INTRODUCTION
Social cognition has been defined as “the mental processes underlying social interactions, including the perception and evaluation of the intentions, tendencies, and behaviours of others, and producing responses to these” [1]. It can be understood from this definition that when a process related to social cognition occurs, there is a series of mental processes related to the event/situation. These evaluations include: 1) perception of the social information related to the event/situation, 2) understanding the cognitive and affective status of the other person, and 3) making a decision considering the aims of the other person [2,3]. Therefore, for social cognition to able to emerge successfully, there is a need for the cognition and neurocognition components of perception, attention, and problem-solving to work perfectly [4,5].
Social cognition is multi-dimensional and includes many components such as theory of mind, mentalisation, emotion processing, social attribution, and social processing [6,7]. This study focuses on the social attribution skill, which is an important component of social cognition. Social attribution was first described and examined by Heider and Simmel [8]. Heider described social attribution as “person perception,” referring to how an individual explains the behaviour of others and what meaning they give to it, rather than the actual reasons for the behavio [9]. A model has been recommended in which social attribution is evaluated in the context of intentional and unintentional behaviours, where intentional behaviour is defined as intention. Malle [10] stated that there were six assumptions in this model, defined as the assumptions and distinctions that an individual uses when attributing behaviour to themselves or others.
Consequently, with the importance of social attribution skills in both the cognitive psychological context and in the evaluation of clinical skills there is a need for tools with proven validity and reliability. The Social Attribution Test (SAT), developed by Klin [11], may be able to fulfil this need as it uses ambiguous visual representations with no verbal clue [12]. SAT aims to evaluate the skill of an individual in giving social meaning to ambiguous visual stimuli. The test material is the Heider and Simmel [8] animation portraying a social drama using mobile geometric shapes. The use of a silent animation reduces the need for auditory processing and verbal memory loading while the animation design also minimises dependence on visual processing. Although the animation contains implied rather than clear clues, the participants assign social identity to the objects and form a social pattern from the movements of the objects, and thus evaluate the video by attributing social meaning within the social context of the movements of the geometric shapes.
The Social Attribution Task-Multiple Choice (SAT-MC) developed by Klin [11] as a multiple choice version of the Heider and Simmel [8] animation is formed of 19 questions [13]. The “correct” responses in the SAT coding scheme for the multiple choice questions, were formed with reference to the Salience Index. By attributing social meaning to ambiguous visual stimuli (images), the participant watching the video identifies social relationships and patterns between mobile geometric shapes (large triangle, small triangle, square, and rectangle) without any verbal contextual clues [12].
It has been shown in literature that the SAT-MC battery is used in the measurement of social cognition and more than one of the subdimensions of social cognition [14,15]. In the context of SAT-MC, there are studies in literature that have examined social attribution skills [3,16-18], social perception [5,19-22], and theory of mind [23]. In a study by Brown et al. [24], SAT-MC was used in the examination of social intelligence structure. There are also panel opinions given by specialists with the aim of determining social cognition measurements and leading future quantitative research [21,22]. SAT-MC is among the potential measures. Although not recommended in the social cognition psychometric evaluation study [19], SAT-MC has been shown to be a strong candidate by panel specialists because of factors such as the form being affected very little by language and cultural differences, and that it is non-verbal [22,23,25-27].
From an examination of studies in the Turkish literature, although some scales and batteries evaluate social cognition and different dimensions of social cognition [28,29], there is no scale or test battery for the evaluation of social attribution skills that has proven psychometric properties. Therefore, the first aim of this study was to contribute to the Turkish literature about the SAT-MC battery.
Different versions of the SAT-MC battery and scoring charts have been used in international literature. The first was the SAT developed by Klin [11], in which there is prominent difficulty in scoring and coding as the points are obtained from 7 different indices. To overcome these limitations, the SAT-MC was developed as a multiple-choice version of the SAT. A new version of the SAT-MC formed by Klin and Jones [13] was then introduced into the literature as the SAT-MC II by Johannesen et al. [27] This was in the same format as the original SAT-MC and used similar mobile geometric shapes. The object movements in the SAT-MC II were designed to enact a new social drama, different from that of the original.
In recent research, an experimental study was conducted on the new meanings attributed to the moving shapes [30]. The aim of the study was to determine whether or not similar observations were obtained by forming a different version from the several forms made by Klin [11] and Klin and Jones [13]. The new set developed was formed of 32 coloured animated shape videos and each video represented a different social situation [30]. Finally, it was observed that although the SAT versions differed in respect of the content and movement patterns of the geometric shapes, the participants spontaneously attributed social meaning to the images they watched.
Social cognition in schizophrenia
In recent years, social cognition has become an increasingly prioritised area in psychiatric research and evaluation, examined under the headings of neurodevelopmental, neurodegenerative, and neuropsychiatric disorders [12,31-38].
Schizophrenia, which is a neuropsychiatric disorder, is referred to as a network disease in which neural and social networks are impaired [39]. There are studies in the literature showing that social cognition is impaired in schizophrenia, which is characterized by social functionality disorders [40-43]. It has been reported that, especially when functional results are examined, the effect of social cognition has a relatively greater share than neurocognition [32,44]. Therefore, interest has increased in social cognition in schizophrenia studies [45].
In comparisons of SAT-MC performances of individuals diagnosed with schizophrenia and autism spectrum disorder, which are known to encompass social cognition disorders such as primarily mentalisation and understanding the thoughts and emotions of others, the points obtained by these two clinical groups have been determined to be significantly lower than those of a healthy control group, and these low points have been reported to be associated with a deficiency in social attribution skills in these clinical groups [3,12,14,18,20,25]. In a study by Bell et al. [25], the SAT-MC points of a schizophrenia sample showed statistically significant relationships with functions such as theory of mind, social problem-solving, and emotion recognition, but no links were found with processes such as verbal and visual learning, attention and alertness. In this context, Klin 11 determined that individuals who had difficulty understanding social intentions, meanings, and consequences focused on the physical properties rather than the social dimensions of the elements in the SAT material.
When psychometric measurement findings were examined in studies conducted with individuals diagnosed with schizophrenia, it was reported that SAT-MC was open to the learning/practice effect [46] and that test-retest reliability and internal consistency levels [19] were weak.
The primary aim of this study was to examine the reliability and validity of the Turkish version of the SAT-MC battery in individuals diagnosed with schizophrenia and a healthy control group. Then the link was examined between social attribution skills and the relationship between socio-demographic characteristics and cognitive functions. Finally, the social attribution skills of individuals diagnosed with schizophrenia were compared with those of the healthy control group and the effect of schizophrenia symptoms on SAT-MC points was evaluated.
METHODS
Research design
The research was conducted as two independent studies with a healthy control group sample and a schizophrenia patient group sample. First, the reliability and validity of the scale were proven with all sample groups (n=207). After seeing that the reliability and validity criteria of the SAT-MC were met, the group comparisons (n=50 for both groups) and predictive variables were examined.
Ethical approval statement
Approval for the study was granted by the Social Sciences and Humanities Ethics Committee of Bartin University (decision no: 2022-SBB-0378, Approval date:19.09.2022; decision no for study 2: 2024-SBB-0674).
Participants
The sample size required for the study was calculated using G*Power software (version 3.1; Heinrich Heine University), with reference to the study by Johannesen et al. [46] In the study by Johannesen et al. [46], the authors tested the psychometric properties of SAT-MC among schizophrenia and healthy adult groups and reported the mean scores pinpointed and effect sizes. In this study, the sample size was calculated based on the sample size of the cited study and the scores obtained. Accordingly, according to the independent samples t-test statistical method, considering the power as 0.95 and the Type I error rate as 0.05, the total sample size was found to be 100.
To compensate for losses and withdrawals, as seen in every study, the sample size was increased by 10%. The control group comprised 78.3% women and 21.7% men with a mean age of 26.32±8.98 years. For the clinical group, a sample size of 50 was determined and these subjects comprised 30% women and 70% men with a mean age of 37.18±10.85 years. The analyses were completed with 157 healthy adults and 50 patients diagnosed with schizophrenia. Table 1 presents the socio-demographic characteristics of the sample groups.
The study inclusion criteria for the control group were determined as, 1) age 18-64 years, 2) no active neurological or psychiatric complaints, 3) no use of any stimulant substance within the last 6 months, 4) no use of any anti-depressants, anxiolytic or psychotropic drugs within the last 6 months, and 5) no history of head trauma. Subjects were excluded from the study if they had a first-degree relative with a neurological or psychiatric disease diagnosis, obtained a score of >9 points on the Beck Depression Inventory (BDI), or a score of >7 points on the Beck Anxiety Inventory. The healthy control group was administered clinical scales to assess only the severity of depression and anxiety symptoms. Participants were considered healthy based on their self-reports of other mental and neurological disorders.
The study included patients with a known diagnosis of schizophrenia who were admitted to the psychiatry outpatient clinic of Mersin City Training and Research Hospital between August 2023 and August 2024. The Structured Clinical Interview for DSM-5-Clinician Version was applied to these patients, and individuals with a confirmed diagnosis of schizophrenia were invited to participate in the study by a psychiatrist. According to the results of the evaluations done, the data collected from the first 50 volunteers who met the inclusion and exclusion criteria and agreed to participate in the study were included in the analyses. Accordingly, 38 of the individuals collected from the clinical population were diagnosed with paranoid schizophrenia, six with residual schizophrenia, fore with atypical schizophrenia, and two with schizoaffective disorder.
The inclusion criteria for the clinical sample group were defined as, 1) age 18-64 years, 2) meeting the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) schizophrenia diagnostic criteria, and 3) to be at least a primary school graduate. Patients were excluded from the study if they 1) had any accompanying neurological or psychiatric disease, 2) were in a manic, psychotic, or depressive attack period which would prevent test compliance, 3) had a history of alcohol or substance abuse, 4) had any change in medical treatment within the last month, 5) had undergone electroconvulsive therapy within the last 6 months, or 6) had any additional physical disease disrupting their general condition.
The study was conducted in the provinces of Bartin and Mersin, and data were collected between April 2023 and May 2024. All the study participants provided written informed consent stating voluntary participation in the study. The study was conducted with separate interviews of each participant. In the study, initial data were collected from healthy individuals. Subsequently, with the inclusion of the schizophrenia group as part of a secondary investigation (Study 2), approval was obtained from two separate ethics committees.
Data collection tools
Sociodemographic data form
All the study participants were administered this form that included questions about age, gender, education level, psychiatric history, and substance use. For the schizophrenia patient group there were also questions about the diagnosis and treatment process of the current pathology.
Scale for the Assessment of Positive Symptoms
This scale was developed to evaluate the level, distribution, and change in severity of positive symptoms in schizophrenia, and is evaluated by a psychiatrist conducting an interview with the schizophrenia patient [46]. In this study it was used to examine the relationship between positive symptoms in schizophrenia and social attribution skills.
Scale for the Assessment of Negative Symptoms
This scale was developed to evaluate the level, distribution, and change in severity of negative symptoms in schizophrenia, and is evaluated by a psychiatrist conducting an interview with the schizophrenia patient [47]. In this study it was used to examine the relationship between the negative symptoms in the schizophrenia group and social attribution skills.
SAT-MC
The SAT-MC Turkish version used in this study was the version revised by Burger-Caplan et al. [12] of the SAT-MC developed by Klin and Jones [13]. The test material used was formed of an animation varying in length from 60-100 seconds and 19 multiple choice questions. The animation was a short, silent video portraying a social drama in the form of four moving geometric shapes [8]. The 19 questions in the test are used to evaluate social attribution skill [12]. Each question has 4 options of response. In the original battery, a total of 19 questions were asked and for each question a correct response was scored as 1 point, to give total points of 0-19.
Raven Standard Progressive Matrices Test
This test was developed by Raven [48] to measure visual spatial perception, perceptual reasoning and abstract thinking, processing memory and deliberation. As there is a relationship with the intelligence component known as fluid intelligence, it then started to be used as a non-verbal intelligence test [49,50]. In this study, the test was used to examine the social meaning attributed to the geometric shapes, and the relationship with perceptual reasoning and abstract thinking of the concept.
Reading the Mind in the Eyes Test
This test was developed by Baron-Cohen et al. [51] to evaluate the skill of making an inference about the mental state of others from a certain part of the face, primarily the eyes, in children with autism or Asperger’s syndrome [52]. The Turkish version consists of 32 pictures and this test was used in this study to evaluate the affective dimension of social cognition.
BDI
This 21-item test, published by Beck et al. [53] in 1961, aims to rapidly acquire information about the level of depression. In this study it was used to exclude individuals with depressive symptoms, using the cutoff point recommended by Hisli.
Beck Anxiety Inventory
This inventory, developed by Beck et al. [54] in 1988, is used to grade the current anxiety symptoms of an individual. As for the BDI, it was used as an exclusion criterion in this study. Subjects with a score <8 points were included in the study.
Procedure
After obtaining permission from the test developers, first the SAT form was translated. Then, the research team restructured the SAT version used in this study by using Adobe Illustrator and Adobe After Effects software. The reason for this was that in the previously used versions the image quality was insufficient and for translation to Turkish of the SAT questions in the sections related to the animation. In this context, the shapes in the animation were re-structured with vectoral-based drawings without disrupting the original structure in the Adobe Illustrator software. The vector-based shapes formed were then made into an animation video in 1920-1080 Full HD format in Adobe After Effects software, taking the original content into consideration. The SAT questions were added to the sections related to the animation video with the Adobe After Effects software. Finally, the video was updated in accordance with the aim of the research with both Adobe Illustrator and Adobe After Effects software. Thus, the SAT test material was converted to test material formed of 19 multiple choice questions integrated with a 71-second silent animation.
The SAT, which was planned in a fully appropriate to the original, consisted of a first video with no questions, then by integrating the questions, a second video to which responses were requested. Thus, the total running time of the two videos was 8 minutes and 55 seconds. Unlike the reference study of Burger-Caplan et al. [12], the multiple choice questions in the video were not pre-recorded and read aloud by a tester. Taking similar application principles into consideration in this adaptation, each participant completed the test on a computer following the test instructions themselves, and therefore this scale cannot be applied to illiterate individuals.
Statistical methods
In the descriptive statistical analyses, results were stated as arithmetic mean, median, and standard deviation values for continuous variables and as numbers and percentages for categorical variables. The conformity of the data to normal distribution was examined using the Shapiro-Wilk test. The known-groups method was used within the scope of construct validity of the study. For this purpose, whether the difference between the schizophrenia and control groups in terms of SAT-MC scores was tested with the independent t-test. Comparisons of categorical variables with normal distribution were examined with the chi-square test. Although the content validity of the social attribution tasks has been performed, the content validity of the test was re-evaluated because there may have been minor differences due to the reconstruction of the stimulus set and because it has exhibited low psychometric properties in validity studies in the literature. The Content Validity Index (CVI) was calculated for content validity, and Pearson correlation analysis was performed for criteria validity. While Polit and Beck [55] and Shi et al. [56] were taken into consideration in the interpretation of the obtained CVI values, Schober et al. [57] were taken into consideration in the interpretation of the correlation values. McDonald’s omega & Cronbach’s alpha coefficient were calculated for internal consistency of the scale, and the Pearson correlation coefficient was examined for test-retest reliability. The Osburn [58] article was considered for internal consistency coefficients. Accordingly, if the internal reliability coefficient is determined between 0.70-0.80, it is acceptable; if it is between 0.80-0.90, it is reported as “good.” Receiver operating characteristic (ROC) curve analysis was performed to determine the cutoff point and power of the scale in differentiating the healthy control group and the schizophrenia patient group. The Youden’s Index was used to determine the cutoff point, and the sensitivity and specificity values of the optimal cutoff point were determined. Hierarchical multivariate regression analysis was used in the examination of the role of age, education level, Raven Standard Progressive Matrices Test (RSPM), and Reading the Mind in the Eyes Test (RMET) in the evaluation of the social attribution skill, and this analysis was based only on the schizophrenia patient group.
RESULTS
Validity of the measurement tool
Language validity
The SAT was first translated from English to Turkish by the research team. This was then checked in terms of language appropriacy by an English teacher and a psychologist with an advanced level of English. The necessary corrections were made in accordance with the recommendations, and the final version of the scale was produced.
Content validity
An expert panel was consulted for content validity. The scale was evaluated by the expert panel in respect of the content of the test and the aim of use. As it was also thought that the test adapted from the expert panel could be applied in the future in Türkiye in individuals with other neuropsychiatric diagnoses (schizophrenia, autism, dementia), evaluations that took the understanding and conceptual capacity of these individuals into account were requested. The expert panel comprised 6 members who were specialists in the fields of pediatric and adolescent mental health, developmental psychology, experimental psychology, and neuroscience. A CVI based on expert opinions was used to calculate content validity [55]. The expert opinions in this study were evaluated with the CVI, and the S-CVI/Ave value was found to be 1.16. Although the acceptable values of the CVI/Ave values calculated for content validity differ from each other [55,59,60], the S-CVI/Ave value calculated in this study is considered above acceptable values.
Criterion validity
In the determination of the criteria validity of the scale, the concurrent validity method was used. The relationship between social attribution skills and perceptual reasoning was assessed with the RSPM, and the relationship with mentalizing skills was assessed with the RMET test. A positive, moderate-level, significant correlation was determined between the RSPM and SAT scores (r=0.583, p<0.001) in all participants. In the calculation of the relationship between the SAT score and RMET, the Pearson correlation coefficient was used, and there was determined to be a positive, moderate level, significant correlation (r=0.564, p<0.001).
Reliability of the measurement tool
Reliability
First, the McDonald’s omega and Cronbach’s alpha internal consistency coefficient of the adapted test was calculated and was determined to be 0.81 and 0.80 (95% confidence interval [CI], 0.77-0.84), respectively. When the correlations of the items with each other and the relationships with the total scale points were examined, it was seen that the internal consistency coefficient of each item made a similar contribution.
Test-retest reliability
For the evaluation of test-retest reliability, the test was repeated after two weeks on 45 control subjects and 16 schizophrenia patients. Using the Pearson correlation coefficient, a high correlation was determined between the two measurements (r=0.755, p<0.001). The test-retest reliability was determined to be at an acceptable level.
Variables predicting social attribution skill in both groups
Hierarchical multiple regression was used to evaluate the ability of the RSPM and the RMET to predict social attribution skills after controlling the effects of age and education level. To evaluate the suitability of the data for regression analysis, linearity, multicollinearity, normality, and whether or not the assumptions of equal variance were met, were checked first. Multicollinearity was checked by first checking correlations between variables. The highest correlation coefficients were 0.67, which were much lower values than the value of 0.90, which is a sign of multicollinearity [61]. To decide on the presence of multicollinearity, the cutoff values for the tolerance and variance inflation factor (VIF) were taken as a reference. A tolerance value of <0.10 and, VIF value >10 indicate a high correlation between the variables in question [62]. The tolerance values were calculated as 0.319-0.760, and the VIF values as 1.266-3.181. Then the autocorrelation value was examined with the Durbin-Watson test and was calculated as 1.926. This value, which is recommended to be 1.5-2.5 [63], was recorded as close to 2, and thus there was understood to be no autocorrelation between the variables. Therefore, it was seen that the multiple collinearity proposition was not violated. The normal P-P graph showed no significant deviations from normality. A scatter graph for linearity was evaluated, and it was seen that the linearity assumption was not violated. Thus, the preliminary regression analyses assumptions were not violated.
Hierarchical multivariate regression analysis was then performed. Age and education level were included in the first block, and these were seen to explain 15.4% of the variance in SAT. RSPM and RMET scores were included in the second block, and variance was found to be 36.5% (F(4,197)=28.345, p<0.001). After controlling age and education level, the RSPM and RMET explained 21.1% of the variance in SAT (R2 change=0.211, F change (2.197)=32.767, p<0.001). In the final model, RSPM and RMET were statistically significant. The beta value of RSPM (beta=0.38, p<0.001) was higher than the beta values of RMET (beta=0.28, p<0.001). The results of the hierarchical multivariate regression analysis are shown in Table 2.
Group comparisons in social attribution skills
The final analysis conducted within the scope of the construct validity of SAT-MC was the comparison of the difference in test scores between schizophrenia and control groups using the known-groups method [64]. As it is known, individuals diagnosed with schizophrenia perform worse by comparison to healthy controls in many social cognition tasks, especially theory of mind skills [3.20,25,40,65]. Based on this, it was also hypothesized that individuals diagnosed with schizophrenia would obtain lower SAT-MC scores than the control group in this study. Therefore, firstly, a matched control group was selected for the schizophrenia group with respect to sociodemographic characteristics such as age, gender, and education level (p>0.05). The Independent samples t-test was used to compare the mean SAT points between the schizophrenia patients and healthy control groups. The mean SAT points of the healthy control group (12.64±3.40) were determined to be statistically significantly higher than those of the schizophrenia group (7.28±3.63) (t(98)=7.62, p<0.001). In determining the size of the difference between the groups, Cohen’s d value was calculated, and there was seen to be a large effect size (Cohen’s d=1.52).
Sensitivity, specificity, and cutoff point of SAT
A ROC curve was drawn to evaluate the ability of SAT to differentiate individuals diagnosed with schizophrenia. The results showed that SAT had the power of differentiation at an acceptable level (area under the curve=0.853; 95% CI=0.780-0.926, p<0.005). The optimal cutoff point for SAT was determined using the Youden’s Index [66], and was calculated as 12 points with sensitivity of 0.920 and specificity of 0.680 [67].
DISCUSSION
The aim of this study was to adapt the SAT-MC to Turkish for a healthy control group and a schizophrenia group. After examining the reliability and validity of the SAT-MC in both sample groups, differences in SAT performance were examined between the groups. Finally, the effects of sociodemographic and clinical variables predicting SAT performance in schizophrenia patients were examined.
Validity analyses were performed first. Validity analyses began with content validity, as the visual stimulus set in the SAT was faithfully re-created, and the multiple-choice questions were translated into Turkish. Content validity encompasses the evaluation of the extent to which the measurement tool and each of its components serve the purpose [68]. However, as evaluations based on expert opinion are qualitative studies, conversion of the evaluation to statistical findings should provide the opportunity for a quantitative study [69]. The CVI and the content validity rates are evaluated in the calculation of the quantitative data obtained [70,71]. There are different approaches that can be used in the examination of content validity [59,72,73]. In the current study, a CVI based on expert opinion was calculated. The S-CVI/ Ave value was 1.16, and as this was over the recommended value of 0.90 [55], it was concluded that the content validity was sufficient.
Taking into consideration the need for more than one approach to validity analyses, criteria validity was also examined [74]. The RSPM and RMET tests were used to calculate the validity of the criteria. The findings obtained from the RSPM test showed that there was an association between perceptual reasoning skills and social attribution skills. In previous studies that have examined the relationship between verbal IQ and social attribution skill, it has been reported that verbal IQ does not predict social attribution skill [11-13,25]. However, in the current study, there was seen to be a relationship of social attribution skill with the aspect of intelligence defined as fluid intelligence. According to the psychometric-based intelligence theory of Cattell [49], fluid intelligence encompasses skills such as abstract thinking in the face of new problems, logic, and reasoning [75]. Within the neuropsychological tests, RSPM is the tool most frequently used to evaluate fluid intelligence [76]. As previously stated, it is necessary to perceive the social information related to the event/situation to successfully reveal social attribution skills [1,17]. This information should be interpreted as a whole, and finally, the decision should be made related to the most recent perception of the situation. This is similar to completing what is missing from the visuals in RSPM, and thus, a relationship is seen between SAT and intelligence in aspects such as perceptual reasoning and abstract thinking.
Secondly, the relationship was examined between social attribution skill and the RMET, which is a test often used in the evaluation of theory of mind skills with the aim of examining the relationship with social cognition in addition to the neurocognitive dimension. In this context, a significant moderate correlation was determined between the RMET and the SAT scores. Statistically significant moderate correlations were found between cognitive measurement tools and SAT points, especially in schizophrenia, in which social cognitive disorders are known to be seen [77]. Consequently, in the context of criteria validity, a relationship was determined between two different measurement tools with Turkish validity and reliability, which evaluated both neurocognition and social cognition with the aim of evaluating the cognitive functions of the SAT adaptation; it was concluded that criteria validity was provided.
The internal consistency of the SAT-MC test has been examined through McDonald’s omega and Cronbach’s alpha reliability coefficients, and this study found that the SAT-MC test showed good reliability. Previous studies have reported that test-retest reliability is not at an acceptable level [25], while studies report the opposite findings [7,46]. In the current study, a high correlation was found in test-retest reliability. In addition to this finding and the correlation obtained from the entire sample covering both groups, when the groups were examined separately, no evidence was found that the effect of practice was detected at the end of the pre-test-post-test process (p>0.05).
The SAT points of the schizophrenia patient group were seen to be much lower than those of the healthy control group. Although these low points were not correlated with the positive symptom measurements of schizophrenia, there was a moderate level correlation with the negative symptom measurements. Negative symptoms have been associated with social function disorders, such as social isolation and lack of empathy [78]. Publications in this field have also reported that SAT scores are not associated with schizophrenia symptoms [25]. However, many researchers have found the lack of association between schizophrenia symptoms and social cognitive disorders to be “structural,” and have reported that it can be found even in the relatives of schizophrenia patients [79].
Patients with schizophrenia often have difficulty interpreting the emotions and intentions of others, and this is known to cause problems in social interactions [80,81]. This leads to various difficulties, such as patients’ inability to interpret social cues correctly, difficulty understanding the perspectives of others, and inappropriate behavior in social situations. Here, social attribution, which tests the ability to infer the intentions of others, is also an indicator of social functioning. Since the SAT does not contain verbal cues, it may better indicate social cognitive functioning. Moreover, since it is well known that verbal skills are impaired in patients with schizophrenia, the use of tests free of verbal skills allows for a more reliable assessment of social cognition [78].
The effects of negative symptoms and cognition on social functioning in schizophrenia patients are clear [32,78]. In addition, it has been found that there are also structural and functional abnormalities in brain regions related to some social cognitive areas in schizophrenia patients [82]. Therefore, treatments and assessments targeting deficiencies in the impairment of social cognition in schizophrenia come to the forefront [83]. In this context, it has been reported that SAT may be promising in clinical studies in the evaluation of social attribution skills in schizophrenia patients [23,25,46]. In the study by Jung et al. [3], empathy and social attribution skills were evaluated in schizophrenia patients with reduced temporal lobe volume and the ability to recognize facial emotion. Here, SAT-MC was used to examine social attribution. It has been suggested that empathy and social attribution training may be effective. As a result, interventions such as social skills training and cognitive enhancement may help schizophrenia patients improve their social interaction skills and live more independent and satisfying lives [84].
The relationship of the SAT scores with demographic variables was examined. No statistically significant findings in respect of age, gender, and education level were found in both the schizophrenia patient group and the healthy control group in this study. In the literature that has examined the relationship between age and the SAT battery, there can be seen to have been different results according to the developmental period. Research conducted on school-age children in particular has shown that in parallel with cognitive development, the scores obtained from SAT increased together with age towards the end of adolescence [7,12]. However, in the limited number of studies conducted on both healthy adults and schizophrenia patients, no significant relationship has been determined between SAT scores and age [7,25,27,46]. In this respect, the current study is consistent with research conducted on adult samples.
When differentiation of social attribution skill between the genders is examined, it can be seen that as for age, there is a statistically significant difference between the genders up to the end of adolescence as the developmental period extends [85,86] and the gender difference is eliminated in adulthood [25,46]. Consistent with previous studies, no significant difference was found between the variable of education level and SAT points for both of the current study sample groups [25,46]. However, education level included as an independent variable in the hierarchical multivariate regression model applied to the schizophrenia patient group was seen to be a predictor of SAT points.
The results of the hierarchical multivariate regression analysis showed that RSPM and RMET were predictive of social attribution skill. Education level and age included in the model in the first block provided a significant contribution to the model. There are studies in the literature supporting that there is no statistical difference in SAT points according to education level [25,46]. The variable of age has also been reported to have no effect on the SAT score [7,25,27,46]. In the second block, RSPM was added to the model based on the literature reporting that social cognitive deficits in schizophrenia are associated with fluid intelligence, and RMET was added as an assessor of the affective dimension of social cognition [87]. RSPM and RMET made a significant contribution to the model and the age and education level variables were not significant.
It has been previously stated that RSPM, which has been found to be a strong predictor of the SAT, is a frequently used measurement tool for fluid intelligence. Deficiencies in fluid intelligence, characterized by complex social behaviours and executive functions [88,89], are associated with cognitive impairments in patients with schizophrenia [90]. These cognitive impairments can disrupt the cognitive processes necessary for social interactions, such as processing social cues, adapting to social contexts, and understanding other people’s perspectives [88]. This relationship established with fluid intelligence supports the understanding of neurocognitive processes affecting the social cognition skills of patients with schizophrenia and makes its use in treatment valuable [5]. The relationship between fluid intelligence and social attribution measured by RPSM may be due to structural or functional abnormalities in the brain prefrontal cortex, which processes social cognition and neurocognition. Furthermore, fluid intelligence may be an essential ability in social functioning. Thus, it can be thought that individuals with more vital abstraction may be better able to interpret non-verbal behaviours and evaluate social cues better. There may be common processing characteristics and pathways between these two tests since both tests require the inference of meaningful relationships between geometric shapes and patterns. However, the SAT is dynamic, and the RSPM is static.
ROC curve analysis is a method used in the determination of the best cutoff point of a scale and in the calculation of the benefits of correct and incorrect responses [91]. In the current study, the capacity of the SAT to differentiate healthy individuals from schizophrenia patients was determined to be at an acceptable level. Using the Youden’s Index, the cutoff point has been calculated to be 12 [92]. In the study by Bell et al. [25], the optimal cutoff point for SAT was found to be 12 points. Although the sensitivity and specificity values obtained were close, the sensitivity value was higher. Bell et al. [25] stated that specificity should be preferred to sensitivity.
This study had some important limitations, primarily that the gender distributions were not equal in both groups, and the control group included more young adults. In addition, it was not possible to represent all education levels equally. The participants completed the scales on a computer by themselves, which required at least a basic level of literacy. Therefore, the results obtained cannot be generalised to illiterate individuals. It has also been reported that in computer-based neuropsychological tests, the ability of the individual in computer use could affect test performance [93]. However, during the data collection process, no participant reported any difficulty in completing the SAT to the researchers. This could be attributed to the participants being relatively young. Therefore, in future studies, computer use skills should be taken into consideration when evaluating the performance of older individuals with neurodegenerative disorders.
In conclusion, the findings obtained in this study demonstrated that the Turkish version of the SAT is a valid and reliable tool for measuring social attribution skills and social cognition skills in healthy individuals and patients diagnosed with schizophrenia. The SAT points were correlated with cognitive skills, and the education level, RSPM, and RMET scores were determined to be predictors of social attribution skills in the schizophrenia patient group. Therefore, the difficulty experienced by schizophrenia patients in understanding clues and the intentions of others can adversely affect negative symptoms such as lack of interest and social withdrawal. It can be said that in addition to pharmacological treatments, treatments that target negative symptoms and social-cognitive functions could be of benefit to patients with schizophrenia. It can be recommended that the test is performed using the original scoring of the cutoff point of 12.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
