Degree Name

Doctor of Philosophy (Clinical Psychology)


School of Psychology


Background: There are two primary definitions of insight used in research. The first, ‘clinical insight’, concerns an accurate understanding of illness, the need for treatment for that illness, and the consequences of the illness being untreated. Cognitive insight, a more recent definition, places emphasis on capacity to reflect on one’s thinking and integrate multiple viewpoints. Insight research is primarily focused on people living with schizophrenia or other psychotic illnesses, however, has also expanded to other mental health disorders. More recently, insight has been explored in people living with a substance use disorder. Clinical insight has been the dominant definition used within the substance use literature, but there are potential confounds between aspects of clinical insight (e.g., problem awareness) and the self-report methods typically used in assessing substance use. For accurate reporting of substance use and its consequences, people need to have good clinical insight. To reduce the impact of such confounds, examining the function of cognitive insight in substance use may be beneficial. A prominent measure of cognitive insight is the Beck Cognitive Insight Scale, made up of two subscales. Self-Reflection concerns the ability to reflect on past experiences and consider different explanations. Self-Certainty focuses on one’s confidence in interpretation and meaning making. In essence, good cognitive insight is made up of high Self-Reflection and low Self-Certainty. To date, cognitive insight has not been extensively studied in people living with a substance use disorder. The aims of this thesis are to examine the profile of cognitive insight in people with a substance use disorder using the BCIS, and explore if there is a relationship between cognitive insight and treatment outcomes for people accessing treatment for substance use disorder.

Methods: Study 1 was a systematic review, comprised of 20 empirical studies examining clinical or cognitive insight in substance use disorder. Study 2 was a psychometric study that aimed to validate the BCIS for people living with a substance use disorder. The sample (n=150) were people attending residential treatment for their substance use disorder. Study 3 was a longitudinal study that explored the relationship of cognitive insight to medication adherence and methamphetamine cessation. The sample consisted of 152 people who were screened for current methamphetamine dependence. Participants were administered the BCIS at the start and end of an encompassing three-month medication trial.

Results: Study 1 found that there is little consistency with regard to how insight is being measured in people with a substance use disorder. From 20 studies included in the review, there were 13 different insight measures. Validation of the measures was also lacking, with few studies providing evidence of construct validity. The most common conceptualization of insight was clinical (i.e., awareness of the substance use problem and its consequences). It was also found that there was a consistent relationship between clinical insight and worse substance use (e.g., frequency, severity) and consequences of use. Factor-analysis conducted in Study 2 confirmed a two-factor structure for the BCIS and after exclusion of two of the original items, it was deemed suitable for use for people with substance use disorder. This study also found similar relationships between BCIS scores and psychological distress as had been found in prior research with samples with other mental health disorders. Study 3 found the BCIS scales did not predict either medication adherence nor reduction in methamphetamine use. Contrary to prediction, Self-Reflection decreased over the study period. Self-Certainty increased in those who maintained or increased their substance use.

Discussion: Clinical insight has almost exclusively been studied within substance use disorder. However, cognitive insight may have clinical utility, by reducing emphasis on problem awareness. The results in this thesis indicate that the BCIS may not be the most appropriate measure, particularly for people who are currently using substances. Implications for conceptualization and measurement of insight, and application to clinical practice are discussed.

FoR codes (2008)

170106 Health, Clinical and Counselling Psychology



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.