Propiedades psicométricas de la Escala de Impulsividad (UPPS-P) en una muestra de adultos argentinos

Abstract

This study aimed to explore the psychometric properties of the Impulsive BehaviorScale (UPPS-P) in its short version in a sampleof Argentinean adults. Its factor structure andinternal consistency were explored, as well asits convergent validity with experiential andrational thinking styles, along with psychopathy.At the same time, the relationship ofimpulsivity with narcissism and Machiavellianismwas explored, controlling the scoresin psychopathy. An intentional sample ofArgentinean adults (N = 517) between 18 and60 years old (54 % female, mean age 27) wasselected. Participants responded on-line to thisscale, a sociodemographic inventory, as well asto the Rational Experiential Inventory and theDark Personality Scale. The results obtainedwere processed with the programs SPSS 23,Factor 8.10, and AMOS 16.0. The samplewas randomly divided into two groups of n= 217 and n = 300. An exploratory factorialanalysis was first performed on the first group,as a calibration study. After that, a confirmatoryanalysis was performed on the secondgroup, as a replication study, since there are noprevious studies on this subject in Argentina. Aparallel analysis was first carried out with theFactor 8.10 program. A number of replications= 100 and percentile representation of simulations=.95 were used. Regarding the exploratoryfactor analysis, The Unweighted LeastSquares method was used along with a Prominrotation. The final solution showed a structureof four factors: urgency, sensation seeking,lack of premeditation, and lack of perseverance,explaining 56 % of the variance. Each ofthese factors explained 16 %, 24 %, 10 %, and6 % of the variance, respectively. The urgencyfactor included both the positive and negativeurgency of the UPPS-P scale. The factor structurerevealed is similar to the UPPS, a scale ofimpulsivity prior to the UPPS-P, made up offour factors. All items on the short scale wereretained. Two items loaded better on a differentfactor. Cronbach’s Alphas were adequate, ensuring internal consistency. Cronbach’s α ofthe full scale was.79. Internal consistency wasanalyzed, resulting in Cronbach’s α values of.76 for Emergency;.75 for Sensation Search;.69 for Lack of Premeditation, and.74 forLack of Perseverance. The resulting Cronbachα values are similar to those of the Spanishversion, which presented internal consistenciesbetween.61 and.79. Subsequently, basedon the factor structure found, confirmatoryfactor analysis was carried out on the secondgroup sample. A relatively good fit was foundfor the model, with CFI and TLI values above.90 and the SRMR and RMSEA below.05.The authors’ original model, which includesfive factors, was also tested. The fit was lesssatisfactory compared to the previous model,because of the CFI, TLI, SRMR, and RMSEAindicators, and the difference in Δ from themodels, which was statistically significantat the p