Widespread cognitive deficits may be seen in individuals with schizophrenia, particularly deficits associated with executive functioning.1 One facet of executive functioning that is of interest, particularly concerning its relationship with delusions, is reasoning ability. Ongoing work in the development of cognitive-behavioral treatment approaches for delusions and schizophrenia,2 which rely to some degree on the reasoning processes of the client, highlight the need for a thorough understanding of the nature of reasoning abilities in schizophrenia.
One model of delusion-formation posits that delusional thought may be the result of impaired reasoning, and, indeed, individuals with a greater degree of delusional ideation exhibit more reasoning biases during tasks of hypothesis-testing and probability judgments.3 However, it remains unclear whether such reasoning deficits are related specifically to delusional thought or are characteristic of all patients with schizophrenia. For example, one study of participants with schizophrenia found deficits in the use of context from various sources when creating global causative inferences, a deficit that appears related more to positive thought-disorder than to the negative symptoms of schizophrenia.4 Patients with schizophrenia and delusions have been shown to incorporate less information into their decision-making process, resulting in poorer judgments,5 but this “jump-to-conclusions” reasoning bias remains even after resolution of the delusional thought.6 Other researchers, however, have found few or no significant differences in reasoning ability between patients with schizophrenia and healthy-control subjects,7 between schizophrenia patients with and without delusions,8 and between patients with schizophrenia and other psychiatric participants.9 This heterogeneity of findings may be due in part to the lack of consensus regarding how to assess the function in question. For example, many of the studies that failed to find a reasoning deficit used tests of formal logic, tests on which even healthy respondents tend to perform poorly.10
The Word Context Test (WCT), a subtest of the Delis-Kaplan Executive Function System,11 was designed to assess inferential reasoning, or the ability to synthesize contextual and previous information in order to make reasoned judgments. Previous research has shown that, in individuals with schizophrenia, poorer overall performance on the WCT predicts decreased awareness of illness and need for treatment.12 However, because the study did not include a healthy-comparison group, it is unclear whether the performance of the patient group was actually impaired relative to the general population.
We used the WCT to further evaluate inferential reasoning in schizophrenia. We hypothesized that participants with schizophrenia-spectrum disorders (SCZ) would show greater deficits of inferential reasoning than healthy-control subjects (HC) on the overall achievement measure of the WCT. We also examined whether deficits of inferential reasoning were independent of other errors of executive functioning, namely perseveration and distractibility, by analyzing the error scores of the WCT. Finally, in order to further understand the relationship between performance on reasoning measures and specific symptoms of schizophrenia, we examined the relationship between WCT scores and the negative and positive symptoms of SCZ.
Participants included 15 individuals with a primary DSM-IV diagnosis of either schizophrenia (N=10) or schizoaffective disorder (N=5), and 15 HC subjects matched for gender (8 men and 7 women in each group) and parental education. Participants ranged in age from 20 to 55, with a significant difference in mean age between SCZ (M=45.07; standard deviation [SD]: 10.15) and HC (M=34.8; SD: 10.87; t =2.67; p=0.01] subjects. Psychiatric symptoms were assessed with the Schedules for Assessment of Negative and Positive Symptoms (SANS and SAPS).13,14 The mean group scores for SANS (M=10.80; SD: 4.36) and SAPS (M=6.13; SD: 3.02) were relatively mild in severity. Exclusion criteria included a history of neurological disorder, mental retardation, traumatic brain injury, current substance use disorder, and significant systemic medical illness. All patients were receiving stable doses of antipsychotic medications at the time of the study. After receiving a full description of the study, participants gave written informed consent and received financial compensation for their participation.
Participants completed a neuropsychological assessment that included the WCT. In this test, examinees are asked to decipher the meaning of a target word from an imaginary language through the use of five clue sentences, revealed one at a time, with each new sentence providing progressively more detail. The examinee gives one response after each sentence, and with each new bit of information can choose to either update their answer or continue giving the same answer. In order to minimize the role of working memory, the old clues continue to be shown to the participant. Successful performance on this test requires the ability to integrate the information from clue sentences in order to make reasonable inferences, as well as the cognitive flexibility to evaluate previous responses and test hypotheses.
The WCT has been found to have acceptable internal split-half and test–retest reliability, and to have moderate correlations with other executive functioning tasks, such as the Wisconsin Card-Sorting Test.11 The primary achievement measure is the Total Consecutively Correct (TCC) variable, based on the number of times a respondent provides a correct response consistently across consecutive clues. Two error measures, Repeated Incorrect Responses (RIR) and Correct-to-Incorrect Errors (CTI), were also used to assess perseveration and distractibility, respectively.
Analysis of covariance, using age as a covariate, revealed a significantly lower TCC score in the SCZ group (M=22.67; SD: 1.79) than the HC (M=29.73; SD:0.93; F [1, 27]=5.35; p=0.03). In contrast, groups did not show a statistically significant difference on RIR (F [1, 27]=2.25; p=0.15), with a mean RIR rate of 7.20 (SD: 1.08) for SCZ and 4.13 (SD: 0.58) for HC. Because of the small number of CTI errors, each group was subdivided into those who made ≥1 CTI errors (6 SCZ and 5 HC), and those who made no such errors (9 SCZ and 10 HC). Chi-square analysis showed no significant difference between groups (χ2=0.14; N=30; p=0.71).
We used partial correlation, controlling for the effects of age, to examine the relationship between TCC and the SANS and SAPS overall scores and subscale scores in the SCZ group; we found no significant correlations.
Compared with HCs, the SCZ group demonstrated poorer inferential reasoning. These participants made fewer consecutively correct responses, indicating impairment in the ability to integrate information in order to make deductions. Previously, some researchers have argued that individuals with delusions have a data-gathering bias and tend to seek less information, but that, when presented with the data, they retain the ability to use the data to make judgments.10 However, the WCT is designed so that data are presented to participants without their having to seek it out, and, in the current study, SCZ participants still showed deficits relative to HC participants on integrating information in a useful way.
Although the SCZ group achieved a lower score on the overall achievement measure of the WCT, both SCZ and HC participants made a similar number of errors of perseveration and distraction. This finding indicates that there may be a core deficit of inferential reasoning in schizophrenia-spectrum disorders that is independent from other aspects of executive functioning tapped by the WCT. Furthermore, despite previous suggestions that impaired reasoning in SCZ may be related to delusional thinking,3,5 no relationship was found between TCC and either positive or negative symptoms, including delusions. This finding is consistent with previous studies reporting independence of reasoning deficits and delusions,6,8 and, taken together, these results suggest that deficient inferential reasoning may be characteristic of patients with SCZ in general as opposed to only those with prominent delusions.
Because of the small sample size of the current study, replication with a larger sample is necessary. We must also be cautious in interpreting these results because they reflect only one measure of reasoning and may not generalize to all forms of reasoning and logic. Furthermore, a number of cognitive processes may contribute to successful reasoning, and a larger study would provide the opportunity to examine the specificity of the reasoning-deficit observed and any relationships between it and other executive or cognitive deficits commonly seen in SCZ. Also, although no correlation was found with any particular symptom in this study, this may be a reflection of the small sample size, the fact that average symptom scores on the SANS and SAPS tended to be relatively low overall and of limited variability, and/or potentially due to the fact that all patients were receiving stable doses of antipsychotic medication. Furthermore, data on patients' age at illness onset was unavailable, and thus it is not known whether chronicity of illness may have had an impact on the reasoning-deficit observed. The number of actual distractibility errors made by either group on the WCT was also modest, indicating that this may not be a highly sensitive measure of this function, and results should therefore be interpreted with caution. Finally, although group differences were present despite statistical control for age, further research with groups who are better matched on this variable will be important.
Strengths of the current study include the use of a measure that minimizes demands on working memory and parses out other influences, such as attention and perseveration. Also, although one previous study used the WCT in SCZ subjects,12 this is the first study to our knowledge that compared performance of such participants with a healthy-comparison group on this measure. Furthermore, despite previous failures to find a reasoning-deficit in SCZ,7,9 the current study provides novel evidence of an inferential-reasoning deficit that also appears to be independent of symptom severity, including delusions, adding further evidence to the hypothesis that there is a specific reasoning-deficit that is characteristic of schizophrenia-spectrum disorders.
Portions of this article were presented at the 28th Annual Meeting of the National Academy of Neuropsychology in New York, NY.
The authors thank Nancy Koven and Jo Cara Pendergrass for assistance with data collection.
This work was supported by grants from the Stanley Medical Foundation, the Hitchcock Foundation, and the Ira De Camp Foundation.
The authors declare no conflict of interest related to the content of this article.