Action-based cognitive remediation (ABCR) will be delivered in group sessions one hour per week. ABCR was developed by Dr. Bowie (CI) and Dr. Best (PI) and has been found efficacious for schizophrenia-spectrum disorders in three clinical trials. ABCR involves practicing computerized training exercises with difficulty level dynamically titrated to improve neurocognitive abilities. Then participants engage in strategy discussions with other group members to develop new cognitive strategies. Finally, participants complete role-play simulations of real-world activities to practice their cognitive strategies in simulations of everyday life. ABCR is more effective for improving functioning than traditional approaches to CR. Homework consists of additional cognitive training and practicing cognitive strategies in everyday life.

Befriending will be delivered according to a manual validated to control for the non-specific effects of CBT, such as duration of therapeutic contact, client expectancy effects, therapeutic alliance, and therapist warmth. Befriending consists of 1-hour individual sessions once per week and involves a series of conversations similar to those one might have with a social acquaintance. These conversations involve discussion of neutral topics without problem-solving, coping strategies, or exploration of emotion. If emotional or mental health-related topics are brought up therapists redirect back to a neutral topic.

Sham CR was developed by Dr. Best and Dr. Bowie (CI) to control for the non-specific effects of CR such as computer practice and group discussion. Participants practice similar computerized exercises to ABCR, however, the exercises do not increase in difficulty. Participants then discuss enjoyment of the exercises but any discussion of cognitive strategies is redirected back to a neutral topic. We have previously found this condition to be an effective control for CR, with similar engagement to the active training group.

QPR is a self-report measure assessing recovery with people experiencing psychosis. This version contains 22 items while the response to each statement is scored on a 5-point Likert scale ranging from "0 = strongly disagree" to "4 = strongly agree". The lowest possible score is 0 and the highest score could be 88. Higher scores would indicate better recovery.

The PANSS is a 30-item semi-structured interview assessing positive, negative symptoms and general psychopathology. Each item is scored on a 7-point scale (1 = absent, 2 = minimal, 3 = mild, 4 = moderate, 5 = moderate severe, 6 = severe, 7 = extreme). The lowest score would be a 30 and the highest score would be 210. A higher score would indicate increased symptomology. A 15-point change on the PANSS is associated with clinically meaningful change.

The PSYRATS assesses frequency and distress associated with the experiences of auditory hallucinations and delusions based on the PANSS interview. Each of the 17 items is scored on a 5-point scale, where a score of 0 indicates no presence, and 4 indicates the highest severity. The lowest score would be a 0 and the highest score would be 68. A higher score would indicate increased symptomology.

Neurocognition will be assessed using a cognitive batter that assess the MATRICS cognitive domains. Assessments include Oral Trail Making Test, Hopkins Verbal Learning Test, Letter Number Span, WMS Symbol Span, Brief Visuospatial Memory Test, DKEFS Verbal Fluency, Hinting Task, CNS Vital Signs Symbol Digit Coding, CNS Nonverbal Reasoning, CNS 4-part CPT. A Neurocognitive Composite score will be calculated based on these subtests. Z-scores will range between -3.00 and +3.00, with higher scores indicating better performance.

The BAPS is a 31-item self-report measure assessing metacognitive beliefs about paranoia. The degree of agreement to each statement is scored on a 4-point scale (1 = not at all, 2 = somewhat, 3 = moderately so, 4 = very much). Scales include positive, negative and normalizing beliefs about paranoia, and paranoia as a survival strategy. The lowest score would be 31 and the highest score would be 124. Higher scores are indicative of more beliefs and are shown to be related to paranoid ideation.

The IVI is a 26-item self-report measure assessing perceptions of auditory hallucinations. Meta-physical beliefs about voices, positive beliefs about voices, and beliefs about loss of control are evaluated. Responses are rated on a 4-point scale (1 = Not at all, 2 = Somewhat, 3 = Moderately so, 4 = Very much). The lowest score would be a 26 while the highest score would be a 104. Higher scores are indicative of more and stronger beliefs.

The BCSS is a 24-item self-report measure assessing positive and negative judgments individuals hold about themselves and others. Responses are first given dichotomously as "no" or "yes". "No" is scored as 0 and if the answers are "yes", the intensity of beliefs are then rated on a 4-point scale (1 = believe it slightly, 2 = believe it moderately, 3 = believe it very much, 4 = believe it totally). The lowest score would be a 0 and the highest score would be a 96. Higher scores in the positive-self subscale indicate more positive beliefs about selves, while higher scores in the negative-self subscale indicate more negative beliefs about selves. Higher scores in the positive-others subscale indicate more positive beliefs about others, while higher scores in the negative-others subscale indicate more negative beliefs about others.

The rating format for the 80 Dysfunctional Attitudes Scale items is a 7-point Likert scale ranging from totally agree to totally disagree. Possible responses are scored from 1 to 7, with the direction depending on whether agreement or disagreement with a particular belief is judged to be a maladaptive response, and with higher scores indicating more distorted thinking.

The TiC-P measures direct and indirect costs of health conditions.

The first part contains 21 questions assessing healthcare costs, calculated by multiplying numbers of access with prices.

The second part contains 12 questions assessing productivity losses. For paid work, productivity losses due to absence are measured by cost prices, with the number of workdays multiplying hours per day and a standard hourly rate. Scores on short-term absence are adjusted with long-term absence to avoid double-counting, while the friction period and measurement times are considered for scores on long-term absence. Productivity losses due to presenteeism are calculated by multiplying the number of workdays impaired with (1 - efficiency score/10) and hours per workday. Unpaid productivity losses are calculated by multiplying the days missed by the number of hours per day needed to make up the work, and a standard hourly rate.

Higher scores indicate higher costs related to health conditions.

The SSTICS is a 21-item self-report measure assessing subjective cognitive complaints among patients with Schizophrenia. Frequency of experiencing each item is rated on a 5-point scale, ranging from "0 = Never" to "4 = Very Often". The lowest score would be a 0 and the highest score would be an 84. Higher scores are indicative of more subjective cognitive difficulties.

DACOBS is a 42-item self-report inventory assessing cognitive processing biases associated with psychosis. The degree of agreement to each statement is scored on a 7-point Likert scale, ranging from "1 = strongly disagree" to "7 = strongly agree". The lowest score would be a 42 and the highest score would be a 294. Higher scores would indicate more cognitive biases.