INTRODUCTION

Studies which have investigated the phenomenology of obsessive compulsive disorder (OCD) have noted that symptoms are very heterogeneous and individualized having patterns of obsessions with contamination, hoarding, religiousness, symmetry, and somatic content; and main compulsions of washing, checking, counting, ordering, and hoarding (1). According to the prevalence studies, the second most common type of obsessions and compulsions comprise those of contamination related themes in 38-45% of the cases with OCD (2,3). While the content of fear of contamination might change, it might appear after a physical contact with a dirty/dangerous object or person. The fear of contamination is described as quite complex, intense, easily triggred, rapidly spreading, not easily subsided, and difficult to control (4). However, Rachman (5) indicated that feeling of contamination might arise without any physical contact with an object or subject, and he proposed the concept of mental contamination. According to Rachman (5), mental contamination refers to feelings of internal dirtiness even though the person has no physical contact with a contaminant.

Although contact contamination and mental contamination are similar concepts, it is emphasized that there are basic differences between the two. For example, while contact contamination requires tangible objects or people that are assumed to be dirty, in mental contamination direct contact is not a necessarry condition. Contact contamination, fear of contamination generally originates from dirty or dangerous objects or places; but in case of the mental contamination, the source of contamination is generally the people or the thoughts that are considered as dirty, dangerous or harmful. In both forms of contamination, individuals feel very strong urge to wash/clean themselves. In case of contact contamination, since the source of the contamination is usually the hands, the individual focuses on this particular part of the body for the removal of the contaminant although the washing behavior is only effective for a short period of time (6). Due to the nature of mental contamination, it does not require any contact, feelings of dirtiness are described as an internal dirtiness without any focus, and is generally spreading all over the body. Therefore, washing behavior is generally inadequate in overcoming these feelings, and individuals with these feelings develop frequently other behavioral patterns besides hand-washing behavior (6). Furthermore, it has been predicted that mental contamination may be associated with some negative feelings. In a qualitative study, it was found that all of the individuals with mental contamination reported negative feelings such as uncomfortableness, anxiety, anger, shame, disgust, distress, and irritability (6). Different from the ones having feelings of contact contamination, individuals who experience mental contamination have tendencies to attribute this feeling of dirtiness to one’s moral character such as being a bad, worthless, and immoral person. These kinds of moral evaluations are not generally observed in contact contamination cases.

In light of these findings, the number of studies attempting to better understand the nature of mental contamination has risen in recent years. When we examined the first studies of mental contamination, Fairbrother and Rachman’s study was remarkable. They examined mental contamination in female subjects who had previous sexual abuse history. In their study, it was found that even remembering and talking about the sexual abuse created the feelings of dirtiness, urge to wash, and even caused washing behavior. Following this study, the phenomenon of mental contamination has been examined through experiments performed on scenarios in which people are asked to imagine to be kissed without their consents. Fairbrother et al. (8) triggered mental contamination in female university students by making them listen to a voice record about a non-consensual kissing case. Consequently, different variables such as attractiveness (9), immoral behavior (10), physical dirtiness (11), and betrayal were added on these scenarios, and the factors affecting mental contamination were investigated. Radomsky and Elliott (13) performed a study in which they let participants listen conditions by manipulating responsibility (the condition that participant did not give consent to be kissed), personal violation (the condition that the participants were kissed by force) and morality (the condition that the kisser was an immoral individual) components into different scenarios, so as to determine their effects on mental contamination. Consistent with expectations, kissing scenarios triggered mental contamination in participants and caused negative feelings. Elliott and Radomsky (11) investigated associations of mental contamination feelings when non-consensual kissing condition was performed by a physically dirty person. According to obtained results, it was shown that non-consensual kissing by a physically dirty person caused higher mental contamination than the condition that the kissing person was clean, and also the condition of consensual kissing by a dirty/clean person. In light of these findings, the need for self-report instruments to measure mental contamination has increased over time.

Vancouver Obsessive Compulsive Inventory-Mental Contamination Scale (VOCI-MC) is one of the first scales developed to evaluate mental contamination based on Rachman’s (4) mental contamination theory. Since mental contamination is associated with moral judgements and the thoughts about internal dirtiness, then thought action fusion, the condition in which the person equates his/her thoughts to actually performing them, might play an important role in mental contamination. Therefore, thought-action fusion in association to mental contamination requires further investigation (14). Thought-Action Fusion-Contamination Scale (TAF-CS) is a measure for fusion of contamination related thoughts, behaviors and feelings, and it is based on thought-action fusion concept which has been proposed by Shafran et al. (14). In a study evaluating psychometric characteristics of VOCI-MC and TAF-CS (14), internal consistency coefficients of both measures were evaluated in different samples including an OCD group with contamination fears, an OCD group without contamination fears, an anxiety group, and a student group. The psychometric properties were at the satisfactory levels.

In summary, mental contamination seems to be a closely related concept to urges to wash and washing behavior in OCD. As washing behavior has a maintaining role in OCD symptoms, it is important that mental contamination should be understood better so as to increase our theoretical and practical understanding of OCD symptoms. There is no study performed about mental contamination and its role in OCD in our country yet. Therefore, it is believed that the Turkish versions of standardized measures which evaluate mental contamination objectively and effectively is very important to increase the number of of such studies. In line with this requirement, the current study presents the Turkish adaptations of VOCI-MC and TAF-CS with their psychometric evaluations.

METHOD

The sample of the present study was composed of 225 university students who were not diagnosed with any psychiatric diagnosis aging between 18-28 years old (females=183, males=70). Of participants, 97.6% (n=249) told that they were single, 0.8% (n=2) were married, and 0.4% (n=1) was divorced.

Two of the participants did not give information about their genders, and 3 did not give information about their marital status. All of the participants reported that they spent all of their life in Turkey.

Measures

Vancouver Obsessive-Compulsive Inventory- Mental Contamination Scale (VOCI-MC): VOCI-MC is a 5-point Likert scale type measurement tool which evaluates symptoms of mental contamination between 0-4 scores and, contains 20 items (16). It was reported that Cronbach’s alpha internal consistency coefficients were between 0.97 and 0.93 in different sample groups (15). Correlations between two measures (Contamination Sensitivity Scale and Thought-Action Fusion Contamination Scale) which were related to mental contamination, and Vancouver Obsessive Compulsive Inventory (VOCI) and contamination subscale of this inventory were examined to evaluate convergent validity of the scale. The results indicated that the scale was significantly and positively related to these scales. In order to evaluate divergent validity, correlation of the scale with Beck Depression Inventory (BDI) (17) and VOCI-contamination subscale was examined. As expected, the correlation of the scale with VOCI-contamination subscale was significantly higher than its correlation with depression. In order to evaluate concurrent validity, VOCI-MC scores of OCD patients with contamination symptoms were compared with the patients without contamination symptoms by ANOVA analysis, and it was found that two groups were significantly different from each other. For predictive validity of the scale, hierarchical regression analysis was performed by defining OCD symptom level as a dependent variable, and it was observed that the scale could continue to explain variation of OCD symptoms. Depending on these findings, it was reported that convergent, concurrent, divergent, and predictive validity of the scale were at the desired level (15). Psychometric properties of the Turkish version of the scale will be evaluated in the present study.

Thought-Action Fusion- Contamination Scale (TAF-CS): TAF-CS is a scale developed to measure feelings and behaviors related to beliefs for thought-action about contamination/getting dirty by Rachman (18). TAF-CS is a measuring tool of 5-point Likert type (0=I disagree, 4=I totally agree) with 9 items. Internal consistency coefficients of the scale are 0.96 for OCD groups with and without contamination symptoms, 0.95 for anxiety group, and 0.93 for the student group. To evaluate convergent validity of the scale, its correlation with VOCI-Contamination subscale was investigated, and it was found that there was a significant positive correlation with university student sample. However, no significant correlation was determined with groups diagnosed with anxiety and OCD. To evaluate divergent validity of the scale, correlation of the scale with BDI, and VOCI -contamination subscale was examined. As expected, the correlation between the scale and VOCI-contamination subscale was significantly higher than the correlation with BDI. To determine concurrent validity, TAF-CS scores of OCD with and without contamination symptoms were compared, and it was observed that the scale could only distinguish between clinical and non-clinical samples. Depending on these findings, it was reported that convergent and divergent validity of the scale was at the satisfactory level (15). Psychometric properties of the Turkish version of the scale will be evaluated in the present study.

Thought-Action Fusion Scale (TAFS): TAFS is a self-reporting, 5-point Likert type scale with 19 items, and was developed by Shafran et al. (14). Previously TAFS was used as three dimensions, then it was reported that two dimensions, TAFS-Likelihood, and TAFS-Moral, were suitable, and TAFS was used in two dimensions in later studies. Psychometric characteristics of the scale were re-evaluated in the study conducted by Rassin et al. (19), and it was indicated that it had the satisfactory level of internal consistency (for total and subscales of TAF α>0.75), but correlations of test-retest were relatively low (total=0.52, likelihood=0.51, self=0.53, others=0.47, moral=0.54). In our country Yorulmaz et al. (20) adapted the scale into Turkish, and psychometric properties of TAFS were evaluated. It was reported that Cronbach’s alpha reliability coefficient of TAF-Likelihood subscale, which was composed of 7 items was 0.92, and TAF-Moral subscale, which was composed of 12 items was 0.88. Performed validity analyses indicated that two factor structure of the scale had also acceptable validity values for the Turkish sample.

Obsessive-Compulsive Inventory Revised (OCI-R): Obsessive-Compulsive Inventory-Revised (OCI-R) is a self-reporting, 5-point Likert type scale, which was developed because of disadvantages of 84-item Obsessive Compulsive Inventory (21) in practice, and improved to 18 items by Foa et al. (22). It measures basic symptoms and mental neutralization of OCD. In the study performed by Abramowitz and Deacon (23), it was reported that internal consistency coefficients of OCI-R (Cronbach’s alpha=0.86), and its subscales (0.83-0.92) had the required consistency. Yorulmaz et al. (24) adapted the scale into Turkish and evaluated its psychometric properties. It was determined that factorial structure of the scale was consistent with the original study. According to reliability analysis, total scale and other subscales had satisfactory levels of internal consistency coefficients, except checking and neutralization (α=0.64 and 0.64) (for total scale α=0.90, for subscales α=0.73-0.84).Besides the scale was a reliable measure, it was found that the scale had satisfactory criterion-related, convergent, and divergent validity values.

The Disgust Scale-Revised (DS-R) Form: Disgust Scale was developed by Haidt et al. (25), and it measures propensity of disgust in various stimuli. The scale has been used previously in different forms, but the latest version which was 5-point Likert type for all 27 items was revised by Olatunji et al. (26). The scale has three subscales; core disgust, animal reminder, and contamination. Cronbach’s alpha reliability coefficient of the scale ranges between 0.34 and 0.64 for all subscales (25). The scale was adapted into Turkish by Inozu and Eremsoy (27). According to reliability analysis, it was reported that internal consistency coefficient of total scale was 0.87, and test-retest reliability was 0.88. Additionally, it was determined that its three factor structure was consistent with its original form. In validity analyses, there were signs indicating that the scale had convergent, divergent, and criteria validities. According to obtained results, it was reported that the scale was a reliable and valid measure of disgust propensity for our country.

Trait Anger-Anger Expression Style Scale (TA-AESS): TA-AESS is a measuring tool with 44 items which is divided into two such as Trait Anger and Anger Expression Style, and developed by Spielberger (28). In the present study, Trait Anger Scale (TAS) which had 10 items, and measured only continuous anger level, was used. The scale was adapted into Turkish by Ozer (29). In reliability-validity analyses, it was found that internal consistency coefficient of the scale was 0.79, and it was reported that the scale was reliable. It was reported that factorial structure of the scale was consistent with original factor structure, and the scale was valid.

Procedure

Translation of Scales

Both scales were translated into Turkish independently by three clinical psychologists who had advanced knowledge in both languages (Turkish-English), and were specialized in OCD. Alternative translations of the scales made by independent translators were compared with the original items and evaluated by the research team, and the best translations reflecting the original item content were selected by the referee agreement. After this procedure, the first Turkish version of scales were obtained. The original English version and the first Turkish version of scales were given to 5 participants who had advanced knowledge in both languages, and were out of this field, and they were requested to score them. Additionally participants were also report whether they understood the same content while they were filling up the English and Turkish versions of the forms, and if it was required, they should correct the statements. After evaluation of the scales, the final versions were prepared, and they were sent to back-translation from Turkish to English to a specialist who had advanced knowledge in both languages, and was not included in the research team. The back-translation version was sent to abroad to researchers who developed these scales, and their approvals were obtained. Finally, scales were prepared to be used.

Data Collection

The study has been approved by a research ethics committee of Dogus University. Only voluntary participants were included into the present study. The study was announced to students before the courses that permission was already obtained, and volunteering students received the questionnaire set, and the students were requested to fill them up in the classroom. The questionnaire set was counterbalanced and took approximately 20-25 minutes to fill in. After filling out the questionnaire, participants received a course credit as an incentive for their participation.

RESULTS

Internal consistency (Cronbach’s alpha) coefficients of VOCI-MC and TAF-CS were found as 0.93 and 0.92, respectively. For TAF-CS, intercorrelation of scale items was ranged between 0.61-0.85; correlation coefficients for VOCI-MC ranged between 0.50 and 0.76. Guttman split-half reliability coefficients were determined as 0.88 for TAF-CS, and 0.90 for VOCI-MC. When split-half reliability was examined, the reliability coefficients for the first part of TAF-CS which contained 5 items was 0.86, and the second part which contained 4 items was 0.89. In addition, the reliability coefficients of the first part of VOCI-MC, containing 10 items, was 0.86, whereas the second part was 0.89. Test-retest reliability which was performed four weeks apart, indicated 0.62 for TAF-CS, and 0.79 for VOCI-MC. These values indicated that both scales had acceptable reliability coefficients.

Validity

Construct Validity

As VOCI-MC and TAF-CS scales were accepted as having a single factorial structure in the original study, an exploratory factor analysis was performed primarily to examine the factor structure in our country, then a confirmatory factor analysis was used to confirm the obtained factor structure (LISREL 8.71).

Exploratory Factor Analysis

To evaluate the structural validity of VOCI-MC, Principal Component Analysis with ‘Varimax rotation was performed. KMO index, which is an index for sampling adequacy for exploratory factor analysis, was 0.93; whereas another marker, Barlett’s Chi-square score was 4591.39 (p<0.001). These results showed that the data properties were ideal for exploratory factor analysis. There were three values greater than eigenvalue of the scale (>1). Three factors reflected by these three values explained 58.77% of total variation. However, scree-plot curve indicated a two-factorial structure. When the item distribution according to Varimax rotation has done, it was observed that 16 items loaded both in the first and second factors with more than 0.30 factor load, and only four items (items: 5,15,11,and 7) loaded on the second factor.

A content examination of the four items was made and it was observed that the items in the two factors are theoretically not distinct from each other. Therefore, it was decided that the scale should be used as single factorial, which was also consistent with the original form. A single factorial structure explained 46.63% of the total variance with item loadings ranging from 0.44 to 0.80.

When the same analyses were used to evaluate TAF-CS factorial structure, it was determined that KMO was 0.087, and Barlett’s Chi-square score was 2163.87 (p<0.001). These values indicated that data were sufficient for factor analysis. Two values were defined with eigenvalues of the scale as >1. Two factors which were reflected by these values explained 67.18% of total variance. However, when item distribution of two factorial structure of the scale was examined by using Varimax rotation, only two items (items 7 and 9) were loaded on the second factor with more than 0.30 factor load. When content of these two items examined, it was observed that its content could not be differentiated from the first factor. Consistent with these findings, scree-plot curve indicated a single factor structure. Therefore, it was decided to use the form as a single factor which was also consistent with the original form. A single factorial structure explained 65.47% of the total variance with item loadings ranging from 0.67 to 0.86.

Confirmation of Unidimentional Structure of

VOCI-MC and TAF-CS

Confirmatory Factorial Analysis (CFA) was performed to test unidimentional factor structure of the scales by using LISREL 8.71 package program (30). In CFA or in model tests, Chi-square should be insignificant to show that the model is acceptable. However, Chi-square value is very sensitive to the sample size. Therefore, if the sample size is large, df ratios of Chi-square (2, 3, and even 5) indicate that the model is acceptable. Besides, it is expected that CFI, GFI, AGF and NNFI values are 0.90 and above, whereas RMSEA value is below 0.08 (31-34).

First, the single factor structure of VOCI-MC scale was examined. Model fit indices obtained by covariation matrices (χ2/df=6.20, AGFI=0.71, GFI=0.83, CFI=0.81, NFI=0.79, RMSEA=0.14, p<0.001) indicated that some modifications should be performed. According to suggested modification indices, the error terms belong to item pairs of 6-16, 5-18, 15-18, 7-11, 1-3 of the scale were covaried. After errors were covaried, the model was tested again (χ2/df=3.51, AGFI=0.84, GFI=0.87, CFI=0.90, NFI=0.87, RMSEA=0.079, p<0.001), and it was determined that added relations improved the model (χ2difference (6)=478.04, p<0.001). Although fit indices were slightly below the expected value, df ratio of Chi-square was determined as 3.51, and this value indicated that the model was at the acceptable level.

The obtained fit indices confirmed a single factorial structure of TAF-CS (χ2/df=14.02, AGFI=0.74, GFI=0.79, CFI=0.81, NFI=0.79, RMSEA=0.18, p<0.001).

According to proposed modification indices (7–9, 4–1, 2–5), errors between these very similar items were covaried, and then the model was tested again (χ2/df=3.38, AGFI=0.91, GFI=0.95, CFI=0.95, NFI=0.95, RMSEA=0.085, p<0.001), and it was observed that added relations improved the model (χ2difference (2)=281.23, p<0.001). Obtained fit indices showed that the model was acceptable.

Convergent Validity Analyses

To evaluate the convergent validity of the scales, correlations between both the new measuring tools and the total and subscale scores of TAFS, DS-R, TAS, OCI-R were examined. As it is shown in Table 1, there were significant positive correlations between both VOCI-MC, TAF-CS scales, and OCI-R total and its subscales, which were showing OCD symptoms, and also total and subscales of TAFS. Taken together, these results supported the convergent validity of the scales.

Divergent Validity

To test the divergent validity of VOCI-MC and TAF-CS, the correlations of these two measures with OCI-R Total (except Washing/Cleaning subscale) and Washing/Cleaning subscale scores were investigated. The results showed that the correlations significantly different from the correlations of trait anger and disgust sensitivity. When the obtained data were reviewed, positive correlations were determined between VOCI-MC and OCI-R Total (except Washing/Cleaning) (r(254)=0.57, p<0.001), and Washing/Cleaning subscale (r(254)=0.61, p<0.001). Similarly, it was determined that VOCI-MC had a significant positive relationship with TAS (r(254)=0.21, p<0.001). However, the correlations between VOCI-MC and OCI-R Total (Z=6.15, p<0.01) and Washing/Cleaning subscale scores (Z=6.52, p<0.01) were significantly higher than the correlation with trait anger subscale. Similarly, the correlations between VOCI-MC and OCI-R Total (except Washing/Cleaning) and Washing/Cleaning subscale were significantly higher than the correlation with disgust scores (Z=5.67, p<0.001; Z=6.66, p<0.001, respectively).

Similar results were obtained when correlations were examined between TAF-CS and OCI-R Total (except Washing/Cleaning) and Washing/Cleaning subscale, anger and disgust scores. It was found that there was a significant positive correlation between TAF-CS and OCI-R Total (except Washing/Cleaning) (r(252)=0.50, p<0.001) and Washing/Cleaning scores (r(252)=0.45, p<0.001). It was also found that TAF-CS had a significant positive correlation with anger (r(252)=0.25, p<0.001). However, correlations between TAF-CS and OCI-R Total (Z=4.21, p<0.01) and Washing/Cleaning symptom subscale scores (Z=3.10, p<0.01) were significantly higher than the correlation with trait anger scale. Similarly, there was a significant difference between TAF-CS and OCI-R Total (except Washing/Cleaning) and Washing/Cleaning subscale scores (Z=4.66, p<0.01; Z=4.07, p<0.01, respectively), than the correlation with TAF-CS and disgust scores (r(251)=0.17, p<0.001). These values indicated that both scales had divergent validity.

To further evaluate the divergent validity of scales, participants who scored below 25th percentile and above 75th percentile on the OCI-R Total (except Washing/Cleaning subscale) score, were classified into two groups as OCD high (n=69; >32 points) and OCD low (n=69; <16 points). One-way ANOVA was performed for VOCI-MC total scale score. The results indicated that VOCI-MC significantly differentiated the group with high OCD symptoms scores from the one with low OCD symptoms scores (F1,136=89.14, p<0.001, η2partial=0.40). The VOCI-MC scores in the high OCD group (M=24.34, SD=15.81) were significantly higher than the low OCD group (M=5.51, SD=4.98). The same analyses were performed to evaluate for divergent validity of TAF-CS. One-way ANOVA was performed for TAF-CS total scale score, and it was found that TAF-CS scores in high OCD group (M=13.63, SD=9.33) were significantly higher than the scores in low OCD group (M=3.45, SD=4.14) (F1,135=68.44, p<0.001, η2partial=0.34). These results showed that both VOCI-MC and TAF-CS scales differentiated groups with high OCD scores and low OCD scores significantly; indicating that the scales had divergent validity.

Predictive Validity

Hierarchical regression analysis was performed to evaluate the degree to which the new measuring tools predicted OCD symptoms. As a dependent variable, new OCI total score, which was formed by subtracting Washing/Cleaning subscale from OCI-R total score, was taken. In the first stage, control variables of age, gender and disgust sensitivity were included in the analysis. In the first stage of the analysis, the control variables explained 9% of the variance, and caused a significant increase in the explained variance (ΔF[3,243]8.75, p<0.001). When standardized regression coefficient was investigated to determine which of variables included in the analysis caused a significant change, age (β=-0.17, t=-2.70, pr=-0.17, p<0.05) and disgust sensitivity (β=0.28, t=4.36, pr=0.25, p<0.001) significantly predicted OCD symptoms. In the second stage, TAF-Likelihood and TAF-Moral subscales which were among OCD beliefs were included in the model. These variables caused 21% and 30% of increases in the explained variance, respectively (ΔF[2,241]=37.01, p<0.001). When it was investigated which variable caused significant change, it was found that only TAF-Moral dimension significantly predicted OCD symptoms (β=0.46, t=4.12, pr=0.26, p<0.001). In the third stage, OCI-R Washing/Cleaning subscale was added into the analysis (β=0.60, t=12.11, pr=0.61, p<0.001) and this variable caused 26% increase in the explained variance (ΔF[1,240]=146.54, p<0.001). In the fourth and last stage, VOCI-MC and TAF-CS were added in the model. Addition of these two measures in the model caused a 2% of increase in the variance, and explained variance was increased to 58% (ΔF[2,238]=4.95, p<0.05). In the last stage, only VOCI-MC predicted significantly OCD symptoms (β=0.14, t=2.44, pr=0.16, p<0.05).

DISCUSSION

In the present study, the aim was to analyze the psychometric properties of two new instruments that were developed to measure a new concept called mental contamination in a Turkish sample.

Reliability analysis showed that internal consistency coefficients were parallel with obtained data in the original study (14). In addition to internal consistency coefficients, test-retest reliabilities of the scales examined. Although four weeks’ interval between the two testing was relatively lower, the scores obtained from the two scales showed consistency over time and similar scores were obtained at different time intervals. Hence, the results supported reliability of the Turkish adaptation of the scales.

Consistent with the original study, validity values of the single structured VOCI-MC and TAF-CS scales were examined through construct, convergent, and divergent validity analyses. The convergent validity is examined through other OCD related variables. The results showed that both the mental contamination scale and the thought-action fusion mental contamination scale together with its subscales and total scale scores were significantly positively related to other OCD related variables. Another important finding supporting the convergent validity of the scales was that both scales showed the most powerful relationship with Washing/Cleaning subscale among OCD symptoms. Another finding which supported the association of mental contamination with OCD symptomatology was obtained in hierarchical regression analysis which was performed to evaluate predictive validity. Regression analysis results indicated that the control variables were still significant on OCD symptom levels despite inclusion of VOCI-MC at the last step in the analysis after effects of TAF-Moral, TAF-Likelihood, and OCI-R Washing/Cleaning subscale scores were controlled. All of these findings were consistent with the literature findings indicating that mental contamination was related to OCD (35,36). Coughtrey et al. (35) performed a study on 177 patients diagnosed with OCD, and they reported that 10% of the patients complained about mental contamination without any physical contact; 15% complained about contamination with only physical contact; and 36% complained about both mental and contact related contamination. Depending on their results Coughtrey et al. (10) underlined that there was an overlap between mental contamination and contact related contamination, but despite this overlap mental contamination had a different structure. Consistent with these findings in the present study, mental contamination still continued to significantly predict the OCD symptoms when contamination symptoms of OCD were controlled. Moreover, VOCI-MC scores predicted significantly OCI-R total scores after controlling Washing/Cleaning subscale scores, therefore mental contamination was not only related to washing/cleaning symptoms of OCD, but they were also relate to other OCD symptoms. As mental contamination feeling had no specific point in the body, and it was defined as an internal feeling of dirtiness, these findings were supportive for additional neutralization behaviors in addition to hand washing in individuals having these feelings (6). Consistent with the original study findings (14), when psychometric characteristics of both scales were examined, there was no significant correlation between TAF-CS and OCD symptoms. It was shown that thought-action fusion was not specific for OCD, it played an important role in other anxiety disorders (37). Findings from this study were consistent with the results of the original study, and the same condition was valid for mental contamination and though-action fusion. Similarly, it seemed to support the idea that the thought-action fusion was not specific for OCD.

The divergent validity of the scales was examined by comparing OCI-Total (exception of Washing/Cleaning) and Washing/Cleaning score correlations to trait anger and disgust sensitivity. The findings were consistent with the expectations, indicating that both two scales had divergent validity. VOCI-MC and TAF-CS scores differentiate significantly between low and high OCD symptom groups, and it was a supportive sign for divergent validity of the scales. Moreover, it seemed to be also consistent with theoretical perspective proposing that mental contamination and related thought-action fusion beliefs should be related to other symptoms of OCD (38,4).

Although the basic underlined feeling in OCD is anxiety according to classical cognitive behavioral theories of OCD (39-42), the recent studies suggested that other feelings such as guilt (43,44), anger (45), and disgust (46-48) might also play important roles in exacerbation of symptoms. Therefore updated models of OCD should be restructured in a way that they should also cover these other feelings. Results of the present study were consistent with this suggestion, and it was shown that both mental contamination and thought-action fusion beliefs related to mental contamination were significantly and positively related to feelings of anger and disgust. Therefore, it is believed that the clinicians who deal with the disorder treatment in daily practice should also consider additional feelings such as anger, guilt, and disgust.

It is believed that the adaptations of these two new scales into Turkish, which objectively measure mental contamination by providing comparable results to the original forms, will contribute to the field by giving rise to investigations about this new concept. However, there are important limitations in interpretation of the study results. Firstly, mental contamination case was investigated among university students who had no diagnoses. Therefore, presence of a non-clinical group with a limited age range limits the generalizability of the results. Although it is a common method to use university students and adult samples in OCD research (49), using a clinical sample diagnosed with OCD and mental contamination, would increase the external validity of the findings. Another important limitation of the study is that the results are based on self-report ratings. Therefore, the results do not indicate a causal relationship, but only indicate a reciprocal relationship between the variables. To increase our understanding about the nature and cognitive components of mental contamination, the use of experimental methods in future studies is important.

In conclusion, the present study indicates that two measures, VOCI-MC and TAF-C scales, which have been developed for better understanding the nature and cognitive components of mental contamination have sufficient reliability and validity to be used in Turkish samples. It is also believed that cultural adaptation of these two new measures, will increase our understanding about OCD, by stimulating interest for future studies in this field. Mental contamination is thought to be a missed symptom group (4) among clinicians. Therefore, an interest in the research realm would also help to increase clinicians’ familiarity with the concept and result in better conceptualization of OCD and treatment interventions. Future studies are needed to explore whether contact related contamination and mental contamination load up to the same factors; on which aspects they differ from each other; and the specific role of mental contamination in OCD symptomatology.

Conflict of Interest: Authors declared no conflict of interest.

Financial Disclosure: Authors declared no financial support.

REFERENCES

1.Akhtar S, Wig NN, Varma VK, Pershad D, Verma SK. A phenomenological analysis of symptoms in obsessive-compulsive neurosis. Br J Psychiatry 1975; 127:342-348. [CrossRef]

2.Calamari JE, Wiegartz PS, Riemann BC, Cohen RJ, Greer A, Jacobi DM, Jahn SC, Carmin C. Obsessive-compulsive disorder subtypes: an attempted replication and extension of a symptom-based taxonomy. Behav Res Ther 2004; 42:647-670. [CrossRef]

3.Rasmussen SA, Eisen JL. Phenomenology of OCD: clinical subtypes, Heterogeneity and Coexistence. In The Psychobiology of Obsessive- Compulsive Disorder, (Eds.) Zohar J, Insel T, Rasmussen S.pp13-43, New York: Springer Publishing Company 1991,13-43.

4.Rachman S. Fear of contamination. Behav Res Ther 2004; 42:1227-1255. [CrossRef]

5.Rachman S. Pollution of the mind. Behav Res Ther 1994; 32:311-314. [CrossRef]

6.Coughtrey AE, Shafran R, Lee M, Rachman SJ. It’s the feeling inside my head: a qualitative analysis of mental contamination in obsessive-compulsive disorder. Behav Cogn Psychother 2012; 40:163-173. [CrossRef]

7.Fairbrother N, Rachman S. Feelings of mental pollution subsequent to sexual assault. Behav Res Ther 2004; 42:173-189. [CrossRef]

8.Fairbrother N, Newth SJ, Rachman S. Mental pollution: feelings of dirtiness without physical contact. Behav Res Ther 2005; 43:121-130. [CrossRef]

9.Herba JK, Rachman S. Vulnerability to mental contamination. Behav Res Ther 2007; 45:2804-2812. [CrossRef]

10.Elliott CM, Radomsky AS. Analyses of mental contamination: Part I, experimental manipulations of morality. Behav Res Ther 2009; 47:995-1003. [CrossRef]

11.Elliott CM, Radomsky AS. Mental contamination: the effects of imagined physical dirt and immoral behaviour. Behav Res Ther 2012; 50:422-427. [CrossRef]

12.Rachman S, Radomsky AS, Elliott CM, Zysk E. Mental contamination: the perpetrator effect. J Behav Ther Exp Psychiatry 2012; 43:587-593. [CrossRef]

13.Radomsky AS, Elliott CM. Analyses of mental contamination: Part II, individual differences. Behav Res and Ther 2009; 47:1004-1011. [CrossRef]

14.Shafran R, Thordarson DS, Rachman S. Thought-action fusion in obsessive compulsive disorder. J Anxiety Disord 1996; 10:379-391. [CrossRef]

15.Radomsky AS, Rachman S, Shafran R, Coughtrey AE, Barber KC. The nature and assessment of mental contamination: A psychometric analysis. J Obsessive Compuls Relat Disord 2014; 3:181-187. [CrossRef]

16.Rachman S. Vancouver Obsessional Compulsive Inventory (VOCI-MC). Rachman Lab, University of British Columbia, 2005a.

17.Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J. An inventory for measuring depression. Arch Gen Psychiatry 1961; 4:561-571. [CrossRef]

18.Rachman S. Contamination Thought-Action Fusion Scale (CTAF). Rachman Lab, University of British Columbia, 2005b.

19.Rassin E, Merckelbach H, Muris P, Schmidt H. The thought-action fusion scale: further evidence for its reliability and validity. Behav Res Ther 2001; 39:537-544. [CrossRef]

20.Yorulmaz O, Yilmaz AE, Gencoz T. Psychometric properties of the Thought–Action Fusion Scale in a Turkish sample. Behav Res Ther 2004; 42:1203-1214. [CrossRef]

21.Foa EB, Kozak MJ, Salkovskis PM, Coles ME, Amir N. The validation of a new obsessive–compulsive disorder scale: The Obsessive–Compulsive Inventory. Psychol Assess 1998; 10:206-214. [CrossRef]

22.Foa EB, Huppert JD, Leiberg S, Langner R, Kichic R, Hajcak G, Salkovskis PM. The Obsessive-Compulsive Inventory: development and validation of a short version. Psychol Assess 2002; 14:485-496. [CrossRef]

23.Abramowitz JS, Deacon BJ. Psychometric properties and construct validity of the Obsessive–Compulsive Inventory—Revised: Replication and extension with a clinical sample. J Anxiety Disord 2006; 20:1016-1035. [CrossRef]

24.Yorulmaz O, Inozu M, Clark DA, Radomsky AS. Psychometric properties of the obsessive-compulsive inventory-revised in a Turkish analogue sample. Psychol Rep 2015; 117:781-793. [CrossRef]

25.Haidt J, McCauley C, Rozin P. Individual differences in sensitivity to disgust: A scale sampling seven domains of disgust elicitors. Pers Individ Dif 1994; 16:701-713. [CrossRef]

26.Olatunji BO, Haidt J, McKay D, David B. Core, animal reminder, and contamination disgust: Three kinds of disgust with distinct personality, behavioral, physiological, and clinical correlates. J Res Pers 2008; 42:1243-1259. [CrossRef]

27.Inozu M, Eremsoy CE. Psychometric Properties of the Turkish Versions of Disgust Scale and Contamination Cognition Scale. Turkish Psychology Articles 2013; 16:1-10. (Turkish)

28.Spielberger CD. State-Trait Anger Expression Inventory (STAXI). Research edition. Odessa, FL: Psychological Assessment Resources, 1988.

29.Ozer AK. Preliminary study for continuous anger (C-Anger) and anger expression style (Anger-Style) scales. Turkish Journal of Psychology 1994; 9:26-35. (Turkish)

30.Jöreskog KG, Sörbom D. LISREL 8. Structural equation modeling with the SIMPLIS command language. Lincolnwood, IL: Scientific Software International, 1993, 133-158.

31.Bollen KA. Confirmatory Factor Analysis. Structural equations with latent variables. New York, NY: Wiley, 1989, 226-317.

32.Hu LT, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal 1999; 6:1-55. [CrossRef]

33.Kline RB. Measurement Models and Confirmatory Factor Analysis. Principles and practices of structural equation modeling. New York: Guilford, 2011, 230-264.

34.Tabachnick BG, Fidell LS. Principal components and factor analysis. Using Multivariate Statistics. United States: Allyn and Bacon Press, 2001, 612-680.

35.Coughtrey AE, Shafran R, Knibbs D, Rachman SJ. Mental contamination in obsessive–compulsive disorder. J Obsessive Compuls Relat Disord 2012; 1:244-250. [CrossRef]

36.Cougle JR, Lee HJ, Horowitz JD, Wolitzky-Taylor KB, Telch MJ. An exploration of the relationship between mental pollution and OCD symptoms. J Behav Ther Exp Psychiatry 2008; 39:340-353. [CrossRef]

37.Abramowitz JS, Whiteside S, Lynam D, Kalsy S. Is thought-action fusion specific to obsessive-compulsive disorder?: a mediating role of negative affect. Behav Res Ther 2003; 41:1069-1079. [CrossRef]

38.Abramowitz JS, Fabricant LE, Taylor S, Deacon BJ, McKay D, Storch EA. The relevance of analogue studies for understanding obsessions and compulsions. Clin Psychol Rev 2014; 34:206-217. [CrossRef]

39.Salkovskis PM. Obsessional-compulsive problems: a cognitive-behavioural analysis. Behav Res Ther 1985; 23:571-583. [CrossRef]

40.Salkovskis PM. Cognitive-behavioural factors and the persistence of intrusive thoughts in obsessional problems. Behav Res Ther 1989; 27:677-682. [CrossRef]

41.Rachman S. A cognitive theory of obsessions. Behav Res Ther 1997; 35:793-802. [CrossRef]

42.Clark DA. Cognitive-behavioral therapy for OCD. New York: Guilford Press, 2004.

43.Nelson JD. Mental pollution and inflated responsibility in obsessive-compulsive disorder: The contribution of anxiety, disgust, and guilt. Unpublished dissertation thesis, Fordham University, New York, 2005.

44.Hale MA, Clark DA. When good people have bad thoughts: Religiosity and the emotional regulation of guilt-inducing intrusive thoughts. J Psychol Theol 2013; 41:24-35.

45.Haciomeroglu B, Inozu M, Gokdemir BP. The association of reassurance seeking with obsessive compulsive symptoms and related emotions. Oral presentation at the 14th European Congress of Psychology (ECP), Milan-Italy, 2015.

46.Olatunji BO, Sawchuk CN, Lohr JM, de Jong PJ. Disgust domains in the prediction of contamination fear. Behav Res Ther 2004; 42:93-104. [CrossRef]

47.Tolin DF, Worhunsky P, Maltby N. Sympathetic magic in contamination-related OCD. J Behav Ther Exp Psychiatry 2004; 35:193-205. [CrossRef]

48.Woody SR, Tolin DF. The relationship between disgust sensitivity and avoidant behavior: studies of clinical and nonclinical samples. J Anxiety Disord 2002; 16:543-559. [CrossRef]

49.Warren R, Gershuny BS, Sher KJ. Cognitions in subclinical obsessive compulsive disorder. In: Frost RO, Steketee G. Cognitive Approaches to Obsessions and Compulsions: Theory, Assessment and Treatment, Pergamon: Elsevier, 2002, 337-360. [CrossRef]