Current Psychology (2023) 42:11403–11414

https://doi.org/10.1007/s12144-021-02436-6

A preliminary psychometric study of the Turkish Schema mode

inventory-forensic (SMI-F)
Gonca Soygüt1 · İ. Volkan Gülüm2 · Ayşe E. Ersayan3,4 · Jill Lobbestael3 · David P. Bernstein3

Accepted: 23 October 2021 / Published online: 29 October 2021

© The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2021

Abstract
The Schema Mode Inventory was the first tool that was developed to assess schema modes (SMI; Young et al., 2007).
Recently, the SMI was expanded to also assess forensic modes (Bernstein et al., 2014). The main purpose of the current
study was to test The Schema Mode Inventory – Forensics’ (SMI-F) reliability and validity. The sample consists of (n = 1271)
volunteer undergraduate students across various universities from Turkey. The sample consisted mostly of females (77.5%).
The mean age of the whole sample was 20.43 (SD = 2.16, range = 18-57). In order to test the psychometric properties of the
SMI-F, we carried out confirmatory factor analysis, multiple regression analysis, correlational analyses for test-retest, and
calculated the correlations and internal reliability coefficients. Overall, the results revealed that the SMI-F has satisfactory
levels of reliability and validity and might be useful for research and clinical purposes. In conclusion, the inclusion of the
forensic modes in this version makes the inventory more comprehensive in reflecting recent developments in the Schema
Therapy Model. It could be possible to assess a wide variety of coping modes and formulate a treatment plan for the general
population and forensic patients with the SMI-F.

Keywords schema therapy · The Turkish Schema mode inventory-forensic (SMI-F) · schema modes · Psychometric study ·
Forensic

Schema Therapy (ST; Young, 1999; Young et al., 2003) is (BPD; Farrell et al., 2009; Giesen-Bloo et al., 2006) and
an integrative form of treatment developed for personality with Cluster C patients (Bamelis et al., 2014). This effective-
disorders (PDs) and other chronic psychiatric problems. ness was also documented when ST was applied in a group
According to ST, toxic early experiences and traumatic format (Farrell et al., 2009). Two randomized clinical trials
events contribute to developing emotional and cognitive of ST are also being conducted to document its effective-
self-defeating patterns (i.e., maladaptive schemas) during ness on forensic patients (Bernstein et al., 2007; Bernstein
childhood and adolescence and continue into adulthood et al., 2012).
(Young et al., 2003). Therefore, schemas activated by life A main concept in ST is schema modes that represent
events evoke powerful emotions and coping efforts that are the person’s active cognitive and emotional state and cop-
common in PDs. Research shows that ST is an effective ing responses at a certain moment in time (Young et al.,
treatment for patients with borderline personality disorder 2003). In patients with severe PDs, the shift between schema
modes is rapid (Lobbestael et al., 2007). ST aims to decrease
the intensity and frequency of these shifts and empower the
* Gonca Soygüt adaptive modes to be more active in a patient’s life (Reiss
[email protected]; [email protected] et al., 2016). Young et al. (2003) have described 10 modes
1
that were grouped into four categories. Child modes develop
Department of Psychology, İstanbul Şehir University,
İstanbul, Turkey when the child’s basic needs were not met during child-
2
hood. Dysfunctional parent modes are the critical and overly
Department of Counseling and Guidance, Dumlupınar
University, Kütahya, Turkey demanding inner voices internalized during childhood. Dys-
3
functional coping modes include overcompensation, avoid-
Department of Clinical and Psychological Science,
Maastricht University, Maastricht, The Netherlands ance, and surrender modes described as the maladaptive
4
actions taken to deal with activated schema modes. Healthy
Department of Psychology, Koç University, İstanbul, Turkey

13
Vol.:(0123456789)
11404 Current Psychology (2023) 42:11403–11414

modes include functional thoughts and behaviors. Healthy (2018) investigated male prisoners’ schema and modes in
modes include the Healthy adult which is capable of giv- Australia. They found that enraged child mode, impulsive
ing healthy decisions and acting maturely, and the Happy child mode, and bully and attack mode significantly pre-
child characterized by playfulness, spontaneity, curiosity, dicted aggression.
and feeling safe. The preliminary results of randomized controlled trials
The Schema Mode Inventory was the first tool that was with forensic patients with PDs showed that ST was better
developed to assess schema modes (SMI; Young et al., than treatment-as-usual to reduce recidivism and promote
2007). The scale included 270 items that intended to meas- resocialization (Bernstein et al., 2012). In another study,
ure the original 10 schema modes and the six modes that forensic inpatients with psychopathic features showed sig-
were later added in the model for a better description of nificant improvement with ST with combined movement
specific PDs (Lobbestael et al., 2007). Based on that ver- therapy and milieu therapy over four years of treatment and
sion, those items with the highest item-loading were retained three-year follow-up (Chakhssi et al., 2014). Overall, ST is
leading to a shorter version of 118 items (Lobbestael et al., a popular tool for treating antisocial personality and psycho-
2010). The 118-item SMI showed the best fit in a 14-factor pathic features (e.g., Bernstein et al., 2012; Chakhssi et al.,
model and acceptable reliability scores (Lobbestael et al., 2014) as well as other axis I (e.g., Morvaridi et al., 2019;
2010). The psychometric properties of the scale have been Simpson et al., 2010) and axis II (e.g., Farrell et al., 2009;
assessed in other languages such as Italian, German, Danish Renner et al., 2013) disorders. It seems important to note
and Urdu and the results have been promising (Panzeri et al., that the studies mentioned above have used a different ver-
2016; Reiss et al., 2012; Reiss et al., 2016; Riaz et al., 2013). sion of the SMI which not including forensic modes or used
In 2007, Bernstein and colleagues proposed four new observational methods to assess forensic mode concepts.
modes that they have found to be common in forensic In recent years, there has been extensive research investi-
patients in their clinical practice. Angry protector mode gating antisocial traits in the general population. For exam-
represents the use of a wall of anger that the person uses to ple, studies in non-forensic samples have investigated traits
protect her/himself from others who s/he sees as a threat. such as the Dark Triad (Psychopathy, Narcissism, and Mach-
In predator mode, the person behaves ruthlessly and in a iavellianism), and psychopathic traits using measures such
calculating manner to eliminate a threat or an enemy. When as Lilienfeld’s Psychopathic Personality Inventory and Pat-
the conning and manipulative mode is activated, the per- rick’s Tripartite Scales. These studies support the notion of a
son cons, lies and manipulates in order to reach a desired continuum of antisocial traits and behavior spanning normal
goal. The suspicious over-controller mode exercises hyper- populations and forensic ones. Research on antisocial traits
vigilance and extreme control to protect her/himself from in normal populations may provide insights into processes
threats. and mechanisms that may be relevant to forensic populations
Recently, the SMI has been expanded to also assess (so-called analogue studies). Whereas these studies have
forensic modes (Bernstein et al., 2014). The Schema Mode been based on dimensional models of antisocial personality
Inventory-Forensic Supplement (SMI-FS) consists of four traits, no measures have yet been developed of personality
supplemental scales which, when added to the SMI, pro- states that may form a continuum between normal and foren-
vide coverage of the four forensic modes, as well as a fifth sic populations. Bernstein and Navot (2020) have argued that
scale which assesses forensic variants of the self-aggran- personality states known as “schema modes” that appear
dizer mode (Bernstein et al., 2014). The self-aggrandizer in forensic populations, such as ones involving grandios-
mode in non-forensic populations is usually characterized ity, manipulation, and predatory aggression (Keulen-de Vos
by superiority and arrogance. However, in forensic popula- et al., 2016, 2017), may also be present in the general popu-
tions, a more extreme form is often observed, characterized lation, though in somewhat different and attenuated form.
by dominance (Bernstein & Navot, 2020). Developing supplementary scales for the SMI that measure
Although the current study presents the first psychometric these so-called forensic modes would enable research on
evaluation of the SMI-F, several previous studies using vari- these personality states in forensic and non-forensic samples.
ous methods or tools to show the role of the forensic mode In the present study, we investigated the dimensionality of
concepts on criminal behavior and the effectiveness of these the SMI-F in a non-forensic student sample as the first step
new modes in forensic treatment. For example, Keulen-de in the validation process.
Vos et al. (2016) used an observational method to evalu- Despite the evidence on their effect on criminal activ-
ate modes in forensic settings. They found that vulnerable ity and the effectiveness of ST in the treatment of forensic
and angry child modes were predictive for the physically modes, there is a lack of reliable and valid schema forensic
aggressive behavior level and overcompensatory modes, in mode assessment tools. The SMI-F has been developed but
general, were a significant predictor for physical violence not been psychometrically tested until the present study.
during criminal activity. In another study, Dunne et al. This study is the first to assess the psychometric properties

13
Current Psychology (2023) 42:11403–11414 11405

Table 1  Participants’ age and N Min Max M SD

gender
Total Sample Female 985 18 57 20.26 2.07
Male 286 18 44 21.01 2.34
Test-Retest Sample Female 34 20 57 22.21 6.28
Male 15 20 44 23.40 5.79

of the SMI-F. More specifically, we expected (i) sufficient the first three authors in 2015.1 Two of the translators are
reliability and validity results; (ii) consistent factor structure advanced; one is standard level schema therapists approved
with the original SMI; (iii) in terms of discriminant validity, by the International Society of Schema Therapy. The SMI-F
dysfunctional modes will predict the psychological symp- was translated to Turkish and back-translated to English by
tomatology. Accordingly, factor structure was tested with a bilingual expert to test translation accuracy. The items had
confirmatory factor analyses, test-retest reliability and inter- to be scored on frequency using a 6-point scale ranging from
nal reliability coefficients were calculated, and the multiple 1 (never or hardly ever) to 6 (always).
regression analysis was conducted.
Brief Symptom Inventory

Methodology To test criterion-related validity of the SMI-F we used the

Brief Symptom Inventory (BSI). The BSI was originally
Participants developed by Derogatis (1992) and adapted in Turkish by
Şahin and Durak (1994). The BSI Turkish form proved to
Psychometric properties of the SMI-F were analyzed from exist only of five subscales (anxiety, depression, negative
a convenience sample of 1271 undergraduate students from self-concept, somatization, and hostility) and 53 items each
various universities in Turkey. The researchers reached scored on a 5-point Likert scale from 0 (not at all) to 4
the students in their universities, asked for their volunteer (extremely). Higher scores in BSI represent higher level of
participation, and provided the study’s link. There was not psychological symptoms in general and/or in a specific area.
any compensation. The sample consisted mostly of females The Turkish BSI subscales` Cronbach alpha levels range
(77.5%). The mean age of the whole sample was 20.43 between α = .75 and .88. The global form explained 66.2%
(SD = 2.16, range = 18-57). All participants were native of the total variance (Şahin & Durak, 1994). In the cur-
speakers of Turkish. Age and gender characteristics of the rent study, subscales’ Cronbach alpha levels range between
participants can be seen in Table 1. α = .75-.89, and α = .96 for the overall inventory.

Procedure and Statistical Analysis

Materials
The current study was approved by the Koç University
Schema Mode Inventory – Forensic Ethics Committee for Social Sciences Research with pro-
tocol number 2016.017.IRB3.014. All participants signed
Bernstein et al. (2014) developed the Schema Mode Inven- informed consent. The data was collected via a web-based
tory Forensic (SMI-F) by building on the SMI short ver- tool (www.s​ urvee​ y.c​ om) from the volunteer college students
sion with 15 subscales and 118 items. Relative to the SMI from different universities across Turkey. To test hypoth-
short version (Lobbestael et al., 2010), the SMI-F has five esized construct of the SMI-F, the factorial structure was
more subscales that include 56 forensic items and a total assessed through confirmatory factor analysis (CFA). The
of 19 subscales. A total of 174 items were translated by goodness-of-fit was evaluated using the comparative fit
index (CFI), the Tucker-Lewis Index (TLI), the Standard-
ized Root Mean Square Residual (SRMR), the Root Mean
1
During the submission process of the current manuscript, a study Square Error of Approximation (RMSEA), the χ2-test and
on the Turkish SMI short version with 14 subscales and 118 items
has been published by a different research group (Aytaç, Köse the degrees of freedom. Following Hu and Bentler (1999),
Karaca, & Karasomanoğlu, 2020). The current study has been con- CFI, TLI, values close to or higher than .95, SRMR value
ducted on the SMI-F with 19 subcales including 5 forensic subscales close to or lower than .08, and RMSEA values close to or
and a total of 174 items. The scale has been translated into Turkish as lower than .06 are considered indicative of a relatively good
a whole to maintain translation integrity. The current study is com-
pletely independent from the aforementioned work. fit between the hypothesized model and the observed data.
Test-retest was assessed over a two weeks period with a

13
11406 Current Psychology (2023) 42:11403–11414

Table 2  The Fit Indices of χ2 df CFI TLI RMSEA SRMR

Tested Models
Measurement Model 67,090.567 14,706 .91 .91 .053 .065
Structural Model 51,202.891 12,784 .93 .93 .049 .060

sub-sample of 49 participants (69.39% were female) using conning and manipulative modes; happy child and angry
Intraclass Correlation Coefficient (ICC). We used ICC to protector and suspicious over-controller modes.
evaluate both correlation and agreement between two meas-
urements. Multiple regression analysis was run with a sub-
sample of 260 participants (53.8% were female) to evaluate Multiple Regression Analysis
the criterion-related validity of the SMI-F. The BSI was only
administered in one university. Thus, this sub-sample was A subsample of n = 260 undergraduate student (53.8%
selected based on the available data. female, ­MAge = 19.61, SD = 1.57) participants completed the
BSI (M = 1.04, SD = 0.65). A power analysis (Champely,
2020) with a significance level of .05 and a power of .80
Results with 19 predictor variables suggested a sample size of 153.
A multiple regression analysis was conducted to determine
Confirmatory Factor Analysis whether schema modes could predict psychological symp-
toms. Prior to the analysis all SMI-F subscales’ and BSI
The CFA analysis was run with using the R Studio soft- scores were transformed using base 10 logarithms. Nineteen
ware and Lavaan package (Yves, 2012). There were mild to modes were included the analysis as predictor variables and
moderate deviations from normality across the SMI-F items the BSI score was the dependent variable. The vulnerable
thus ratings were transformed using base 10 logarithms. The child, angry child, undisciplined child, compliant surrender,
CFA analysis were run with the diagonally weighted least suspicious over-controller, and bully and attack modes pre-
squares estimation several times and appropriate modifica- dicted the BSI score significantly, F(19,240) = 23.4, p < .001,
tions were made. The fit indices of the measurement and R2 = .65, R2Adjusted = .62. Model coefficients related to multi-
structural models of the 19-factor model are presented in ple regression analysis were presented in the Table 6.
Table 2. The structural model’s CFI and TLI levels were
close to .95; the RMSEA level was below .06, and SRMR
level was below .08, which could be considered acceptable. Discussion
We run a post-hoc power analysis using the R Studio
and semPower package (Moshagen & Erdfelder, 2016). In order to test the psychometric properties of the SMI-F, we
The results suggest that, sample size of N = 1271 with a carried out confirmatory factor analysis, multiple regression
power larger than >99.99% to reject a wrong model (with analysis, correlational analyses for test-retest, and calculated
df = 12,784) with an amount of misspecification correspond- the correlations and internal reliability coefficients. Overall,
ing to RMSEA = .049 on alpha = .05. the result revealed that the SMI-F has satisfactory levels of
The SMI-F subscales’ internal consistency levels range reliability and validity and might be useful for research and
between α = .62 and .91. The factor loadings, internal con- clinical purposes.
sistency levels and the factor structure were presented in To test the factor structure of the SMI-F we ran a CFA.
Table 3. In total, 11 items’ factor loadings were below the.30 There were eleven items with low factor loadings and those
level and were not included in further analyses. In the end, items were deleted. Although Hu and Bentler’s (1999)
there were 5 to 13 items per factor. recommendation regarding the cutoff levels of fit indices
The test-retest analyses were done with the 49 partici- (CFI and TLI levels close .95, RMSE level close to .06,
pants (69.39% female) in a two-week period. Table 4 shows and SRMR level close to .08) generally considered as rule
the baseline and retest mean and standard deviations as well of thumb, some researchers argued that this criterion are
as ICCs. ICC was conducted using absolute agreement and “too restrictive” especially for the complicated models (e.g.,
two-way mixed effects model. Marsh et al., 2004; Marsh et al., 2005). With 163 items and
The inter-correlations of the SMI-F subscales are pre- 19 factors, our model might be interpreted as a complicated
sented in Table 5. Most of the inter-correlations were sig- model and, the CFI and TLI levels were close to .95 and
nificant except the correlations between healthy adult mode RMSEA and SRMR levels were below .05. In sum, the final
and enraged child, impulsive child, detached protector, and CFA results showed that the SMI-F has adequate fit indices

13
Current Psychology (2023) 42:11403–11414 11407

Table 3  Factor loadings and structure of the SMI-F (N = 1271)

Factors and Items Factor
Load-
ings

Vulnerable Child (n = 11, α = .91)

5. I feel humiliated .64
18. I feel left out or excluded. .68
29. I often feel alone in the world. .72
52. I feel weak and helpless. .66
79. I feel desperate. .71
97. I feel lost. .73
137. Even if there are people around me, I feel lonely. .71
120. I feel that nobody loves me. .74
159. I feel fundamentally inadequate, flawed, or defective. .69
174. I feel lonely. .67
30. I’ve been cheated or treated unfairly. .54
Angry Child (n = 8, α = .75)
67. I feel enraged toward other people. .66
90. I feel like telling people off for the way they have treated me. .59
113. If someone is not with me, he or she is against me. .50
145. I have a lot of anger built up inside of me that I need to let out. .63
147. I’m angry with someone for leaving me alone or abandoning me. .46
162. I feel like lashing out or hurting someone for what he/she did to me. .61
130. I’m angry that people are trying to take away my freedom or independence. .32
144. It makes me angry when someone tells me how I should feel or behave. .36
Enraged Child (n = 9, α = .89)
39. Once I start to feel angry, I often don’t control it and lose my temper. .77
75. I throw things around when I’m angry. .58
155. I have rage outbursts. .72
146. I physically attack people when I’m angry at them. .68
101. My anger gets out of control. .69
170. I destroy things when I’m angry. .70
31. When I’m angry, I often lose control and threaten other people. .76
41. If I get angry, I can get so out of control that I injure other people. .71
8. I can become so angry that I feel capable of killing someone. .58
Impulsive Child (n = 7, α = .75)
135. I break rules and regret it later. .54
89. It is impossible for me to control my impulses. .56
143. I blindly follow my emotions. .39
126. I have trouble controlling my impulses. .63
99. I act impulsively or express emotions that get me into trouble or hurt other people. .61
149. I act first and think later. .47
40. I say what I feel, or do things impulsively, without thinking of the consequences. .54
Undisciplined Child (n = 5, α = .62)
16. I’m lazy. .47
164. I get bored easily and lose interest in things. .56
10. If I can’t reach a goal, I become easily frustrated and give up. .53
80. I can’t bring myself to do things that I find unpleasant, even if I know it’s for my own good. .43
142. I don’t discipline myself to complete routine or boring tasks. .38
Happy Child (n = 8, α = .82)
129. I feel that I have plenty of stability and security in my life. .67
45. I feel content and at ease. .71

13
11408 Current Psychology (2023) 42:11403–11414

Table 3  (continued)
Factors and Items Factor
Load-
ings

121. I feel loved and accepted. .68

117. I feel that I fit in with other people. .47
111. I feel safe. .62
98. I feel optimistic. .58
100. I feel listened to, understood, and validated. .43
153. I feel spontaneous and playful. .39
*33. I trust most other people.
167. I feel connected to other people.
Compliant Surrender (n = 6, α = .70)
7. I try very hard to please other people in order to avoid conflict, confrontation, or rejection. .42
22. I change myself depending on the people I’m with, so they’ll like me or approve of me. .64
123. I let other people get their own way instead of expressing my own needs. .50
47. I act in a passive way, even when I don’t like the way things are. .53
105. In relationships, I let the other person have the upper hand. .45
65. I allow other people to criticize me or put me down. .48
*53. I can put up with anything from people who are important to me.
Detached Protector (n = 9, α = .82)
82. I feel nothing. .59
116. I feel detached (no contact with myself, my emotions or other people). .68
6. I don’t feel connected to other people. .50
56. I feel indifferent about most things. .50
27. I don’t want to get involved with people. .42
103. I don’t care about anything; nothing matters to me. .50
106. I feel flat. .63
160. I feel cold and heartless toward other people. .67
169. I feel distant from other people. .65
Detached Self-Soother (n = 4, α = .62)
87. I like doing something exciting or soothing to avoid my feelings (e.g., working, gambling, eating, shopping, sexual activities, watch- .58
ing TV).
9. In order to be bothered less by my annoying thoughts or feelings, I make sure that I’m always busy. .46
151. I want to distract myself from upsetting thoughts and feelings. .54
50. I work or play sports intensively so that I don’t have to think about upsetting things. .36
Self-Aggrandizer (n = 10, α = .78)
173. It’s important for me to be Number One (e.g. the most popular, most successful, most wealthy, most powerful). .46
34. I do what I want to do, regardless of other people’s needs and feelings. .56
4. I get irritated when people don’t do what I ask them to do. .54
136. I do things to make myself the center of attention. .56
68. I have to be the best in whatever I do. .47
92. I’m demanding of other people. .58
138. I’m quite critical of other people. .53
2. I feel special and better than most other people. .31
1. I won’t settle for second best. .34
110. I feel I shouldn’t have to follow the same rules that other people do. .36
Bully and Attack (n = 6, α = .71)
71. Equality doesn’t exist, so it’s better to be superior to other people. .61
125. I mock or bully other people. .55
81. If you don’t dominate other people, they will dominate you. .62
127. Attacking is the best defense. .56

13
Current Psychology (2023) 42:11403–11414 11409

Table 3  (continued)
Factors and Items Factor
Load-
ings

157. By dominating other people, nothing can happen to you. .72

150. If you let other people mock or bully you, you’re a loser. .39
*128. I demand respect by not letting other people push me around.
*166. I am invulnerable.
*168. I belittle others.
Punishing Parent (n = 10, α = .83)
156. I’m a bad person. .63
23. I have impulses to punish myself by hurting myself (e.g., cutting myself). .48
93. I’m angry at myself. .68
104. It’s my fault when something bad happens. .50
133. I deny myself pleasure because I don’t deserve it. .56
48. I’m a bad person if I get angry at other people. .43
60. I don’t allow myself to do pleasurable things that other people do because I’m bad. .59
46. I don’t deserve sympathy when something bad happens to me. .55
38. I can’t forgive myself. .62
140. I deserve to be punished. .64
Demanding Parent (n = 5, α = .69)
77. I’m trying not to make mistakes; otherwise, I’ll get down on myself. .59
14. I sacrifice pleasure, health, or happiness to meet my own standards. .44
74. I’m pushing myself to be more responsible than most other people. .46
19. I’m under constant pressure to achieve and get things done. .62
20. I’m hard on myself. .62
*28. I know that there is a ‘right’ and a ‘wrong’ way to do things; I try hard to do things the right way, or else I start criticizing myself.
*154. I don’t let myself relax or have fun until I’ve finished everything I’m supposed to do.
Healthy Adult (n = 10, α = .77)
107. When necessary, I complete boring and routine tasks in order to accomplish things I value. .38
139. I assert what I need without going overboard. .65
148. I know when to express my emotions and when not to. .39
108. I feel that I’m basically a good person. .36
17. When there are problems, I try hard to solve them myself. .40
85. I can solve problems rationally without letting my emotions overwhelm me. .46
165. I have a good sense of who I am and what I need to make myself happy. .45
59. I can stand up for myself when I feel unfairly criticized, abused, or taken advantage of. .63
94. I can learn, grow, and change. .50
161. I’m capable of taking care of myself. .56
Conning and Manipulative (n = 11, α = .82)
49. I get other people to do what I want without them even knowing it. .67
112. I am a good liar. .51
13. I get people to trust me to take advantage of them. .50
102. I am good at manipulating people. .68
72. I make up stories about myself to impress other people. .52
36. I only do things to get a favor in return. .68
115. I use people’s secrets against them to get what I want. .54
64. I use my charm to make people do something for me. .47
3. I use peoples’ weak points to my advantage. .61
163. I exaggerate my misfortunes to get what I want. .68
109. I pretend to like people to get them to do something for me. .67
* 62. I can convince someone of anything.

13
11410 Current Psychology (2023) 42:11403–11414

Table 3  (continued)
Factors and Items Factor
Load-
ings

Predator (n = 10, α = .82)

70. I make plans to get rid of people. .65
171. I wait until the right time for revenge comes. .61
66. I do not mind hurting someone to get what I want. .60
95. I would not feel bad about killing someone. .42
44. I would physically hurt someone to make a point. .58
84. People who oppose me will pay a heavy price. .70
131. I could hurt someone badly and not lose sleep over it. .57
42. If I need to hurt someone, it is his own fault. .51
43. No one would realize I had plans to harm them. .62
122. I would kill someone for a lot of money. .38
Angry Protector (n = 8, α = .81)
76. I make a silent “wall of anger” that keeps people at a distance. .67
51. I use controlled anger to give the people the message “stay away from me”. .63
119. My hostile look and body language are enough for people to know that I am angry. .65
86. I let people know that I am angry just by how I look at them. .47
158. I use my anger to push people away. .82
61. I let people see my anger in non-verbal ways. .49
21. When I am irritated, I don’t want people coming near me. .39
134. I would let people know that I am angry just by how I look at them. .39
*15. Even if I don’t say that I am angry, people notice it.
*63. When I am angry, I do not say it; but people see it.
Suspicious Over-Controller (n = 13, α = .84)
25. If others are nice to you, they have hidden reasons for this. .55
88. By being on your guard, you can avoid others taking advantage of you. .60
124. Other people are plotting against me. .58
58. You have to be on your guard with others. .64
54. I have to be alert not to let others harm me. .48
132. I look for clues that reveal others’ plans to hurt me. .58
73. People are not as harmless as they seem. .44
141. I am vigilant for any possible threats against me. .55
152. I do not share personal information because people will use it against me. .48
37. It is okay to lie about other people. .53
11. If I don’t fight, I will be abused or ignored. .50
172. I try to find the motives behind what people do. .38
26. I have to be careful about the true intentions of others. .36
Forensic Self-Aggrandizer (n = 13, α = .83)
118. I do what I want to do, regardless of other people’s needs and feelings. .55
69. I feel superior to most other people. .57
32. I value power and status above everything else. .55
12. I put my needs above those of others. .42
83. I buy things so others can see I am successful (like expensive cars and clothing, and a beautiful house). .56
96. Other people see me as arrogant and aloof. .50
91. Being ordinary is the worst thing you can be. .46
55. I show other people that I am the boss. .62
57. I like to be the “top-dog”. .57
35. I don’t have patience towards people who are stupid. .45
78. I make sure that people see how powerful I am. .54

13
Current Psychology (2023) 42:11403–11414 11411

Table 3  (continued)
Factors and Items Factor
Load-
ings

24. I don’t like hanging around with losers. .32

114. Knowing me is a privilege. .38

*Due to low factor loadings these items were not included in analyses but kept in the questionnaire for future studies

Table 4  Mean and standard Baseline Baseline SD Retest Retest ICC 95% CI
deviations of baseline and retest Mean Mean (SD) SD
measurement and correlations
(N = 49) Vulnerable Child 2.89 1.09 2.61 1.15 .78 .61-.87
Angry Child 2.65 .78 2.50 .77 .80 .67-.88
Enraged Child 1.86 .71 1.66 .59 .59 .37-.75
Impulsive Child 2.28 .75 2.13 .64 .64 .44-.78
Undisciplined Child 3.26 .86 3.17 .89 .69 .51-.81
Happy Child 3.89 .92 3.74 .92 .88 .78-.93
Compliant Surrender 2.59 .86 2.38 .84 .82 .66-.90
Detached Protector 2.60 .88 2.57 1.02 .82 .70-.89
Detached Self-Soother 3.67 .97 3.46 .97 .65 .46-.79
Self-Aggrandizer 2.83 .77 2.54 .78 .77 .48-.89
Bully and Attack 2.15 .95 2.04 .94 .79 .66-.88
Punishing Parent 1.88 .72 1.74 .67 .67 .48-.80
Demanding Parent 2.89 1.00 2.80 1.02 .68 .49-.80
Healthy Adult 4.17 .63 4.08 .68 .76 .61-.86
Conning and Manipulative 2.06 .82 1.80 .66 .61 .38-.77
Predator 1.90 .73 1.71 .61 .65 .44-.80
Angry Protector 2.89 .99 2.79 .92 .80 .68-.88
Suspicious Over-Controller 2.73 .76 2.50 .87 .84 .67-.92
Forensic Self-Aggrandizer 2.72 .93 2.56 .87 .83 .71-.90

for 19-factor model when taken into consideration the χ2/df, The internal reliability coefficients of subscales ranged
CFI, TLI, RMSEA, and SRMR levels together. between α = .62 to .91. Accordingly, the internal reliability
Among 19 factors, additional to the happy child and coefficients can be considered at satisfactory levels accept-
healthy adult modes, the SMI-F has two dysfunctional par- able to excellent. Moreover, in the test-retest analysis, the
ent modes, five maladaptive child modes, five dysfunctional interclass correlation coefficients were satisfactory and
coping modes, and five forensic modes. The results sup- suggested that the schema modes are relatively stable con-
ported the 14-factor model, as congruent with the original structs. In congruent with the model, most of the inter-cor-
and the other versions such as Italian, German, and Urdu relations were statistically significant. Nonetheless, some of
(Panzeri et al., 2016; Reiss et al., 2012; Riaz et al., 2013) the intercorrelations were relatively low. A possible explana-
and suggested five other forensic modes. To test the concur- tion might be that the low but statistically significant cor-
rent validity of the SMI-F for general psychiatric symptoms, relations might be due to the large sample size. Additionally,
we administrated the BSI to a subgroup. To evaluate the overcompensation, avoidance, and compliant surrender cop-
relationships between psychological symptoms and schema ing patterns might neutralize the potential intercorrelations.
modes, we ran a multiple regression analysis. Due to high Some items of the questionnaire had low factor loadings.
inter-correlations between subscales of the SMI-F and BSI, Therefore, these eleven items were not included into later anal-
and sample characteristics, most of the loadings were non- yses but kept in the questionnaire for further studies especially
significant. Nevertheless, the adjusted R square value sug- on forensic samples and clinical settings. In sum, the current
gested that the SMI-F modes explained considerate amount study’s result exhibit considerable overlapping with the other
of the BSI variance, 64%. studies carried on clinical samples (e.g., Panzeri et al., 2016;

13
11412 Current Psychology (2023) 42:11403–11414

Table 5  Inter-correlations between the sub-scales

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

2 .56*
3 .31* .60*
4 .44* .56* .54*
5 .44* .45* .29* .46*
6 −.50* −.14* −.10* −.10* −.09***
7 .55* .32* .22* .43* .37* −.21*
8 .74* .60* .39* .47* .45* −.42* .45*
9 .25* .36* .21* .36* .22* .14* .18* .22*
10 .27* .50* .38* .45* .28* .08* .27* .39* .23*
11 .32* .56* .47* .46* .29* −.09* .32* .46* .27* .62*
12 .66* .50* .45* .54* .34* −.35* .57* .60* .18* .32* .41*
13 .49* .45* .32* .33* .21* −.12* .41* .44* .38* .48* .42* .52*
14 −.09* .22* .02 .01 .07*** .60* −.11* −.02 .37* .25* .14* −.13* .22*
15 .32* .48* .45* .50* .30* −.13* .43* .45* .17* .63* .68* .48* .36* .02
16 .31* .60* .63* .49* .29* −.09* .30* .49* .16* .51* .68* .48* .33* .09** .70*
17 .32* .61* .55* .41* .31* .02 .21* .42* .37* .40* .45* .33* .39* .30* .39* .50*
18 .44* .62* .42* .40* .31* −.03 .35* .51* .40* .52* .63* .44* .52* .32* .52* .57* .53**
19 .24* .56* .42* .41* .28* .10* .23* .40* .28* .80* .69* .30* .46* .33* .64* .59* .48* .59*

Note: *p ≤ .001. **p = .002. ***p = .009. 1: Vulnerable Child, 2: Angry Child, 3: Enraged Child, 4: Impulsive Child, 5: Undisciplined Child, 6:
Happy Child, 7: Compliant Surrender, 8: Detached Protector, 9: Detached Self-Soother, 10: Self-Aggrandizer, 11: Bully and Attack, 12: Punish-
ing Parent, 13: Demanding Parent, 14: Healthy Adult, 15: Conning and Manipulative, 16: Predator, 17: Angry Protector, 18: Suspicious Over-
Controller, 19: Forensic Self-Aggrandizer

Table 6  Multiple Regression 95% CI

Analysis Model Coefficients
Predictor Estimate SE Lower Upper t p Stand. Estimate

Intercept −1.06 .10 −1.26 −.86 −10.58 < .001

Forensic Self-Aggrandizer .27 .20 −0.12 .66 1.35 .178 .12
Suspicious Over-Controller .33 .16 .00 .65 1.20 .05 .142
Angry Protector −.04 .13 −.30 .22 −0.32 .75 −.02
Predator −.03 .143 −.31 .26 −.18 .86 −.01
Conning and Manipulative −.12 .13 −.38 .14 −.91 .36 −.06
Healthy Adult .13 .24 −.34 .60 .55 .59 .04
Demanding Parent .14 .12 −.08 .37 1.23 .22 .07
Punishing Parent −.07 .13 −.33 .20 −.50 .62 −.03
Bully and Attack −.44 .14 −.72 −.16 −3.11 .00 −.23
Self-Aggrandizer −.24 .19 −.61 .12 −1.30 .20 −.10
Detached Self-Soother .08 .11 −.13 .29 .75 .46 .08
Detached Protector .16 .15 −.13 .45 1.10 .27 .08
Compliant Surrender .22 .11 −.00 .44 1.96 .05 .10
Happy Child .10 .19 −.27 .47 .54 .59 .04
Undiciplined Child .25 .12 .01 .48 2.04 .04 .10
Impulsive Child .13 .12 −.12 .37 1.04 .30 .06
Enraged Child .26 .13 −.02 .49 1.81 .07 .12
Angry Child .35 .17 .017 .67 2.07 .04 .16
Vulnerable Child .83 .15 .53 1.12 5.51 < .001 .44

13
Current Psychology (2023) 42:11403–11414 11413

Reiss et al., 2012; Riaz et al., 2013). As mentioned earlier, the Conclusions
original SMI was developed and examined on clinical samples;
and studies on the Italian and German versions (Panzeri et al., Overall, the psychometric results indicate that the Turkish
2016; Reiss et al., 2012) were carried out mostly on clinical SMI-F is a satisfactorily reliable and valid measure that can
samples. Accordingly, we figured that, considerable similari- be of use for mode assessment in clinical and research set-
ties between the findings of the current and aforementioned tings. The inclusion of the forensic modes makes the inven-
studies might give some initial support for the clinical utility of tory a valuable tool that reflects the recent developments
the Turkish SMI-F. Besides, overlapping findings for 14-factor in the Schema Therapy Model. Given the limitations of
structure among the Urdu version (Riaz et al., 2013), which the study, there is a need for future research with forensic
was also conducted on a university student sample and the populations. Finally, we would like to re-emphasize that the
current study might be thought of as additional support for SMI-F is not only a forensic assessment tool but also can be
the Turkish SMI-F. used for a variety of clinical and research purposes, which
Overall, since the SMI-F has forensic modes as well as are a very crucial aspect of community health.
child, parent, coping, and healthy modes, it might be utilized
in not only forensic but also a variety of clinical and research Availability of Data and Material Data not available due to
purposes. Namely, assessing forensic issues and general men- ethical and legal restrictions.
tal wellbeing might be a crucial aspect of community health
services. In this context, the SMI-F has the potential utiliza-
tion of these purposes. Additionally, it could be possible to
assess a wide variety of coping modes and formulate a treat-
ment plan for the general population and forensic patients with
the SMI-F. Declarations

Limitations Conflict of Issue Disclosure On behalf of all authors, the corresponding

author states that there is no conflict of interest.
The major limitation of the study is the sample characteristic. Consent to Participate Informed consent was obtained from all indi-
Our sample includes only university students. Considering sev- vidual participants included in the study.
eral difficulties in collecting data from clinical settings at the
time of the study, we proceeded with the student sample as a
preliminary step for the adaptation of the SMI-F. Despite our
efforts to compensate for convenience sampling bias by using a
References
relatively large sample size, this might limit the generalizabil-
ity of our findings. Future studies should examine the predic- Aytaç, M., Köse Karaca, B., & Karaosmanoğlu, A. (2020). Turkish
tive validity of the SMI-F in forensic samples using clinically adaptation of the ShortSchema Mode Inventory. Clinical Psychol-
meaningful outcome variables such as risk assessments and ogy & Psychotherapy, 27(3), 346–363. https://​doi.​org/​10.​1002/​
recidivism rates. Second, we used a convenience sampling cpp.​2432
Bamelis, L. L., Evers, S. M., Spinhoven, P., & Arntz, A. (2014). Results
method, and our sample is not representative of the forensic of a multicenter randomized controlled trial of the clinical effec-
population in terms of gender and age. Most of the partici- tiveness of schema therapy for personality disorders. The Ameri-
pants in our sample were females. However, female offenders can Journal of Psychiatry, 171, 305–322. https://d​ oi.o​ rg/1​ 0.1​ 176/​
constitute only 4% of the Turkish prison population (Turkish appi.​ajp.​2013.​12040​518
Bernstein, D. P., Arntz, A., & de Vos, M. (2007). Schema focused
Statistical Institute, 2019). The average age of our sample is therapy in forensic settings: Theoretical model and recommenda-
20.43. 70% of the forensic population in Turkey is above 25 tions for best clinical practice. International Journal of Forensic
(Turkish Statistical Institute, 2019). There is a need for studies Mental Health, 6(2), 169–183. https://​doi.​org/​10.​1080/​14999​013.​
with a randomly selected forensic sample representing of the 2007.​10471​261
Bernstein, D. P., Ersayan, A. E., & Lobbestael, J. (2014). Schema mode
age and gender of the forensic population. Third, given that inventory forensic supplement (SMI-FS). Maastricht University.
we did not implement a longitudinal study design, the current Unpublished questionnaire.
cross-sectional findings limit the predictive value of the SMI- Bernstein, D. P., & Navot, L. (2020). Bridging the gap between forensic
F. Fourth, in terms of construct validity, this study presents and general clinical practice: Working in the “here and now” with
difficult schema modes. In G. Heath & H. Startup (Eds.), Creative
satisfactory discriminant and factor validity. As a further test methods in Schema therapy: Advances and innovation in clinical
of the construct validity, it would be beneficial to include a practice. Routledge.
personality assessment tool to examine the convergent valid- Bernstein, D. P., Nijman, H. L. I., Karos, K., Keulen-de Vos, M., de
ity of the scale. Vogel, V., & Lucker, T. P. (2012). Schema therapy for forensic
patients with personality disorders: Design and preliminary

13
11414 Current Psychology (2023) 42:11403–11414

findings of a multicenter randomized clinical trial in the Neth- Morvaridi, M., Mashhadi, A., Shamloo, Z. S., & Leahy, R. L. (2019).
erlands. International Journal of Forensic Mental Health, 11, The effectiveness of group emotional schema therapy on emo-
312–324. https://​doi.​org/​10.​1080/​14999​013.​2012.​746757 tional regulation and social anxiety symptoms. International
Chakhssi, F., Kersten, T., de Ruiter, C., & Bernstein, D. P. (2014). Journal of Cognitive Therapy, 12(1), 16–24. https://​doi.​org/​10.​
Treating the untreatable: A single case study of a psychopathic 1007/​s41811-​018-​0037-6
inpatient treated with schema therapy. Psychotherapy, 51(3), Moshagen, M., & Erdfelder, E. (2016). A new strategy for testing struc-
447–461. https://​doi.​org/​10.​1037/​a0035​773 tural equation models. Structural Equation Modeling, 23, 54–60.
Champely, S. (2020). Pwr: Basic functions for power analysis. R. Pack- https://​doi.​org/​10.​1080/​10705​511.​2014.​950896
age version 1.3-0. https://​CRAN.R-​proje​ct.​org/​packa​ge=​pwr Panzeri, M., Carmelita, A., De Bernardis, E., Ronconi, L., & Dadomo,
Derogatis, L. R. (1992). The brief symptom inventory (BSI): Adminis- H. (2016). Factor structure of Italian short Schema mode inven-
tration, scoring and procedures manual-II. Clinical Psychometric tory (SMI). International Journal of Humanities and Social Sci-
Research Inc.. ence, 6(2), 43–55.
Dunne, A. L., Gilbert, F., Lee, S., & Daffern, M. (2018). The role of Reiss, N., Dominiak, P., Harris, D., Knörnschild, C., Schouten, E.,
aggression-related early maladaptive schemas and schema modes & Jacob, G. A. (2012). Reliability and validity of the German
in aggression in a prisoner sample. Aggressive Behavior, 44(3), version of the Schema mode inventory. European Journal of Psy-
246–256. https://​doi.​org/​10.​1002/​ab.​21747 chological Assessment, 28(4), 297–304. https://​doi.​org/​10.​1027/​
Giesen-Bloo, J., van Dyck, R., Spinhoven, P., van Tilburg, W., Dirksen, 1015-​5759/​a0001​10
C., van Asselt, T., Kremers, I., Nadort, M., & Arntz, A. (2006). Reiss, N., Krampen, D., Christoffersen, P., & Bach, B. (2016). Reli-
Outpatient psychotherapy for borderline personality disorder: ability and validity of the Danish version of the Schema mode
Randomized trial of Schema-focused therapy vs transference- inventory (SMI). Psychological Assessment, 28(3), 19–26. https://​
focused psychotherapy. JAMA Psychiatry, 63(6), 649–658. https://​ doi.​org/​10.​1037/​pas00​00154
doi.​org/​10.​1001/​archp​syc.​63.6.​649 Renner, F., van Goor, M., Huibers, M., Arntz, A., Butz, B., & Bern-
Farrell, J. M., Shaw, I. A., & Webber, M. A. (2009). A schema-focused stein, D. (2013). Short-term group schema cognitive-behavioral
approach to group psychotherapy for outpatients with borderline therapy for young adults with personality disorders and personal-
personality disorder: A randomized controlled trial. Journal of ity disorder features: Associations with changes in symptomatic
Behavior Therapy and Experimental Psychiatry, 40, 317–328. distress, schemas, schema modes and coping styles. Behaviour
https://​doi.​org/​10.​1016/j.​jbtep.​2009.​01.​002 Research and Therapy, 51(8), 487–492. https://​doi.​org/​10.​1016/j.​
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covari- brat.​2013.​05.​011
ance structure analysis: Conventional criteria versus new alterna- Riaz, M. N., Khalily, T., & Kalsoom, U. (2013). Translation, adapta-
tives. Structural Equation Modeling, 6(1), 1–55. https://​doi.​org/​ tion, and cross language validation of short Schema mode inven-
10.​1080/​10705​51990​95401​18 tory (SMI). Pakistan Journal of Psychological Research, 28,
Keulen-de Vos, M., Bernstein, D. P., Clark, L. A., de Vogel, V., 51–64.
Bogaerts, S., Slaats, M., & Arntz, A. (2017). Validation of the Simpson, S. G., Morrow, E., van Vreeswijk, M., & Reid, C. (2010).
schema mode concept in personality disordered offenders. Legal Group schema therapy for eating disorders: A pilot study. Fron-
and Criminological Psychology, 22, 420–441. https://​doi.​org/​10.​ tiers in Psychology, 1, 182. https://​doi.​org/​10.​3389/​fpsyg.​2010.​
1111/​lcrp.​12109 00182
Keulen-de Vos, M. E., Bernstein, D. P., Vanstipelen, S., de Vogel, V., Şahin, N. H., & Durak, A. (1994). Kısa semptom envanteri (Brief
Lucker, T. P. C., Slaats, M., Hartkoorn, M., & Arntz, A. (2016). Symptom Inventory-BSI): Türk gençleri için uyarlanması. Türk
Schema modes in criminal and violent behaviour of forensic clus- Psikoloji Dergisi, 9(31), 44–56.
ter B PD patients: A retrospective and prospective study. Legal Turkish Statistical Institute (2019). Ceza Infaz Kurumu Istatistikleri.
and Criminological Psychology, 21, 56–76. https://​doi.​org/​10.​ TÜİK Haber Bülteni (Publication no. 33625). Retrieved from:
1111/​lcrp.​12047 https://​data.​tuik.​gov.​tr/​Bulten/​Index?p=​Ceza-​Infaz-​Kurumu-​Istat​
Lobbestael, J., van Vreeswijk, M., & Arntz, A. (2007). Shedding light istik​leri-​2019-​33625
on schema modes: A clarification of the mode concept and its Young, J. E. (1999). Cognitive therapy for personality disorders: A
current research status. Netherlands Journal of Psychology, 63, schema focused approach. Professional Resource Press.
76–85. https://​doi.​org/​10.​1007/​BF030​61068 Young, J. E., Klosko, J. S., & Weishaar, M. E. (2003). Schema therapy:
Lobbestael, J., van Vreeswijk, M., Spinhoven, P., Schouten, E., & A practitioner’s guide. Guilford Press.
Arntz, A. (2010). Reliability and validity of the short Schema Young, J. E., Arntz, A., Lobbestael, J., Weishaar, M. E., & Van-
mode inventory (SMI). Behavioral and Cognitive Psychotherapy, Vreeswijk, M. F. (2007). The Schema mode inventory. Schema
38, 437–458. https://​doi.​org/​10.​1017/​S1352​46581​00002​26 Therapy Institute.
Marsh, H. W., Hau, K., & Wen, Z. (2004). In search of golden rules: Yves, R. (2012). Lavaan: An R package for structural equation mod-
Comment on hypothesis testing approaches to setting cutoff values eling. Journal of statistical software, 48(2), 1-36. https://​doi.​org/​
for fit indexes and dangers in overgeneralizing Hu and Bentler’s 10.​18637/​jss.​v048.​i02.
(1999) findings. Structural Equation Modeling, 11(3), 320–341.
https://​doi.​org/​10.​1207/​s1532​8007s​em1103_2 Publisher’s Note Springer Nature remains neutral with regard to
Marsh, H. W., Hau, K., & Grayson, D. (2005). Goodness of fit in jurisdictional claims in published maps and institutional affiliations.
structural equation models. In A. Maydeu-Olivares & J. McAr-
dle (Eds.), Contemporary psychometrics (pp. 275–340). Lawrence
Erlbaum Associates.

13