Review of the self-sacrifice dimension of the dimensional clinical personality inventory

In some cases, psychological functioning can be related to personality disorders (Millon et al. 2004; Millon 2011). A personality disorder can be construed as the representation of a pattern where the individual’s personality cannot function properly in their environment. This can significantly impair living (Millon 1993; Widiger and Trull 2007; Millon et al. 2010). This definition supports findings in literature used to support the Diagnostic and Statistical Manual of Mental Disorders’ Fifth Version ([DSM-5]; American Psychological Association 2003). According to Skodol et al. (2011), an individual is diagnosed with a personality disorder when they show significant self- (identity and self-directedness dimensions) and interpersonal (empathy and intimacy dimensions) functioning impairment.

In literature, there are various proposals for assessing typical personality disorder characteristics, one of which is Theodore Millon's proposal. Millon’s theory offers a robust frame of reference for understanding personality and its disorders, as well as tools for measuring constructs (for further information, please see Millon 2011; Millon and Davis 1996; Millon et al. 2004; Millon and Grossman 2007a; 2007b; Millon et al. 2010). Based on Millon's proposal and DSM-IV-TR Axis II (American Psychological Association 2003), as well as the empirical support of a dimensional model (Schroeder et al. 1992), Carvalho and Primi (2015) developed the Dimensional Clinical Personality Inventory (Inventário Dimensional Clínico da Personalidade — IDCP). Particularly, the authors used Millon's pathological personality characteristics and DSM-IV-TR personality disorder diagnostic criteria as the basis for operationalizing the instrument's items.

IDCP is a self-report instrument for assessing pathological personality aspects consisting of 163 items divided in 12 dimensions, namely: Dependency, Aggressiveness, Mood Instability, Eccentricity, Attention Seeking, Mistrust, Grandiosity, Isolation, Avoidance of Criticism, Self-Sacrifice, Conscientiousness, and Impulsiveness. Each dimension is more closely related to one of Milton’s pathological personality patterns (2011).

Carvalho and Primi (2015, in press) and Carvalho et al. (2014a) investigated IDCP's psychometric properties. Resulting data show acceptable reliability indexes (Nunnally 1978) and validity evidence based on internal structure (analuzed with Classical Test Theory and Item Response Theory) and external variables (psychiatric diagnoses and NEO-PI-R). However, despite the instrument’s psychometric properties being acceptable, Carvalho and Primi (2015) have recommended the instrument be revised.

They specifically recommend revising particular aspects of certain dimensions. Some of these include the Conscientiousness (Carvalho et al. 2014c) and Attention Seeking (Carvalho et al. 2014b) dimensions, whose high selection rates suggest their items lean more towards assessing healthy characteristics instead of pathological characteristics. Additionally, Conscientiousness presented with an internal consistency (Cronbach’s alpha) of 0.69, lower than 0.70, the cut-off point. Carvalho and Primi (2015) have also suggested continuous revision of all of the instrument's dimensions in order to further refine and update IDCP.

This study aims to fulfill that need, revising IDCP dimensions using procedures figured in the research papers cited (Carvalho et al. 2014c; Carvalho et al. 2014a, 2014b, 2014c) contributing to the instrument's continued revision and improvement. This study will focus on one of IDCP's dimensions, namely, Self-Sacrifice. This dimension includes a tendency to help others instead of oneself—and even to act in a detrimental fashion towards oneself, in favor of others (Carvalho and Primi 2015). According to the authors, this dimension's definition closely relates to masochist functioning. However, no empirical studies directly supporting these data have been found. In studies of correspondence between IDCP dimensions and psychiatric diagnoses (Carvalho and Primi in press) and in profile assessments of patients diagnosed with personality disorders (Abela 2013), there has been no data on the masochist personality disorder due to it not figuring in DSM-IV-TR Axis II (American Psychological Association 2003).

However, considering this disorder relates to significant impairment of self in favor of others (Millon 2011), correlation data between IDCP and NEO-PI-R facets shown by Carvalho and Primi (in press) have been consistent. It is also worth noting that correlation values to the Depressivity (r = 0.37), Embarrassment (r = 0.31), and Modesty (r = 0.25) facets have all been significant, which indicates that individuals scoring high in Self-Sacrifice tend to devalue themselves, see themselves as incapable, and not consider their own deeds relevant. Based on these data and the need for continuous improvement of IDCP dimensions, this research paper aims to revise the Self-Sacrifice dimension and confirm its new version's psychometric properties. We have also sought to establish factors for the Self-Sacrifice dimension, which should make it possible to investigate different profiles according to this dimension's specific components.

The method has been divided into two parts to better fit this study's objectives. Firstly (Part I), the Self-Sacrifice dimension review procedures are developed. Then (Part II), we present information on data collection for verifying the new dimension's psychometric properties.

In the first phase, as previously mentioned, DSM-5, PID-5, SWAP, and Clark’s (1990) proposals were used. The adopted models generally do not include typical Self-Sacrifice characteristics or do include few of them. Due to that, we chose to revise Millon’s theory according to its updated literature (Millon et al. 2010; Millon 2011). Self-Sacrifice is construed as a set of characteristics that relate to an excessive lack of concern for oneself (self), evidently presenting along with tendencies to help others. This behavior is intimately related to a masochist or self-destructive personality disorder (American Psychological Association 2003; Millon 2011). No diagnostic categories typically related to this dimension were found in DSM-5 (American Psychological Association 2013). Out of PID-5’s 25 facets, Depressivity and Submission facets were considered. This instrument assesses personality traits according to DSM’s most recent version. Depressivity entails feelings of inadequacy, inferiority, emptiness, and lack of future perspective. Submission refers to the extent an individual is willing to self-negate to the benefit of others.

In SWAP, only the Dysphoria dimension was contemplated. It consists of items that indicate an individual's tendency to feel inferior, inadequate, prostrated, and to avoid social situations. In addition, in Clark (1990), characteristics related to the Self-Sacrifice dimension were selected from four different dimensions: Anhedonia, Self-Destructive, Passive-Aggressive, and Pessimism. Overall, aspects presented by Clark refer to lack or absence of the capacity to feel pleasure, devaluation of personal gain, lack of self-confidence, lack of interest for personal goals, influence of past experiences, and overstating difficulties. As previously mentioned, these models present few Self-Sacrifice-related dimensions. In phase 3, new items were established based on these proposals. A total of 189 items were established and distributed according to selected constructs.

In phase 4, out of 189 items, the ones considered to be more adequate were selected based on a consensus, resulting in 60 pre-selected items. In phase 5, items were divided into categories arbitrarily created by the authors to verify whether the they were representative of the dimension. Items from the dimension's original version were also included in this division. Categories are: Depressivity, referring to low mood, difficulty in finding meaning in life, negative self-image, and overstating difficulties; Submission/Self-Negation, containing items about an individual's need for satisfying others and putting them first; Anhedonia, containing items concerning an individual's difficulty in feeling pleasure; ‘Self-Devaluation’, consisting of sentences where the individual devalues personal gain; Self-Deprecation, including psychological suffering due to one's own actions; and Culpability, containing items regarding feeling guilty due to personal actions perceived as failures or mistakes.

Five items from the original dimension were included in the Submission/Self-Negation category and two of them were included in the Self-Deprecation category. These items were considered for the new pre-selected item set (N = 60). Thus, final selection consisted of 34 items, varying in five (Anhedonia and Self-Deprecation) to eight (Submission/Self-Negation) items. After completing this phase, IDCP Self-Sacrifice dimension's final version was obtained, containing 7 original items and 34 new ones, coming to a total of 41 items.

Values vary from 0.13 to 0.65. Total score relates more expressively to Depression and Self-Consciousness facets. The same is true for all Self-Sacrifice factors. Facets that related less to Self-Sacrifice factors were Impulsiveness and Anxiety. Additionally, as previously verified, most relations between Self-Sacrifice and Agreeableness presented values lower than 0.20. The highest values shown were between Hopelessness & Self-Devaluation and Trust (r = -0.34), followed by Depressivity and Trust (r = -0.31). Hopelessness and Self-Devaluation is the Self-Sacrifice factor that showed the highest number of significant correlations with Agreeableness facets.

Then, PID-5’s Depressivity and Submission dimensions were related to IDCP dimensions. All correlation values were shown to be statistically significant, varying mainly from moderate to high, from 0.31 (Masochism and Depressivity; Depressivity and Submission) and 0.77 (Hopelessness & Self-Devaluation and Depressivity). Total score presented similar correlation to both facets (r = 0.66 for Depressivity and 0.60 for Submission). Some Self-Sacrifice factors (Depressivity [r = 0.71] and Hopelessness & Self-Devaluation [r = 0.77; p = 0.01]) were shown to more expressively relate to Depressivity. Other factors (Masochism [r = 0.51; p = 0.01] and Submission (r = 0.53; p = 0.01]) were more closely related to PID-5's Submission facet.

Our first goal was to develop a new set of Self-Sacrifice dimension items that would also include original IDCP items (Carvalho and Primi 2015). We obtained 189 initial items from which 34 new items and 7 original items were selected for the instrument's final version. Categories created to better understand these items indicate their relation to typical Self-Sacrifice dimension characteristics, according to IDCP literature. For instance, some elements in PID-5 (Krueger et al. 2011), Clark (1990), and SWAP's (Westen and Shedler 1999) proposals are evidently related to the Depressivity category. In PID-5 and Millon's (2011) proposals, there are Submission/Self-Negation and Anhedonia-related elements. In PID-5, Clark, and Milllon's proposals we find correlation to Self-Devaluation and Self-Deprecation. In PID-5 and Clark's proposals, there are Culpability-related elements. Based on these data, the revised dimension was empirically investigated and four factors were found.

These factors are consistent with how literature—as previously mentioned—construes Masochism (tendency to help others more than oneself, including with self-detrimental behavior), Depressivity (tendency to have feelings of sadness and devaluation), Hopelessness and Self-Devaluation (believing one's own actions cannot result in a favorable outcome and self-blaming), and Submission (tendency to self-derogate in favor of others and perceive other people to be better). On the one hand, there were no a priori expectations about the number of factors found. On the other hand, the set of items was considered to be interpretable and consistent with Self-Sacrifice aspects (Millon 2011)—such as a tendency to prioritize others and devalue oneself. These factors and the revised dimension's total score's internal consistency coefficients were considered acceptable (Embretson 1996; Nunnally 1978). This datum suggest that it is possible to use the Self-Sacrifice dimension with an acceptable room for measurement errors, i.e., construct assessment may be prioritized when applying this IDCP dimension.

The Self-Sacrifice dimension's final 18-item set—four of them from the original dimension—was internally and externally (to other instruments' dimensions and facets) correlated. The revised dimension's factor and total score correlations indicate, on the one hand, cohesion among factors—mainly regarding high values in total score correlations. On the other, they make it possible for individuals who present total scores in a same range (for instance, a high score) to have a distinct profile in the Self-Sacrifice dimension. The possibility of creating profiles inside IDCP dimensions has been proved in literature (Carvalho et al. 2014c; Carvalho et al. 2014a, 2014b, 2014c). I.e., it is possible for two individuals to score high in the dimension but still have different profiles due to its four factors. Considering that one of revising IDCP dimensions' main goals is to develop internal profiles for each dimension, this possibility can be considered favorable evidence of fulfilling this study's objectives.

Self-Sacrifice and Neuroticism factor values were also shown to be consistent, given that these values were positive. This indicates that both factors converge in a similar range, i.e., a pathological range. Correlations with Agreeableness (all values lower than 0.20) indicate an absence of relation to this NEO-PI-R dimension, its revision, and its factors. Considering Self-Sacrifice aspects (Carvalho and Primi 2015)—regarding prioritizing others—and Agreeableness aspects (Costa and McCrae 2009)—regarding care and consideration for other people's wishes—, it was expected for the IDCP dimension's higher scores to be related to higher Agreeableness scores, which was not verified. However, due to each instrument's aim—IDCP assesses extreme pathological traits and NEO-PI-R assesses healthier traits—, the occurrence of opposite correlations would also have been coherent. Also, given that NEO-PI-R's dimension assesses healthier aspects and IDCP's dimension assesses pathological aspects, negative correlations could have been expected. In fact, both phenomena were observed in Agreeableness facets, as explained below.

To better understand these data, PID-5 facets were correlated to Agreeableness as an additional procedure mechanism. A similar pattern was observed: a -0.20 Depressivity correlation and a 0.12 Submission correlation. This datum suggests that aspects assessed by NEO-PI-R's Agreeableness dimension are not in line with typical Self-Sacrifice aspects, which could be expected. Future studies should investigate this issue further.

Specifically regarding Neuroticism facets and Self-Sacrifice variables, it is worth noting that some of IDCP dimension factors cannot be directly corresponded (for instance, Masochism and Submission). Other facets, such as Depressivity and Hopelessness & Self-Devaluation, present higher correlation values to the Depression facet. This is consistent with the fact that both facets relate to self-devaluation (Costa and McCrae 2009). Self-Sacrifice's total score was more expressively related to Depression and Self-Consciousness, which is consistent with Self-Sacrifice's subjacent sad mood, culpability, and submission components (Millon 2011). Regarding Agreeableness facets, Masochism presented more expressive and positive correlations to Altruism and Modesty. The NEO-PI-R facets suggest a tendency of thinking of other people's needs and not putting oneself above others (Costa and McCrae 2009), which is consistent with the intended assessment focus of IDCP's revised dimension. Submission also presented more expressive and positive correlation values to Modesty, indicating that this factor is partially related to a tendency of not putting oneself above others. Depressivity, however, correlated significantly and negatively to Trust. The same was true for Hopelessness and Self-Devaluation, suggesting that these two Self-Sacrifice factors are more related to depressive aspects and to difficulty in trusting others. This relation should be further investigated in future studies.

Lastly, PID-5 facets were the ones to show more expressive correlation values. This was expected, given that both the IDCP and PID-5 instruments were developed to assess pathological personality characteristics (Carvalho and Primi 2015; Krueger et al. 2011). These data can also be considered favorable to the Self-Sacrifice dimension. However, similarly to NEO-PI-R, PID-5 does not include any facets to specifically assess masochist functioning. Even so, IDCP's Masochism factor still presented higher correlation to the Submission facet. This is consistent with how significant submission is to a self-sacrificial functioning (Millon 2011). Additionally, Depressivity and Hopelessness & Self-Devaluation factors presented expressive correlation values with PID-5's Depressivity facet. As previously observed with NEO-PI-R, this is conceptually consistent with the fact that IDCP factors and this PID-5 facet include depressive aspects (Krueger et al. 2011). Correlation values between both the Submission factor and facet are also consistent, given that in both cases aspects assessed include the individual putting themselves after other people.

This study aimed to develop a new item set for IDCP's Self-Sacrifice dimension with adequate psychometric properties. Particularly, validity evidence was observed based on internal structure and external variables. Dimension and factor reliability indexes (internal consistency) were also verified. The revised dimension's high reliability coefficients and its relation to other instruments used suggest that this study's objective was reached and that the revised dimension proved to be more adequate than the original Carvalho and Primi (2015) one.

Regarding internal structure, we suggest the evidence presented favored the new Self-Sacrifice dimension, as four interpretable factors were observed according to proposals used as a basis for developing this dimension. Comparing IDCP's original dimension's score to the revised dimension's score, new item contents seem to contain more pathological characteristics than the original item set (please see NEO-PI-R and PID-5 correlations), which was expected and required given that the instrument assesses pathological personality characteristics. Additionally, it is worth pointing out that each factor could have consisted of a higher number of items, from a psychometric standpoint. However, we sought to keep the number of items per factor to a minimum in order to make the instrument useable in the clinical setting.

This study also included limitations that must be stated. The first of these limitations relates to population: the number of participants and their characteristics. Future studies should aim to replicate diverse groups and their structure, particularly including patients diagnosed with personality disorders. Additionally, when replicating these factors, it is paramount that reliability indexes be confirmed based on the level of subjects in the latent construct, for instance, through local precision (Daniel 1999). We also suggest investigating the use of this revised dimension in the clinical setting.