Introduction: Voice emotional recognition (VER) is the ability to identify others’ emotion through their vocal expressions, which contains of 7 emotions: happiness, sadness, anger, disgust, fear, surprise, and calm. Patients with schizophrenia tend to have deficits of VER that affect their interpersonal interaction.
Objectives: The purpose of this study is to develop a computerized voice emotional recognition test (COVERT) and examine its psychometric properties.
Method: This study had three phases. (1) Forming an VER item bank. (2) Constructing the assessment system of the COVERT. (3) Examining the psychometric properties of the COVERT.
Results: We selected 168 items (24 for each emotion) from a validated data bank to construct the COVERT and confirm its content validity. Then, we recruited the 20 patients with stable symptom severities to complete the COVERT twice with a 2 week interval and 10 healthy adults to complete the COVERT once. The COVERT showed good test-retest reliability (intraclass correlation coefficient = 0.82), small random measurement error (percentage of standard measurement error = 2.2%), zero practice effect (Cohen’s d = 0), and satisfactory known-groups validity (Cohen’s d = -3.1).
Conclusion: The COVERT has the potential to become a reliable and valid test of VER in patients with schizophrenia.