Impulsive and rigid temperament subtypes and executive functioning an exploratory study of temperament configurations and neuropsychological performance Raegan Murphy University College Cork, Ireland Nafisa Cassimjee University of Pretoria, South Africa 42 nd Annual PSI conference, Cork, Ireland, November 2012
Introduction
Temperament, its biological basis and how it influences behaviour and pathogenesis has enjoyed widespread attention dating as far back as 350 BC with the contributions of the Greek philosophers Hippocrates and Galen.
More recently, facilitated by developments in the fields of neurophysiology, neuropsychology and neuroimaging, various studies have focused on investigating the links between heritable temperament dimensions and specific psychiatric disorders and on the relationship between psychiatric disorders and neuropsychological functioning.
Moreover, research has set out to elucidate the neurophysiological aspects underlying and contributing toward temperament and executive functions all of which have been aimed at the detection and diagnosis of subclinical putative risk factors for psychopathology.
Although the results of these research endeavours have shown that various temperament traits and executive functioning share common underlying neural systems, with each individually implicated in the development of some forms of psychopathology, few studies have been conducted exploring specific temperament configurations and their accompanying executive functioning correlates.
It is suggested that analysing trait configurations may emerge as a valuable factor in identifying predictive and subclinical attributes in the development of some forms of psychopathology.
Overview
The temperament traits harm avoidance (HA), novelty seeking (NS), reward dependence (RD) and persistence (P) are measured by the Temperament and Character Inventory (TCI), an instrument derived from Cloninger s psychobiological theory of personality.
According to Cloninger temperament is the precedent for our pre-potent responses to commence, sustain and halt behavioural responses.
Novelty seeking is an individual s tendency to action behaviours and is associated with the genetic bias to impulsivity, quick loss of temper and exploratory activities. Harm avoidance is associated with a tendency to inhibit behaviours, and is associated with a fearful, cautious, pessimistic and shy approach.
These traits are related to independent yet dynamic neural networks, namely the behavioural inhibition system (BIS) and the behavioural activation system (BAS).
Differences in temperament dispositions may be an antecedent for observed differences in executive abilities.
Executive functioning involves the integration of higher order processing comprising motivational, affective, cognitive, and behavioural components necessary for decision-making, planning and goal-oriented behaviours. Thus, executive functioning is a complex neuropsychological construct sub serving various levels and processes of neurocognitive attenuation and augmentation.
Associated behavioural configurations of HA and NS constitute the Rigid temperament subtype (low NS and high HA) and Impulsive temperament subtype (high NS and low HA).
Rigidity is characterised by passivity, as well as the inability to tolerate conflict and ambiguity, and Impulsivity is characterised by impatience and recklessness. According to Cloninger the Impulsive subtype is reflective of histrionic and antisocial personalities, and the Rigid subtype of the passive dependent and obsessional personalities.
Questions
The motivation for this study is twofold: firstly, previous research has shown HA and NS trait configurations are associated with characteristics of specific psychopathological disorders. secondly, specific psychopathological disorders are characterised by executive performance profiles.
Hence this exploratory study aims to investigate differences in executive performance as a function of Rigid and Impulsive temperament subtypes, which reflect HA and NS configurations.
Method
A non-experimental relational design was employed. The PennCNP comprises three computerized neuropsychological test batteries (Emotions, Executive Function and Abstract Reasoning). Self-report socio-demographic Q. The Temperament and Character Inventory (TCI), a 238 item forced choice true-false standardised self-administered questionnaire, derived from Cloninger s psychobiological personality model.
Sample
First year psychology students at a residential university in South Africa (ethical clearance). 630 students from the 1,124 registered students agreed to participate in the study. 210 cases were deleted due to incomplete records and cases with medical history (epilepsy, head injury, depression). A realized sample of 420 was utilised for the final data analyzes.
These participants were categorised according to their scores on HA and NS. Mean splits were used to categorise high and low HA and high and low NS. Those with high HA and low NS were labelled as Rigid Temperament (n = 131). Those with low HA and high NS as Impulsive Temperament (n = 121).
Procedure
The computer laboratory at the University was used for the group administration of tests. Participants were required to choose a scheduled session and were assigned to one of 30 groups. 3 researchers, 8 research assistants.
Method of administration
The PennCNP was administered using clickable icons on desktop computers, in a fixed order. The tests were implemented on IBM computers. An Applescript routine is used to collect participant IDs and basic demographic information and to present the tests in a prescribed order.
The results are uploaded to a data repository using an automated script, and scored using a program written in the Python programming language. There is no timelag effect and internet connectivity speed is not a confounding issue.
Results
In order to investigate performance variations between the Rigid and Impulsive subtypes, tests of difference were conducted in addition to controlling for possible confounds using covariant analyses on all the executive performance outcomes.
The following table shows results from an independent group t-test, which highlights the significant differences between the subtypes on a number of executive performance outcomes. All significant differences were for reaction time and not for performance accuracy.
Executive Domain Temperament Group Statistics Group N Mean Std. Deviation Std. Error Mean p value Abstraction and ConceptFormation Reaction Time incorrect [Block 2] Rigid Impulsive 96 89 2787.94 2445.07 1235.05 1068.99 126.05 113.31 Attention and Working Memory Reaction Time Rigid 131 439.59 111.45 9.73 Impulsive 120 412.25 76.18 6.95 Reaction Time [Block 1] Rigid 131 456.54 127.82 11.16 Impulsive 120 416.83 103.05 9.40 Sensorimotor Ability Reaction Time [Trial 1] Rigid 131 830.79 319.62 27.92 Impulsive 121 744.64 235.60 21.41 Reaction Time [Trial 2] Rigid 131 612.18 107.63 9.40 Impulsive 121 575.44 88.10 8.00 Abstraction in Executive Functioning Reaction Time Rigid 131 1994.36 518.91 45.33 Impulsive 120 1870.77 466.27 42.56 Abstraction and Mental Flexibility Reaction Time Rigid 130 18968.42 8906.19 781.12 0.046 0.025 0.008 0.016 0.003 0.049
The impulsive group proved to be faster on abstraction and concept formation; attention and working memory; sensorimotor ability; abstraction in executive functioning as well as abstraction and mental flexibility.
In order to determine whether intelligence influenced reaction times on performance outcomes, a number of analyses of covariance were conducted on the executive functioning variables.
To assess whether there was an interaction between the groups and measured intelligence with a variety of executive functioning outcomes, a preliminary ANCOVA was run using the SPSS GLM with custom model that included an intelligence by group interaction term with intelligence being the covariate. This was used to indicate the absence of a significant violation of the homogeneity of regression assumption.
The following table conveys the unadjusted and adjusted scores for the various performance outcomes for which main effects (rigid and impulsive) were significant when using intelligence scores as a co-variate. The rank orderings of the group means was not changed by adjustment for the covariate; however, after adjustment, the means were slightly higher/lower for each group. All adjusted scores are statistically significant.
Executive Domain Temperament Group Statistics Group N Unadjusted scores Adjusted scores Intelligence mean score (sd) Abstraction and Concept Formation Reaction Time incorrect [Block2] Rigid 96 2787.94 ms 2788.55 ms 43.73 (9.34) Impulsive 89 2445.07 ms 2425.98 ms 43.94 (8.91) Abstraction in Executive Functioning Reaction Time Rigid 131 1994.36 ms 1993.43 ms 43.73 (9.34) Impulsive 120 1870.77 ms 1859.66 ms 43.94 (8.91) Abstraction and Mental Flexibility Reaction Time Rigid 130 18968.42 ms 18963.34 ms 43.73 (9.34) Impulsive 119 15946.11 ms 15951.66 ms 43.94 (8.91)
Other outcome measures were evaluated for correct assumptions before ANCOVAs were conducted. All interactions were statistically significant, which indicates a violation of the homogeneity of regression assumption and hence no ANCOVA was conducted.
This violation indicates that the slope to predict outcomes from intelligence differs significantly for the rigid and impulsive groups and hence it would be incorrect to use the same regression slope to create adjustment means for the two groups.
The following brief statistics are offered for each of the respective outcomes: Attention and Working Memory (RT): F(1, 246) = 4.34, p =.038; Attention and Working Memory RT [Block 1]; F(1, 246) = 8.8, p =.003; Sensorimotor ability RT [Trial 1]; F(1, 246) = 4.5, p =.035 and Sensorimotor ability RT [Trial 2]; F(1, 246) = 6.67, p =.01.
We conducted the same but this time controlling for parental education.
Executive Domain Temperament Group Statistics Group N Unadjuste d scores Adjusted scores t value and p value Abstraction and Concept Formation (controlling for father s education) Reaction Time incorrect [Block 2] Rigid 96 2787.94 2785.06 t = 1.998 Impulsive 89 2445.07 2448.18 p = 0.047 Abstraction and Concept Formation (controlling for mother s education) Reaction Time incorrect [Block 2] Rigid 96 2787.94 2786.52 t = 1.98 Impulsive 89 2445.07 2446.61 p = 0.049 Attention and Working Memory (controlling for father s education) Reaction Time Rigid 131 439.59 439.68 t = 2.25 Impulsive 120 412.25 412.15 p = 0.025 Attention and Working Memory (controlling for mother s education) Reaction Time Rigid 131 439.59 439.30 t = 2.19 Impulsive 120 412.25 412.56 p = 0.029
Executive Domain Temperament Group Statistics Group N UnadjustedAdjusted scores scores t value and p value Attention and Working Memory (controlling for father s education) Reaction Time (Trial 1) Rigid 131 456.54 456.58 t = 2.69 Impulsive 120 416.83 416.78 p = 0.008 Attention and Working Memory (controlling for mother s education) Reaction Time (Trial 1) Rigid 131 456.54 455.95 t = 2.6 Impulsive 120 416.83 417.47 p = 0.01 Sensorimotor Ability (controlling for father s education) Reaction Time [Trial 1] Rigid 131 830.79 829.85 t = 2.36 Impulsive 121 744.64 745.66 p = 0.019 Reaction Time [Trial 2] Rigid 131 612.18 612.11 t = 2.93 Impulsive 121 575.44 575.51 p = 0.004
Executive Domain Temperament Group Statistics Group N Unadjusted Adjusted scores scores t value and p value Abstraction and Mental Flexibility (controlling for father s education) Reaction Time Rigid 130 18968.42 18968.42 t = 2.84 Impulsive 119 15946.11 15946.11 p = 0.005 Abstraction and Mental Flexibility (controlling for mother s education) Reaction Time Rigid 130 18968.42 19012.70 t = 2.93 Impulsive 119 15946.11 15897.73 p = 0.004
Discussion
To reiterate, this study attempted to explore differences in performance accuracy and reaction time between subtypes of temperament on computerised measures of abstract reasoning and executive functioning.
This study found significant differences in overall median reaction time on executive domains of attention and working memory; abstraction in executive functioning, abstraction and mental flexibility and sensorimotor processing.
Interestingly, no differences were found in accuracy or speed on the measure of verbal analogical reasoning. Researchers have reported the influence of a general factor such as intelligence and parental education levels on executive abilities. Based on these reported findings, an ANCOVA was conducted with SRAVEN, maternal and paternal scores as covariates.
When considering these potential moderating factors on reaction time on executive tasks, the significant differences between the groups were still observed.
The results of this study show that the Rigid subtype consistently performed slower than the Impulsive subtype in tasks that required conflict resolution strategies and cognitive control (complex abilities), and in tasks of attention and working memory task (simpler abilities) that required no conflict resolution.
Moreover, the faster reaction times of the Impulsive subtype were also evidenced across the complexity levels within tasks for example, in the LNB2-0 - back (easiest trial condition and does not involve memory per se); the 1- back (includes a memory load) and the 2-back (greater information load and longer delay from stimulus presentation to response). Based on the variations in task demands, effortful control, when considered as a facet of executive functioning, may allude to processes of response inhibition, and when considered as a facet of temperament may allude to self-regulation processes sub serving voluntary inhibition.
Cloninger (1987) contended that HA (avoidance system) inhibits NS (approach system) and this mechanism is consistently seen in the slower reaction times exhibited by the Rigid subtype across task demands and subcomponents of executive measures.
Conclusion
The contribution of this exploratory study is based on the findings that firstly, reaction time differences are significant when considering the executive functioning profiles of temperament subtypes and secondly, the reaction time differences across task complexity and structure hint at the multidimensional and multidetermined nature of constructs such as impulsivity and inhibition.
This study has some limitations. The tasks administered in the present study were limited to a select group of tests and single measures of specific facets of executive functioning. The sample was also relatively homogenous in age and gender, two sociodemographic aspects that limit the generalisalibility of results. Future studies should include a diverse sample and non-executive neuropsychological measures to determine if information processing differences between temperament subtypes are specific to neuropsychological domains.