Investigating human emotions empirically is still considered to be challenging, mostly due to the questionable validity of the results obtained when employing individual types of measures. Among the most frequently used methods to study emotional reactions are self-report, autonomic, neurophysiological, and behavioral measures. Importantly, previous studies on emotional responding have rarely triangulated the aforementioned research methods. In this paper we discuss main methodological considerations related to the use of physiological and self-report measures in emotion studies, based on our previous research on the processing of emotionally-laden narratives in the native and non-native language, where we employed the SUPIN S30 questionnaire as a self-report tool, and galvanic skin response (GSR) as a physiological measure (Jankowiak & Korpal, 2018). The findings revealed a more pronounced reaction to stimuli presented in the native relative to the non-native language, which was however reflected only in GSR patterns. The lack of correlation between GSR and SUPIN scores might have resulted from a number of methodological considerations, such as social desirability bias, sensitive questions, lack of emotional self-awareness, compromised ecological validity, and laboratory anxiety, all of which are thoroughly discussed in the article.

The relationship between self- and informant reports of personality using psychometric instruments is constantly the focus of attention for researchers in the field of clinical assessment in psychology. The research shows weak agreement between clinicians and patients’ assessments of personality disorders (PDs). The current study aimed at the convergence of measurement of PDs using the Shedler-Westen Assessment Procedure (SWAP-200), the self-report Character Styles Questionnaire-R (CSQ-R) and Borderline Personality Inventory (BPI). Paper-pencil questionnaires were administered to 102 inpatients (88.2% female, aged 18-64, M = 38.4) in a voivodeship hospital and outpatient health care centre. The SWAP-200 allowed us to gather expert (clinician) personality ratings basing on the intensive contact with patients. Results show that only a few SWAP-200 PD scales showed low positive correlations with corresponding self-reported PD scales from the CSQ-R. With the canonical correlation analysis, we identified two functions (borderline and internalising) that described similarities between the SWAP-200 and CSQ-R. SWAP-200 Obsessive-Compulsive PD correlated negatively with BPI scales. Consistent with previous studies, the self-report and the clinical assessment were only marginally convergent. Furthermore, OCPD stands out from other disorders in that it correlates positively with health indicators and negatively with some of the other personality disorders. The highest agreement was observed in the description of Borderline PD.