
2 5.4005 6.11715
Iowa Gambling Task
Selections Deck A 0.801 .378 0.025 1 17.0833 5.55073

Downloaded from https://academic.oup.com/acn/article/32/8/1026/3979241 by guest on 24 February 2023

2 15.2381 5.77845
Selections Deck B 5.238 .029 0.145 1 27.7500 5.70685
2 21.3333 8.66795
Selections Deck C 0.159 .693 0.005 1 29.2500 5.87947
2 30.8095 12.73428
Selections Deck D 3.336 .077 0.097 1 25.9167 5.99179
2 32.6190 11.81726
Performance Block 0–20 0.115 .737 0.004 1 –1.6667 4.24978
2 –2.3810 6.53051
Performance Block 20–40 4.337 .046 0.123 1 2.6667 2.99495
2 7.0476 6.88822
Performance Block 40–60 1.790 .191 0.055 1 2.3333 6.02017
2 5.6190 7.17270
Performance Block 60–80 1.573 .219 0.048 1 3.3333 8.23886
2 7.5238 9.73457
Performance Block 80–100 3.315 .078 0.097 1 3.6667 8.85575
2 9.0476 7.76193
(C + D) – (A + B) 4.488 .042 0.126 1 10.3333 14.21480
2 26.8571 24.67850
Mayer–Salovey–Caruso Emotional Intelligence Test
Scores A_E 0.092 .765 0.005 1 6.3333 1.63299
2 6.0714 1.81720
Scores B_F 4.440 .049 0.198 1 4.5000 3.56371
2 6.9286 1.68543
Frequencies B_F 5.508 .031 0.234 1 0.3133 0.10801
2 0.3914 0.04418
Delis–Kaplan Executive Function System Color-Word Interference Test
T-scores (Section 3 – Section 1) 2.377 .133 0.071 1 12.9167 2.81096
2 11.6667 1.85293
T-scores (Section 4 – Section 1) 0.206 .653 0.007 1 11.7500 4.22385
2 11.1429 3.36579

A total score. The results of MANOVA indicated a significant difference between healthy controls [M(SD) = 26.8571
(24.67850)] and the clinical group: [M(SD) = 10.3333(14.21480)] (F(1,31) = 4.488, p = .042, partial n2 = 0.126) (see
Table 2).

Performance per block. MANOVA across two groups indicated a significant difference in performance in the block 20–40
(F(1,31) = 4.337; p = .046, partial n2 = 0.123) (see Table 2). This suggest that clinical group in the beginning of the game
shows significantly lower performance than healthy controls (M(SD) = 2.6667(2.99495), M(SD) = 7.0476 (6.8882)).

Preference per deck. Table 2 shows the results of statistical analysis across two groups for their preferences measured as a
number of selections for each deck. This suggests that selections from disadvantageous deck B (F(1,31) = 5.238, p = .029,
partial n2 = 0.145) significantly differ between the clinical group (M(SD) = 27.7500(5.7068)) and healthy controls (M(SD) =
21.3333(8.66795)).
1032 O. Zinchenko, E. Enikolopova / Archives of Clinical Neuropsychology 32 (2017); 1026–1036

Learning curves. Fig. 1 shows repeated-measures ANOVA results across two groups for their performance through five
blocks of IGT. No significant differences were found for clinical group (F(2,141,23,546) = 1.35, p = .289, partial n2 =
0.107). However, the IGT performance through five blocks differed significantly in control group (F(4,80) = 9.748, p = .000,
partial n2 = 0.328). Post-hoc analysis with Shidak correction has been applied. A performance between first and all next
blocks differed significantly (p < .05), while all next comparisons (2–3, 3–4, 4–5) were not significantly different.

Wisconsin Card Sorting Test

MANOVA revealed a significant difference between the clinical group and healthy controls in branches of parameters,
such as Percentage errors (F(1,31) = 9.476, p = .004, partial n2 = 0.234) and perseverative errors (F(1,31) = 7,614, p = .01,
partial n2 = 0.197) (see Table 2). This suggests that participants from the clinical group made more errors then healthy con-

Downloaded from https://academic.oup.com/acn/article/32/8/1026/3979241 by guest on 24 February 2023

trols (M(SD) = 28.9133(14.97841), M(SD) = 16.9238(7.4935)) and made more perseverative errors (M(SD) = 18.3550
(10.2149), M(SD) = 11.0171(5.1292)).
No other significant differences between groups were found.
In further analysis only survived parameters of IGT and WCST has been included: performance per block 20–40 in IGT,
preference per deck B in IGT, total score in IGT, Percentage errors and perseverative errors in WCST.

Wisconsin Card Sorting Test and Iowa Gambling Task. To evaluate the possible interaction between the WCST and IGT
parameters, first a correlation analysis with the Pearson correlation was implemented to access the direction and strength of
the relationship between IGT and WCST significantly differed variables between groups (WCST percentage errors and per-
severative errors, selections from Deck B, performance in block 20–40 and overall score in IGT). Second, if any variables ap-
peared to be related, linear regression analysis was implemented to evaluate the percentage of one variable dispersion that
could be predicted by another.
No significance was found in correlation and regression analyses of statistically differed parameters between groups
(WCST Percentage errors and perseverative errors, selections from Deck B, performance in block 20–40 and overall score in
IGT).

D-KEFS Color-Word Interference Test. No significant differences in overall performance and correlations with performance
in the Iowa Gambling Task were observed.

Emotional intelligence. Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT v.2.0)—Both groups completed
Sections A and E in MSCEIT, while only 6 out of 12 patients and 14 out of 21 controls completed Sections B and F.
MANOVA revealed no significant differences between subgroup of healthy controls and the clinical group in scores and
frequencies in emotional perception and identification (Sections A and E). There is a significant difference between overall
frequencies in Sections B and F (F(1,18) = 5,508, p = .031, partial n2 = 0.234) between controls (M(SD) = 0.3914(0.04418))

Fig. 1. Learning curves (Iowa Gambling Task) for clinical and control groups.
 O. Zinchenko, E. Enikolopova / Archives of Clinical Neuropsychology 32 (2017); 1026–1036 1033

and the clinical group (M(SD) = 0.3133(0.10801)) and overall scores (F(1,18) = 4.440, p = .049, partial n2 = 0.198) between
controls (M(SD) = 6.9286(1.68543)) and the clinical group (M(SD) = 4.500(3.56371)).
Emotional intelligence and performance in the Wisconsin Card Sorting Test—In patients group there is a significant effect of
scores in Sections A and E (beta = –0.661, p = .019, R2-corrected =0.381) on percentage errors in WCST, while there is no
such effect in healthy controls (p = .315).
No other significant effects were found.
Emotional intelligence and performance in the Iowa Gambling Task—No effects of scores in Sections A and E were found
on significantly differed between groups parameters of IGT.

Subgroup analyses. We performed a separate analysis for subjects who completed Sections B and F in MSCEIT to reveal the
possible influence of emotions on the thinking process, as the authors of MSCEIT initially proposed the value for these scales.

Downloaded from https://academic.oup.com/acn/article/32/8/1026/3979241 by guest on 24 February 2023

The scores of subsection F and frequencies converted from overall scores in Sections B and F were included in the further
analysis for subgroup performance in IGT and WCST (number of participants included in the clinical subgroup = 6, healthy
controls = 14). We expected to find an interaction/influence between the parameters of emotional intelligence on IGT perfor-
mance, but not on WCST.

Iowa Gambling Task Performance

A total score. The results of MANOVA indicated a significant difference between healthy controls [M(SD) = 30.2857
(26.6355)] and the clinical group: [M(SD) = 5.3333(15.2665)] (F(1,18) = 4,.531, p = .047, n2 = 0.201)

Performance per block. MANOVA across two groups indicated a significant difference in performance in the block 80–100
(F(1,18) = 5.121; p = .036, n2 = 0.221) between healthy controls [M(SD) = 10.4286(7.7331)] and the clinical group: [M(SD) =
0.666(11.2190)], while a difference between groups in block 20–40 approaches significance (F(1,18) = 3.896; p = .064, n2 =
0.178).

Preference per deck. The selections from disadvantageous deck B (F(1,18) = 5.262, p = .034, n2 = 0.226) significantly differ
between the clinical group (M(SD) = 29.6667(4.6332)) and healthy controls (M(SD) = 20.2143(9.5125)).

Emotional intelligence and performance in the Iowa Gambling Task. A linear regression analysis revealed no interaction/
influence of frequencies converted from overall scores in Sections B and F on selection from deck B in clinical subgroup.

Wisconsin Card Sorting Test performance. There were no significant differences between clinical subgroup and healthy con-
trols in different parameters of WCST.

Emotional intelligence and performance in the Wisconsin Card Sorting Test. A linear regression analysis revealed no influ-
ence of frequencies converted from overall scores in Sections B and F on WCST parameters significantly differed across
groups.

Discussion

The current study assesses decision-making under uncertainty in patients with right dorsolateral prefrontal cortex tumors
compared to healthy controls and revealed the significant impairment in EF and IGT performance and EI in the clinical group.
Moreover, this study sought to replicate the findings of poor performance in IGT in patients with dorsolateral prefrontal cortex
lesions (Manes et al., 2002) because of the contradictory findings of Bechara and colleagues (1994; 2000a, 2000b). This pres-
ent study is also the first to obtain data on the Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT v.2.0) from a
clinical sample with tumor lesions.
We found that total IGT scores and performance per second block (20–40) differed significantly between the group of pa-
tients with dorsolateral lesions and the control group. According to the learning curves obtained, both control and clinical
groups started to display positive total score from block 2 (20–40). A preference for disadvantageous deck B is more common
in the clinical group compared to the control group. The former results are consistent with previous findings (Cardoso et al.,
2014) obtained for a frontal stroke lesions group. This indicates that patients with dorsolateral prefrontal cortex lesions lose
1034 O. Zinchenko, E. Enikolopova / Archives of Clinical Neuropsychology 32 (2017); 1026–1036

the ability to learn to avoid disadvantageous decks, though nothing could be said about their sensitivity to advantageous
decks. This means that patients prefer immediate large rewards and are not able to evaluate the long-term consequences of
their actions—compare worse outcome in Deck B preferences and better outcome in advantageous decks with a long-term
perspective. A significant difference in IGT performance in block 20–40 also shows that patients choose a disadvantageous
strategy more often at first trials then healthy controls.
Does an impairment of the “cold” component or “hot” component result in such behavior? Brand, Recknor, Grabenhorst,
and Bechara (2007) suggest that two different mechanisms underlie the IGT performance: decisions under uncertainty in the
first trials and decisions under risk in the latter trials. Thus, an impaired “cold” component should have a higher affect on IGT
performance in the first trials and no affect in the latter trials. To address this question, we performed regression analysis to
investigate the possible interaction between EF tests, EI measurements and IGT performance. In the current study, we did not
see any significant interaction with regard to regression analysis between WCST parameters and IGT performance per block,
though we expect that it could be reproducible in larger samples (for instance, n = 97 in the study by Brand and colleagues

Downloaded from https://academic.oup.com/acn/article/32/8/1026/3979241 by guest on 24 February 2023

[2007]). However, some studies imply (Hawthorne & Pierce, 2015) that a higher cognitive load could result in impaired IGT
performance, especially a greater selection of Deck B. Since we did not measure the attentional capacity of our subjects
directly, we expect that this indirect influence in the clinical group will be proved in further research.
In accordance with previous findings (Manes et al., 2002), we found that patients with dorsolateral lesions showed an
impairment in cognitive flexibility (WCST, perseverative errors) and the IGT. It is important to note that we were able to repli-
cate the results of previous studies on a larger sample: n = 12 compared to n = 4 in the study by Manes and colleagues (2002).
However, some findings imply that there is no influence of EF on overall IGT decision-making in frontal stroke patients, but
there are some correlated partial parameters, such as net scores for the first block and the number of correct answers in the
modified version of WCST (Cardoso et al., 2015). Toplak’s review (2010) also suggests that interactions between set-shifting,
cognitive flexibility measurements and IGT performance are inconsistent across studies. In accordance with previous studies
(Cardoso et al., 2015; Toplak, Sorge, Benoit, West & Stanovich, 2010), we found no significant correlations or interactions
between IGT performance and scores on “cold” EF tests. We also did not reveal this specificity in the performance of different
IGT blocks, as some previous studies have reported (Brand et al., 2007; Cardoso et al., 2015).
Our findings do not suggest that performance in the Color-Word Interference Test is impaired in patients with right frontal
lobe tumors or that the results of the Color-Word Interference Test do not predict performance in IGT. Previous findings also
reported the absence of any correlation or interaction between the Stroop Test, which corresponds mainly with the structure
of the D-KEFS Color-Word Interference Test, and IGT (Bechara et al., 2001; Mimura, Oeda, & Kawamura, 2006). The
review by Toplak and colleagues (2010) also provides evidence that the correlation between the performance of IGT and mea-
sures of inhibition are relatively low across 11 studies. We expect that such results could be obtained because of differences
in the neural basis of cognitive switching performance: recent findings suggest that the gray matter volume of the left tempo-
ral lobe predicts the outcome of the switching condition in the D-KEFS Color Word Interference Test (Adolfsdottir et al.,
2010).
The preliminary results suggest that patients with right dorsolateral prefrontal cortex tumors and healthy controls do not
differ in their ability to identify and perceive emotions correctly, but in the ability to understand and use emotions to facilitate
activity. According to our results, the ability to identify and verbalize subliminal emotional conditions is highly impaired in
patients with right frontal lobe tumors. Although no significant differences in scores A and E were found between the clinical
and control group, we found a significant influence of this parameter on the quantity of percentage errors in WCST only in
the clinical group. Our results are partially consistent with the findings of Bar-On and colleagues (2003) on the affect of emo-
tional intelligence on IGT performance: the interaction was observed in the clinical group, but not in the controls. Webb and
colleagues (2014) and Demaree and colleagues (2010) also show that emotional intelligence scores significantly predict IGT
performance until starting to control for the intelligence quotient.
The results support the hypothesis that right frontal lobe damage (not only ventromedial, as has been shown before, but
dorsolateral prefrontal) can greatly impair behavioral performance in IGT. We assumed that EF will be impaired in our clini-
cal group due to the tumor frontal localization. Since we have found only an interaction between emotional intelligence and
IGT performance in the specifically chosen clinical group, we suggest that IGT impaired performance in right frontal lobe
damage mostly reflect the “hot” component of the decision-making process.

Limitations

The main limitation of the neuropsychological methodology used in this study is that we performed all visits in Russian
and measures were then translated. Here we present preliminary results of the influence of emotional intelligence on some
 O. Zinchenko, E. Enikolopova / Archives of Clinical Neuropsychology 32 (2017); 1026–1036 1035

parameters of cognitive tests, especially Sections B and F of MSCEIT. We expect that the latter will be confirmed with
increasing sample size. Since T-scores were used in the analysis of the D-KEFS Color-Word Interference Test, we expect cul-
tural differences in the normative sample compared to the study sample to be an additional limitation of the study. However,
patients were not tested specifically to evaluate mood/emotional status and it could be a confounding variable as only patients
who voluntarily agreed to participate and were neutral were asked to take part in the study. Despite the small sample size,
which could be a limitation, we checked for group differences to be sure that the main and subgroup analysis were statistically
correct. Moreover, neither the main groups nor subgroups differ for any sociodemographic variable.

Funding

During the data analysis and writing of this paper OZ was supported by the Russian Academic Excellence Project “5-100”.

Downloaded from https://academic.oup.com/acn/article/32/8/1026/3979241 by guest on 24 February 2023

Conflict of Interest

The authors declare no competing financial interests.

Acknowledgments

The authors gratefully acknowledge Dr Sveltana Buklina, MD, PhD for supervision of the current work during data collec-
tion and observation.

References

Adolfsdottir, S., Ystad, M. A., Haasz, J., Wehling, E., Lundervold, A., & Lundervold, A. J. (2010). Is cognitive switching performance on the D-KEFS Color
Word Interference Test predicted by frontal lobe volume? Frontiers in Human Neuroscience Conference Abstract: The 20th Annual Rotman Research
Institute Conference, The frontal lobes. doi:10.3389/conf.fnins.2010.14.00089
Bar-On, R., Tranel, D., Denburg, N. L., & Bechara, A. (2003). Exploring the neurological substrate of emotional and social intelligence. Brain, 126 (Pt 8),
1790–1800.
Bechara, A., Damasio, A. R., Damasio, H., & Anderson, S. W. (1994). Insensitivity to future consequences following damage to human prefrontal cortex.
Cognition, 50, 7–15.
Bechara, A., Damasio, H., Tranel, D., & Anderson, S. W. (1998). Dissociation of working memory from decision making within the human prefrontal cortex.
Journal of Neuroscience, 18, 428–437.
Bechara, A., Damasio, A. R., & Damasio, H. (2000a). Characterization of the decision-making deficit of patients with ventromedial prefrontal cortex lesions.
Brain, 123, 2189–2202.
Bechara, A., Damasio, A. R., & Damasio, H. (2000b). Emotion, decision-making and orbitofrontal cortex. Cerebral Cortex, 10, 295–307.
Bechara, A., Dolan, S., Denburg, N., Hindes, A., Anderson, S. W., & Nathan, P. E. (2001). Decision-making deficits, linked to a dysfunctional ventromedial
prefrontal cortex, revealed in alcohol and stimulant abusers. Neuropsychologia, 39, 376–389.
Buelow, M. T., & Blaine, A. L. (2015). The assessment of risky decision making: A factor analysis of performance on the Iowa Gambling Task, Balloon
Analogue Risk Task, and Columbia Card Task. Psychological Assessment, 27, 777–785.
Brand, A. (1985). Hot cognition: emotions and writing behavior. Journal of Advanced Composition, 6, 5–15. Retrieved from http://www.jstor.org/stable/
20865583.
Brand, M., Recknor, E. C., Grabenhorst, F., & Bechara, A. (2007). Decisions under ambiguity and decisions under risk: Correlations with executive functions
and comparisons of two different gambling tasks with implicit and explicit rules. Journal of Clinical and Experimental Neuropsychology, 29, 86–99.
Cardoso, C., de, O., Branco, L. D., Cotrena, C., Kristensen, C. H., Schneider Bakos, D. D. G., et al. (2014). The impact of frontal and cerebellar lesions on
decision making: Evidence from the Iowa Gambling Task. Frontiers in Neuroscience, 8, 61. http://doi.org/10.3389/fnins.2014.00061.
Cardoso, C., de, O., Branco, L. D., Cotrena, C., & Fonseca, R. P. (2015). Correlational analysis of performance in executive function tasks after stroke.
Psychology & Neuroscience, 8, 56–65. http://dx.doi.org/10.1037/h0101021.
Damasio, A. R. (1994). Descartes’ error: emotion, reason, and the human brain. New York: Putnam.
Delis, D. C., Kramer, J. H., Kaplan, E., & Holdnack, J. (2004). Letter to the editor: Reliability and validity of the Delis-Kalplan Executive Function System:
An update. Journal of the International Neuropsychological Society, 10, 301–303.
Demaree, H. A., Burns, K. J., & DeDonno, M. A. (2010). Intelligence, but not emotional intelligence, predicts Iowa Gambling Task performance.
Intelligence, 38, 249–254.
Fellows, L. K., & Farah, M. J. (2005). Different underlying impairments in decision-making following ventromedial and dorsolateral frontal lobe damage in
humans. Cerebral Cortex, 15, 58–63.
Gansler, D. A., Jerram, M. W., Vannorsdall, T. D., & Schretlen, D. J. (2011). Does the Iowa Gambling Task Measure Executive Function? Archives of
Clinical Neuropsychology, 26, 706–717.
Guillaume, S., Jollant, F., Jaussent, I., Lawrence, N., Malafosse, A., & Courtet, P. (2009). Somatic markers and explicit knowledge are both involved in decision-
making. Neuropsychologia, 47 (10), 2120–2124.
1036 O. Zinchenko, E. Enikolopova / Archives of Clinical Neuropsychology 32 (2017); 1026–1036

Hawthorne, M. J., & Pierce, B. H. (2015). Disadvantageous Deck Selection in the Iowa Gambling Task: The effect of cognitive load. Europe’s Journal of
Psychology, 11, 335–348. doi:10.5964/ejop.v11i2.931.
Heaton, R. K., Chelune, G. J., Talley, J. L., Kay, G. G., & Curtis, G. (1993). Wisconsin Card Sorting Test (WCST). manual revised and expanded. Odessa:
Psychological Assessment Resources Inc.
Huettel, S. A., Song, A. W., & McCarthy, G. (2005). Decisions under uncertainty: Probabilistic context influences activation of prefrontal and parietal corti-
ces. Journal of Neuroscience, 25, 3304–3311.
Kahneman, D. (2003). A perspective on judgement and choice. American Psychologist, 58, 697–720. doi:10.1037/0003-066x.58.9.697.
Luria, A. R. (1973). The scheme of neuropsychological studies. M.: Moscow State Univesity Publishing House.
Manes, F., Sahakian, B., Clark, L., Rogers, R., Antoun, N., Aitken, M., et al. (2002). Decision-making processes following damage to the prefrontal cortex.
Brain, 125, 624–639.
Mayer, J. D., & Salovey, P. (1997). What is emotional intelligence? In P. Salovey, & D. Sluyter (Eds.), Emotional development and emotional intelligence:
Educational implications (pp. 3–31). New York: Perseus Books Group.
Mayer, J. D., Salovey, P., & Caruso, D. R. (2002). Mayer–Salovey–Caruso Emotional Intelligence Test (MSCEIT) user’s manual. Toronto, Canada: MHS
Publishers.

Downloaded from https://academic.oup.com/acn/article/32/8/1026/3979241 by guest on 24 February 2023

Mayer, J. D., Salovey, P., Caruso, D. R., & Sitarenios, G. (2003). Measuring Emotional Intelligence with the MSCEIT V2.0. Emotion (Washington, D.C.), 3,
97–105.
Mimura, M., Oeda, R., & Kawamura, M. (2006). Impaired decision-making in Parkinson’s disease. Parkinsonism & Related Disorders, 12 (3), 169–175. doi:
http://dx.doi.org/10.1016/j.parkreldis.2005.12.003.
Roiser, J. P., & Sahakian, B. J. (2013). Hot and cold cognition in depression. CNS Spectrums, 18, 139–149.
Salovey, P., & Mayer, J. D. (1990). Emotional intelligence. Imagination, Cognition, and Personality, 9, 185–211.
Séguin, J. R., Arseneault, L., & Tremblay, R. E. (2007). The contribution of “cool” and “hot” components of decision-making in adolescence: Implications
for developmental psychopathology. Cognitive Development, 22, 530–543.
Sergienko, E. A., & Vetrova, I. I. (2009). Emotional intelligence: Russian-language adaptation of The Mayer–Salovey–Caruso Emotional Intelligence Test,
Version 2.0. Psikhologicheskie Issledovaniya, 6. http://psystudy.ru. [in Russian, abstr. in English].
Shafir, E., Simonson, I., & Tversky, A. (1993). Reason-based choice. Cognition, 49, 11–36. doi:10.1016/0010-0277(93)90034-S.
Smith, B. W., Mitchell, D. G. V., Hardin, M. G., Jazbec, S., Fridberg, D., Blair, R. J. R., et al. (2009). Neural substrates of reward magnitude, probability,
and risk during a wheel of fortune decision-making task. Neuroimage, 44, 600–609.
Toplak, M. E., Sorge, G. B., Benoit, A., West, R. F., & Stanovich, K. E. (2010). Decision-making and cognitive abilities: A review of associations between
Iowa Gambling Task performance, executive functions, and intelligence. Clinical Psychology Review, 30, 562–581.
Vickery, T. J., & Jiang, Y. V. (2009). Inferior parietal lobule supports decision-making under uncertainty in humans. Cerebral Cortex, 19, 916–925.
Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press.
Webb, C. A., DelDonno, S., & Killgore, W. D. (2014). The role of cognitive versus emotional intelligence in Iowa Gambling Task performance: What’s emo-
tion got to do with it? Intelligence, 44, 112–119.