Abnormal numerical processing in math-anxious individuals: Evidence from event-related brain potentials

• Autores: Macarena Suárez Pellicioni
• Directores de la Tesis: María Isabel Núñez Peña (dir. tes.), Elisabet Tubau Sala (dir. tes.), Angels Colomé González (dir. tes.)
• Lectura: En la Universitat de Barcelona ( España ) en 2014
• Idioma: inglés
• Tribunal Calificador de la Tesis: Carles E. Escera Micó (presid.), Luis Carretié Arangüena (secret.), Dénes Szücs (voc.)
• Materias:
  • Psicología
    • Psicología experimental
      • Actividad cerebral
      • Emoción
• Enlaces
  • Tesis en acceso abierto en:  TDX  Dipòsit Digital de la UB 
• Resumen

  • This PhD thesis comprises five studies aiming to investigate differences between low math- anxious (LMA) and high math-anxious (HMA) individuals in numeric processing by means of behavioral and event-related potential (ERPs) measures. The excellent temporal resolution of the ERP technique was expected to provide detailed information that would shed light about the difficulties HMA individuals face when they have to deal with numbers. The first study aimed to adapt into Spanish and validate the Shortened Mathematics Anxiety Rating Scale (sMARS; Alexander & Martray, 1989) as a starting point of this thesis, in order to make sure that the construct of math anxiety (MA) was going to be assessed with an instrument providing valid and reliable measures. The adaptation into Spanish of the sMARS scale gave sound evidence of its good psychometric properties: strong internal consistency, high 7-week test-retest reliability and good convergent/discriminant validity. Study II aimed to investigate, with the ERP technique, the use of the plausibility strategy in math-anxious individuals by studying Faust et al. (1996)´s finding on flawed scores for dramatically incorrect solutions (large-split) in an arithmetic verification task. We were able to replicate, for the first time, those findings, by finding a greater percentage of flawed scores for large-split solutions for the HMA group as compared to the LMA one. Moreover, ERP analysis showed that large-split solutions generated a P600/P3b component of larger amplitude and delayed latency for the HMA group as compared to the LMA one. Given the functionality of this component, this finding suggested that large-split solutions demanded more cognitive resources and required more time to be processed for the HMA group than for the LMA one. These findings were interpreted according to the Attentional Control Theory (ACT; Eysenck, Derakshan, Santos, & Calvo, 2007): HMA individuals, being more influenced by the stimulus-driven attentional system, would have succumbed to the distractor nature of the large-split solution, devoting more time (P600/P3b latency) and cognitive resources (P600/P3b amplitude) to process this clearly wrong solution, instead of using the plausibility strategy. Study III consisted of finding the electrophysiological correlates of numeric interference in LMA and HMA individuals, by means of a numeric Stroop task. We found that HMA individuals needed more time to solve this task as compared to their LMA peers, suggesting that they were distracted by the task-irrelevant dimension of the stimuli (i.e. physical size of numbers). ERP data analysis showed that LMA and HMA individuals differed in the way they adapted to conflict: the LMA group presenting a greater N450 component for the interference effect preceded by congruence than when preceded by incongruity while the HMA group showed the same enhancement but for the subsequent Conflict sustained potential. These results suggested that both groups showed a different implementation of attentional control, which was executed in a proactive way by LMA individuals and in a reactive way by HMA ones. A reactive recruitment of attentional control in HMA individuals would have made them more influenced by bottom-up input (i.e. stimulus-driven attentional system), making them more vulnerable to distraction. The two remaining studies of this PhD thesis aimed to explore two possible factors contributing to the development of MA. Given that errors are crucial for mathematical learning, because of its cumulative nature, one concept building on the next, Study IV aimed to assess whether LMA and HMA individuals differed in the way they processed a numeric error as compared to a non- numeric one. We found that HMA individuals showed an increased error-related negativity (ERN) when they committed an error in the numeric Stroop task, but not in the classical Stroop task. Furthermore, standardized low resolution electromagnetic tomography (sLORETA) analysis showed significant greater voxel activation at the right insula for the errors committed in the numerical task as compared to the classical one for the HMA group and not differences at all for the LMA one. Given that the right insula has been associated with the discomfort with one´s own physiological responses and given that errors are considered to generate a cascade of physiological responses, this finding suggests that HMA individuals´ may have experienced a discomfort with the physiological responses generated by a numeric error. This negative bodily reaction towards numeric errors may be at the base of the development of negative attitudes towards mathematics and of the tendency of HMA individuals to avoid math-related situations. Finally, Study V aimed to investigate, by means of an emotional Stroop task, whether MA is characterized by an attentional bias towards math-related information, given that an attentional bias towards threatening information is considered to be a contributory factor in the origin and maintenance of several types of anxiety. This study showed that HMA individuals showed a clear tendency of responding slower to math-related words as compared to neutral words. Given that this slowdown in an emotional Stroop task has traditionally been interpreted as an attentional bias towards threatening or emotional stimuli, this study demonstrates that MA is also characterized by an attentional bias, in this case, towards math-related words, which could probably be at the base of its development and maintenance. To sum up, this PhD thesis has shown that MA is characterized by a vulnerability to distraction, which was shown when a large-split solution was presented for a simple addition task (Study II) and when physical size interfered with numerical magnitude in a numeric Stroop task (Study III). Moreover, HMA individuals also showed a reactive recruitment of attentional control after conflict detection (Study III), a greater sensitivity or emotional response to numeric errors (Study IV) and a clear tendency of an attentional bias towards math-related stimuli (Study V).