یک روش آماری برای مقایسه چندین ابزار هنر درمانی در برآورد سطح حالتی روانشناختی

An approach using a statistical method has been proposed to compare the effectiveness of several art therapy tools in estimating the level of a psychological state, which includes various symptoms and disorders. For each of the art therapy tools, a regression model is developed, whose dependent variable (to be explained) is the level of a psychological state. Various elements in drawings are considered as independent (explaining) variables, some of which are evaluated by the computer systems for art therapy assessments. A set of independent variables “well” explaining the dependent variable is determined by the stepwise regression. The effectiveness of art therapy tools is analyzed and measured in a comparable quantitative term, and thus can be compared with other tools. In this paper, the approach has been applied to three widely used art therapy tools in estimating the level of dementia. The same method can be adopted as a predictor of a psychological state of an individual.

There are many studies estimating psychological states based on the elements in drawings. For example, Reddy, Bhadramani, and Samiullah (2002) distinguished the differences in the Kinetic Family Drawings (KFD) (Burns & Kaufman, 1972) done by the children who were neglected and the children who were considered to be raised in a normal family setting. The results showed that the drawings differed in the contents of the family members. Children from neglected families were more likely to put the family members ‘irregularly’ and be ‘reluctant’ to draw. Louw and Ramkisson (2002) identified sexually abused girls from the House-Tree-Person (HTP) Test (Buck, 1948) and the Draw-a-Person (DAP) Test (Naglieri, 1988), where the sexually abused girls differed significantly on the four scales of the HTP and the DAP, compared to the same age group who did not have any history of sexual abuse. Veltman and Browne (2001) found that the KFD was three times better than the Favorite Kind of Day (FKD) (Manning, 1987) for identifying maltreated children. Kim, Betts, Kim, and Kang (2009) estimated the level and predicted the existence of dementia from the Structured Mandala Coloring (SMC). They found that the level of dementia was related to the accuracy in coloring area given, the number of color clusters generated, and the use of light-green, brown, and green, and that severe dementia was related to the degree of concentration in coloring a structured mandala and the use of brown. In this study, an approach using a statistical method has been developed to compare the effectiveness of several art therapy tools in estimating the level of a psychological state, including the symptoms and disorders of the child abuse or neglect, sexual abuse, trauma, schizophrenia, dementia, depression, attention deficit, etc. As a statistical method, a multiple linear regression analysis has been adopted, which is a widely used method explaining one variable from the others (Kutner, Nachtsheim, Neter, & Li, 2005). The dependent (to be explained) variable of the model is the level of a psychological state. The level may be scored by the formal tests such as the Mini-Mental State Evaluation (MMSE) (Folstein, Folstein, & McHugh, 1975), the Beck's Depression Inventory (BDI) (Beck, Ward, Mendelson, Mock, & Erbaugh, 1961), the Symptom Checklist-90-Revisional (SCL-90-R) (Derogatis, 1977), the Myers–Briggs Type Indicator (MBTI) (Myers & McCaulley, 1985), and the Murphy–Meisgeier Type Indicator for Children (MMTIC) (Murphy & Meisgeier, 1987), or the evaluation from experts such as art therapists and psychologists. The independent (explaining) variables can be the elements in drawings including ones in the Descriptive Assessment for Psychiatric Art (DAPA) (Hacking, Foreman, & Belcher, 1996), the Diagnostic Drawing Series (DDS) (Cohen, Hammer, & Singer, 1988), the Formal Elements Art Therapy Scale (FEATS) (Gantt & Tabone, 1998), the Face Stimulus Assessment (FSA) (Hamilton, 2008), and the Computer_Color Related Elements Art Therapy Evaluation System (C_CREATES) (Kim, 2010 and Kim et al., 2007), such as the number of colors used, space usage, details, accuracy in coloring, etc. Some elements are automatically evaluated by the computer systems for art assessments, which include the C_CREATES, the determination of unusual placement (Kim, Kang, & Kim, 2008), the judgment of main color (Kim, 2008), the assessment of completeness and accuracy in coloring a pattern (Kim, Kang, & Kim, 2009), and the rating of the color variety (Kim & Hameed, 2009); others are manually evaluated by art therapists. Among various independent variables, by stepwise regression, some variables “well” explaining the dependent variable are selected. The effectiveness of tools is analyzed via scatter plots and is measured by the coefficient of determination (R2). Also, the relative importance of selected independent variables is compared by the standardized regression. As an application and case study of this approach, the effectiveness of three art therapy tools in estimating the level of dementia, the Person Picking an Apple from Tree (PPAT) (Gantt, 1990), the FSA (Betts, 2003), and the SMC, were compared. The samples were collected from 58 persons with suspected dementia in a psychiatric unit, ages ranging from 60 to 90 years old. The level of dementia, a dependent variable of the model, was measured by the Korean version of the MMSE (MMSE-K). A total of 64 elements were considered as the independent variables of the model. These elements came from the DAPA, the DDS, the FEATS, the FSA, and the C_CREATES. Some independent variables, such as the number of colors, number of clusters, length of edge, and number of pixel painted, were automatically evaluated by several computer systems, while others, such as the color fit, implied energy, integration, and logic, were manually evaluated by human art therapists. Many art therapists may struggle to convince other mental health professionals on the effectiveness of art therapy tools. The approach to choose an effective tool and analyze the relationship between the test scores and its estimated values may aid many art therapists who hope to find evidence to support their observations and impressions to official documents. Use of this approach is not to suggest how to diagnose the clients; rather, it is to aid the judgment process for art therapists. The approach developed in this study is a generalized one, and thus can compare the effectiveness of any special art therapy tools in estimating the levels of various psychological states. Also, the approach can be applied to statistically identify the psychological disorder, which could aid the art therapists’ process of evaluating their clients. Additionally, the approach can support the psychological tests for clients with limited ability in speaking, reading, and comprehension.

We developed a generalized approach using the regression model to compare any kind of art therapy tools such as the DDS, the KFD, the DAP, and the HTP to estimate the levels of any kind of psychological disorders such as ADHD, sexual abuse, and depression. This study applied the approach to compare the PPAT, the FSA, and the SMC in estimating the level of dementia measured by the MMSE-K. The scatter plots of correlations between the MMSE-K scores and the estimated MMSE-K scores were shown to compare the effectiveness of the regression model for the three art therapy tools. The coefficients of determination via the regression model for the PPAT, the FSA, and the SMC were 0.61, 0.45, and 0.59, respectively, which indicated that all three art therapy tools were marginally satisfactory or applicable and that the PPAT was the best. The SMC was the second best. The ranking of the estimated MMSE-K and the MMSE-K were partly correlated. The Spearman's rank correlation coefficient was 0.786 for the PPAT, 0.576 for the FSA, and 0.639 for the SMC. The level of dementia was measured by the MMSE-K for 58 samples from patients with suspected dementia at a psychiatric unit. There are other tests such as the Geriatric Deterioration Scale (GDS) (Yesavage et al., 1983) and the Clinical Dementia Rating (CDR) (Morris, 1993), which measure the level of dementia. The size of our samples, 58, may not be large enough, and samples from a broad area are desirable. More tests and a larger sample size from broader areas might result in a different conclusion. Probably, we can obtain improved results targeted at people with slight dementia (Wang et al., 1998). Also, we can obtain improved R2 through the nonlinear regression models. We note that our method can be applied to predict the probability of existence of a psychological disorder. The approach is expected to promote the use of art therapy and pave the way for the development of new technology for the analysis and application of art therapy, as well as its theoretical understanding.