Validation of a Korean Version of the Short UPPS-P Impulsive Behavior Scale for Children
Article information
Abstract
Objective
Impulsivity is a multifaceted construct that plays an important role in various problem behaviors in children and adolescents. The purpose of this study was to validate a Korean version of the short UPPS-P Impulsive Behavior Scale for Children.
Methods
Participants were 330 children (166 female) from 2 elementary schools in Korea and 94 attention deficit hyperactivity disorder (ADHD) children (23 female) from two major hospitals. The Korean short UPPS-P Impulsive Behavior Scale for Children (UPPS-P-C) (20 items), Child Behavior Checklist for Ages 6–18 (CBCL 6–18), and Barratt Impulsiveness Scale-11 (BIS-11) were administered. 107 children from the control group were retested 6 months later.
Results
Confirmatory factor analysis (CFA) conducted in the control group supported a 5-factor hierarchical model in which 1) positive and negative urgency factors are loaded on a higher-order factor of general urgency; 2) lack of perseveration and lack of premeditation factors are loaded on a higher-order factor of lack of conscientiousness; and 3) sensation seeking remained as a separate dimension. Reliability analysis demonstrated that the 5 factors of the Korean short UPPS-P-C had acceptable internal consistency and test-retest reliability. Lack of premeditation and lack of perseveration subscales showed significant correlations with measures of problem behaviors in CBCL and all the subscales were correlated with the BIS-11. The ADHD group showed significantly higher scores in lack of premeditation, lack of perseveration, positive urgency, and negative urgency subscales.
Conclusion
This study indicates that the Korean version of short UPPS-P-C has adequate reliability and validity. It may be a valid tool to assess impulsivity of healthy children as well as ADHD.
INTRODUCTION
Impulsivity is a multifaceted construct with significant implications in various problem behaviors and psychopathology among children and adolescents. The definitions of impulsivity are diverse, depending on different theoretical perspectives, and its sub-dimensions vary accordingly. Whiteside and Lynam 1 developed the urgency-premeditaion-perseverance-sensation seeking (UPPS) Impulsive Behavior Scale, which includes negative urgency, lack of premeditation, lack of perseverance, and sensation seeking, by conducting factor analyses on existing self-report impulsivity scales and the NEO-PI-R. Subsequently, Cyders et al. [2] introduced a fifth factor, positive urgency, thereby completing the UPPS-P (urgency-premeditaion-perseverance-sensation seeking-positive urgency) model. These five factors are closely related to Costa and McCrae’s five-factor personality model [3].
The UPPS-P model assess five dimensions of impulsivity. Firstly, negative urgency is characterized by the tendency to act impulsively under intense negative emotional states, whereas positive urgency represents the tendency to act impulsively in intense positive emotional states. Both these factors are conceptually associated with neuroticism in the five-factor model. Secondly, lack of premeditation refers to the tendency not to deliberate on the potential consequences before initiating an action, and lack of perseverance is defined as the inability to sustain attention on tedious or challenging tasks. These two dimensions are conceptually tied to conscientiousness within the five-factor model. Lastly, sensation seeking corresponds to extraversion in the five-factor model, characterized by the pursuit of exciting and thrilling activities, notwithstanding their potential risk [1,2]. The UPPS-P has been translated into various languages, including German [4], French [5], and Spanish [6], and its psychometric properties have been validated. In Korea, Lim and Lee [7] translated and validated a 59-item version of the UPPS-P, and an additional short version consisting of 20 items was subsequently developed [8].
Impulsivity is a factor that is widely associated with various psychiatric disorders in children and adolescents, with attention deficit hyperactivity disorder (ADHD) being the most frequently diagnosed disorder where impulsivity is a prominent concern. However, research focusing on impulsivity in children, based on the UPPS-P model, is relatively scarce when compared to studies conducted on adults. For children, there exists a 59-item version developed by Zapolski and Smith [9] and Zapolski et al. [10], as well as a 20-item version developed by Geurten et al. [11] based on the adult short version of Billieux et al. [12] In Geurten et al.’s study, 11 a five-factor structure was identified through confirmatory factor analysis (CFA), and the lack of premeditation, along with two urgency subscales, significantly discriminated between ADHD and the control group.
In the validation study of the Japanese version of the short UPPS-P Impulsive Behavior Scale conducted by Hasegawa et al. [13], a five-interrelated factor model demonstrated the best fit. Similarly, a study involving Italian youth with the short UPPS-P [14] discovered that the five-factor model remained invariant across gender and age. A validation study was also undertaken with Turkish adolescents on the short UPPS-P [15], wherein the lack of premeditation, positive urgency, and negative urgency subscales effectively distinguished between ADHD adolescents and the control group. Since impulsive children typically have shorter attention spans and easily lose motivation during testing sessions, it becomes challenging to administer self-report assessments with a large number of items. Furthermore, children often show difficulty in completing boring and unattractive tasks such as long self-reports. Consequently, shorter versions are more commonly preferred and utilized across various countries.
Currently, there is no translated version of the short UPPS-P scale for children available in Korea, and any Korean version of UPPS-P scale for children does not exist. The aim of this study was to translate the items from Geurten et al. [11] into Korean, validate its reliability and validity, and assess its clinical utility by comparing ADHD and control groups.
METHODS
Participants and procedure
Participants in the control group comprised 330 children (164 male, 166 female), aged between 10 to 12 years (11.02±0.79), enrolled in two elementary schools in Korea. Among this sample, 107 children were retested 6 months later. The Korean versions of the short UPPS-P Impulsive Behavior Scale for Children (UPPS-P-C) (20 items) [11] and Child Behavior Checklist for Ages 6–18 (CBCL 6–18) [16] were administered in both waves, and the Barratt Impulsiveness Scale-11 (BIS-11) [17] was added in the second wave. The ADHD group consisted of 94 children (71 male, 23 female) recruited from the psychiatry departments of two major hospitals in Seoul, Korea, aged between 7 to 12 years (8.85±1.61). This group completed the Korean version of the short UPPS-P-C during their psychological assessment, with official diagnoses made by staff psychiatrists after the assessment.
Written informed consent was obtained from all children and parents in the control group. The children in the ADHD group and their parents provided informed consent about utilizing the test results for research purpose prior to the psychological assessment. This study was approved by the Institutional Review Board of Dongduk Women’s University (DDWU1905-3), Severance Hospital (4-2021-0218) and Sanggye Baik Hospital (SGPAIK2021-06-024).
Measures
Korean version of the short UPPS-P-C
The short UPPS-P-C consists of 20 items rated on a 4-point Likert scale, measuring five dimensions of impulsivity: lack of premeditation (4 items), lack of perseverance (4 items), positive urgency (4 items), negative urgency (4 items), and sensation seeking (4 items) [11]. Geurten and colleagues [11] provided the items in both French and English, and the Korean version was translated from the English items. One of the authors of this study, a bilingual clinical psychologist, initially translated the items from English into Korean. Subsequently, the other authors revised a few ambiguous and awkward items. This preliminary version was administered to 10 children (8 male, 2 female), and Cronbach’s alpha was calculated for the five subscales. Most of the scales demonstrated adequate internal consistency, ranging from 0.61 to 0.93, comparable to the results found in Geurten et al.’s study [11], except for the negative urgency subscale (Cronbach’s alpha 0.56). The authors reviewed and modified some expressions within the negative urgency subscale to enhance understanding for children. Korean version of the short UPPS-P Impulsive Behavior Scale for Children is provided in Supplementary Material (in the online-only Data Supplement).
Korean version of the CBCL 6–18
The Korean version of CBCL 6–18 [16] is a parent-reported questionnaire assessing behavioral and emotional problems of school-age children. It was originally developed by Achenbach and Rescorla [18]. The empirically based syndrome scales comprise 120 items, divided into three internalizing problem scales (anxious/depressed, withdrawn/depressed, somatic complaints), two externalizing problem scales (rule-breaking behavior, aggressive behavior), and four additional scales (social problems, thought problems, attention problems, and other problems). Each item is rated on a 3-point scale ranging from 0 (not applicable at all) to 2 (often), with higher scores indicating a greater tendency toward problem behaviors. In the Korean standardization study, the Cronbach’s alpha for each scale ranged from 0.62 to 0.84. For this study, we utilized the raw scores of the total problems, internalizing problems, externalizing problems, attention problems, rule-breaking behavior, and aggressive behavior scales in the correlational analysis.
Korean version of the BIS-11
The BIS-11 is a 30-item self-report questionnaire that measures impulsiveness. It was developed by Patton et al. [19], and the Korean version was validated by Heo et al. [17] Items are rated on a 4-point Likert scale, ranging from 1 (rarely/never) to 4 (almost always/always). The BIS-11 consists of six first-order factors (attention, cognitive instability, motor, perseverance, self-control, cognitive complexity) and three second-order factors (attentional, motor, and non-planning impulsiveness). In this study, we used raw scores of the three second-order factors and the total score. The Cronbach’s alpha for the three factors in the Korean validation study ranged from 0.61 to 0.70, and was 0.80 for the total score [17].
Data analysis
First, we performed a CFA to examine the factor structure of the Korean short UPPS-P-C using Mplus (version 8.3; Muthén & Muthén, LosAngeles, CA, USA) [20]. Prior to performing CFA, Little’s MCAR test was executed to ensure that the data were missing at random [21]. The result of this test was non-significant (χ2 (358)=389.97, p=0.118), leading us to utilize the maximum likelihood estimation in the CFA. The following four models were tested: 1) a model representing a single impulsivity factor; 2) a 3-factor model of urgency, conscientiousness, and sensation seeking; 3) a 5-intercorrelated factor model including lack of premeditation, lack of perseveration, positive urgency, negative urgency, and sensation seeking; 4) a 5-factor hierarchical model wherein the higher-order construct of “urgency” encompasses negative urgency and positive urgency, and the higher-order construct of “conscientiousness” encompasses lack of premeditation and lack of perseveration, leaving sensation seeking as an independent construct.
In the reliability analysis, internal consistency coefficient of Cronbach’s alpha and the test-retest correlation were computed. Correlation analyses were carried out with the subscales of the Korean short UPPS-P-C, CBCL 6–18, and BIS-11 across two waves. Lastly, an analysis of covariance (ANCOVA) was performed to compare the scores of the Korean short UPPS-P-C between the ADHD and control groups. Since a significant difference in age between the two groups was observed (t=12.59, p<0.001), age was employed as a covariate in the ANCOVA. All statistical analyses were conducted using IBM SPSS Statistics (version 26; IBM Corp., Armonk, NY, USA).
RESULTS
CFA
The fit indices for the four models tested in the CFA are presented in the supplement (Supplementary Table 1 in the online-only Data Supplement). Both Model 3 (Tucker-Lewis Index [TLI]=0.903, comparative fit index [CFI]=0.918, root mean square error of approximation [RMSEA]=0.050) and Model 4 (TLI=0.905, CFI=0.919, RMSEA=0.050) demonstrated acceptable fits to the data. However, Model 4, with two higher-order factors and an independent factor of sensation seeking, exhibited slightly higher TLI and CFI values. Additionally, the chi-square difference test was not significant (Δχ2(3)=1.732, p>0.05), suggesting that Model 4, with its higher degree of freedom, can be accepted as the best model for this study. The final 5-factor hierarchical structure of the Korean short UPPS-P-C is depicted in Figure 1. In this final model, the lack of premeditation and lack of perseveration are loaded onto a higher-order factor of “conscientiousness,” whereas positive urgency and negative urgency are loaded onto a higher-order factor of “urgency.” Meanwhile, sensation seeking remains as an independent and distinct factor.
Descriptive statistics and reliability analysis
Table 1 offers the descriptive statistics and reliability coefficients for five subscales and the total score of the Korean short UPPS-P-C. The normality of the scale was found to be satisfactory, as the skewness (0.059–0.421) and kurtosis (-0.629–0.420) indices meet the normality requirements [22]. The internal consistency coefficients (Cronbach’s alpha) for the subscales and total score varied from 0.62 to 0.80, and the test–retest correlations at 6 months interval ranged from 0.53 to 0.70.
Correlations
Table 2 presents the correlations between the subscales of the Korean short UPPS-P-C. All subscales, except for sensation seeking, exhibited significant positive correlations with one another. The sensation seeking subscale was not correlated with the lack of premeditation and lack of perseveration, although it was significantly correlated with the two urgency subscales.
Table 3 shows the correlations among the Korean short UPPS-P-C, CBCL 6–18, and BIS-11 across two waves. In wave 1, lack of premeditation, lack of perseveration, and sensation seeking subscales were significantly correlated with internalizing/externalizing problems, attention problems, rule-breaking behavior, and aggressive behavior scores from CBCL 6–18. Negative urgency demonstrated significant correlations with all CBCL 6–18 scores except for attention problems in wave 2. However, positive urgency did not show significant correlations with CBCL 6–18 scores in wave 1 but was positively correlated with total problems, externalizing problems, and aggressive behavior in wave 2. Sensation seeking was only correlated with externalizing problems in wave 1. In both waves, the Korean short UPPS-P-C exhibited significant positive correlations with subscales of BIS-11.
Comparison of ADHD and control group
Lastly, Table 4 presents the results of the ANCOVA comparison between the ADHD and control groups. Children diagnosed with ADHD showed significantly higher scores in all subscales of the Korean short UPPS-P-C, except for sensation seeking. No group difference was found in sensation seeking (F=0.49, p>0.05).
DISCUSSION
The purpose of this study was to translate and validate the Korean version of the short UPPS-P-C, focusing on assessing its validity and reliability as a measure of impulsivity, a major factor in various child and adolescent psychopathologies. Developing a valid instrument to measure different facets of impulsivity is highly valuable. This study’s findings affirm that the Korean version of short UPPS-P-C maintains a theory-consistent factor structure and stable reliability. Correlations with CBCL 6–18 and BIS-11 underscore that the Korean short UPPS-P-C is an effective tool for evaluating the multidimensional construct of impulsivity and associated problematic behaviors. Furthermore, children diagnosed with ADHD scored significantly higher on most subscales of Korean short UPPS-P-C, highlighting its clinical utility.
The results of the CFA indicated that both a 5-intercorrelated factor model and a 5-factor hierarchical model fit the data well. However, the hierarchical model was chosen as the final model, as it offered a greater degree of freedom without a significant chi-square difference test. This finding aligns with Lim and Lee’s [7] validation study of Korean UPPS-P in an adult sample and is supported by other previous studies [12,15]. The hierarchical model concurs with Whiteside and Lynam’s theoretical proposal that negative/positive urgency correlates with the neuroticism factor of the five-factor personality model (FFM), while lack of premeditation/perseverance associates with conscientiousness [1,3]. Sensation seeking is related to excitement seeking of extraversion in the FFM. The 5-factor hierarchical model of this study robustly supports this theoretical stance. Recent research has supported the hierarchical model of impulsivity in both adolescent and adult samples using UPPS-P [23-25]. Nevertheless, some studies have demonstrated that the 5-intercorrelated factor model may provide a better fit for child and youth versions [11,14,26]. It seems that both the 5-intercorrelated factor model and the 5-factor hierarchical model are consistent with the commonly observed structure across diverse age samples.
The internal consistency coefficients for the five subscales in this study ranged from 0.62 to 0.78, values that are analogous to those reported in Geurten et al.’s study [11]. This finding is also comparable to the reliability analysis of the short version of the Korean UPPS-P for adults [8]. In contrast, the test–retest correlations observed over a 6-month interval in our study, ranging from 0.43 to 0.64, were somewhat lower than those documented in Geurten et al.’s research [11]. However, it is essential to consider that the measurement interval in Geurten et al.’s study [11] was substantially shorter, spanning 6–8 weeks, making the test-retest reliability observed in this study appear acceptable, with all p values <0.01.
The validation of the Korean adaptation of the short UPPS-P-C was examined through correlation analysis with other widely used scales concerning problem behaviors and impulsivity. Correlations with parent-reported CBCL 6–18 revealed that lack of premeditation, defined as a failure to consider consequences before initiating an action, is significantly associated with internalizing/externalizing problems and attentional problems across both wave 1 and wave 2. This highlights the importance of a child’s capacity to contemplate the consequences of their actions to prevent various problematic behaviors. Furthermore, lack of perseveration, an inability to endure a tedious or challenging task, was also related to most problem behaviors, especially attentional problems. However, positive urgency and negative urgency yielded some inconsistent outcomes. Negative urgency was significantly related to all problem behaviors in waves 1 and 2, except for attention problems in wave 2. In contrast, positive urgency displayed no significant correlations with problem behaviors in wave 1 but did show associations with externalizing problems and aggressive behaviors in wave 2. Negative and positive urgency describe the tendency to act impulsively under negative or positive moods, respectively. These findings are consistent with previous research, suggesting that negative urgency is linked to a broad range of problem behaviors and psychopathology, both concurrently and prospectively [27,28]. However, considering that positive urgency was only associated with externalizing problems in wave 2, it is possible that the urgency factor, closely tied to the neuroticism of FFM [3], may not be so stable in its relationship with various problem behaviors during childhood. At the same time, there is a possibility that positive urgency might be a more predictive factor for future externalizing behaviors, particularly in terms of deficits in behavioral inhibition compared to negative urgency [29]. Sensation seeking, defined as the tendency to seek thrilling activities regardless of potential danger, was only significantly correlated with rule-breaking and aggressive behaviors in wave 1. This finding is consistent with a previous study that investigated the validity of an abbreviated youth version of the UPPS-P, where sensation seeking was solely related to externalizing problems [26]. Lastly, the correlations with the subscales of BIS-11 affirmed that the Korean version of the short UPPS-P-C is a valid instrument to evaluate the multidimensional facets of impulsivity.
We compared the subscale scores of the Korean version of short UPPS-P-C between the ADHD group and the control group to investigate the discriminant power of the scale. The findings indicated that the ADHD group had higher scores in four out of five subscales, with sensation seeking being the exception. Based on a meta-analytic review concerning the behavioral implications of the UPPS [27], sensation seeking was found to have a near-zero negative association with ADHD. ADHD pathology seems to be more closely associated with two of the “lack” subscales, both of which consistently correlate with behavioral tasks [30,31]. Moreover, it is noteworthy that sensation seeking is a genetically distinct facet from other dimensions of impulsivity [32].
Limitations
Several limitations should be considered in interpreting the findings of the present study. Firstly, the translation of the items from the English version presented in Geurten et al.’s [11] work raised concerns as they validated the scale using the French item. As a result, a comparison with the validation studies of the English version of short UPPS-P-C in children is necessary. Secondly, the inclusion of children aged 10 to 12 was necessary to obtain reliable self-report data, but this age range limits the generalization of the study’s findings to all school-age children. Thirdly, the exclusive focus on the ADHD group to assess the discriminant validity of the scale neglects the broader spectrum of psychopathologies connected to impulsivity, such as gambling, eating disorders, and various addictive behaviors. Inclusion of diverse patient groups would further delineate the clinical utility of the UPPS-P-C. Lastly, the utilization of the Korean version of short UPPS-P-C as a self-report questionnaire demands comparison to other impulsivity measures, like behavioral tasks or parent/teacher-report assessments, to ensure its validity.
Conclusions
Despite the aforementioned limitations, this study demonstrates that the Korean version of the short UPPS-P-C serves as a reliable and valid assessment tool for measuring multidimensional impulsivity in children. The findings from the CFA support a 5-factor hierarchical model, and the reliability analysis revealed satisfactory internal consistency as well as test-retest stability. Furthermore, the subscales of the Korean version of the short UPPS-P-C were found to correlate with relevant measures of problem behaviors and impulsivity. Notably, it significantly differentiated between the ADHD diagnosis group and the control group, thus underlining its practical application in clinical settings.
Supplementary Materials
The online-only Data Supplement is available with this article athttps://doi.org/10.30773/pi.2023.0277.
Notes
Availability of Data and Material
The datasets generated or analyzed during the study are available from the corresponding author on reasonable request.
Conflicts of Interest
The authors have no potential conflicts of interest to disclose.
Author Contributions
Conceptualization: all authors. Data Curation: Ji Hyeon Kang, Min-Jung Shin, Jeon Hoon Park, Ji Young Choi. Formal analysis: Joo Young Lee. Funding acquisition: Ji Young Choi. Investigation: Joo Young Lee, Ji Hyeon Kang, Dong Ho Song, Ji Young Choi. Methodology: Joo Young Lee, Ji Young Choi. Project administration: Ji Young Choi. Supervision: Ji Young Choi. Validation: Ji Hyeon Kang, Min-Jung Shin, Jeon Hoon Park, Dong Ho Song, Ji Young Choi. Writing—original draft: Joo Young Lee. Writing—review & editing: Ji Hyeon Kang, Min-Jung Shin, Jeon Hoon Park, Dong Ho Song, Ji Young Choi.
Funding Statement
This study was supported by the Jisan Cultural Psychiatry Research Fund from the Korean Neuropsychiatric Association.