Taxonomy and hierarchy of psychological abuse strategies in intimate partner relationships

Título: Taxonomía y jerarquización de las estrategias de abuso psicológico en la violencia de pareja. Resumen: Este estudio se centra en analizar los componentes del abuso psicológico en la violencia de pareja adulta y contra la mujer desde un enfoque psicosocial. Tras realizar una revisión de los estudios sobre el tema, se plantea como primer objetivo el proponer una nueva taxonomía de estrategias de abuso psicológico, acompañadas de sus respectivas definiciones operativas. A partir de esta nueva clasificación, se traza un segundo objetivo que se centra en someter esta clasificación al juicio de un grupo de expertos a través de un estudio Delphi, en el que han participado 32 expertos de los ámbitos académico-universitario y profesional. Estos expertos debían juzgar la adecuación del sistema de categorías, además de evaluar cuantitativamente la severidad que aporta cada una de las estrategias al fenómeno global del abuso psicológico en la pareja. Los resultados muestran que los expertos ratifican la nueva clasificación de estrategias y sus definiciones operativas, avalando así su validez de contenido y constructo. Además, jerarquizaron las estrategias de abuso, atribuyendo mayor severidad a las de tipo emocional, seguidas de las relacionadas con el contexto cercano, de las de tipo cognitivo y las de carácter conductual. Se discuten los resultados, sus limitaciones y las implicaciones que pueden derivarse de ellos. Palabras clave: abuso psicológico; violencia de pareja; método Delphi; taxonomía; severidad. Abstract: Taking a psychosocial perspective this study analyses the components of psychological abuse in intimate adult partner violence against women. Following a review of studies on the subject the main objective that emerged was the need for a new taxonomy of psychological abuse strategies, one which should include their corresponding operational definitions. The proposed new classification was then evaluated by means of a Delphi study involving 32 experts from the academic and professional worlds. These experts were asked to assess the suitability of the system categories and to rate the severity of the impact made by each of the strategies on the global phenomenon of psychological abuse in couples. The results show that the experts ratified the new classification of strategies and the corresponding operational definitions, thereby endorsing their content and construct validity. When rating the strategies according to the severity of their impact, those of an emotional nature were considered the most severe, followed by those related to the immediate context, those of a cognitive nature and, finally, behavioural strategies. We discuss the results and their implications.


Introduction
Intimate partner violence against women is considered the most widespread and worrying form of interpersonal aggression and it has become a main concern of health professionals, jurists and politicians.Indeed, it is now regarded as a major public health problem (Heise, & García-Moreno, 2002) impinging on all countries, cultures and social backgrounds (Levinson, 1989), one which has serious consequences for the women who endure it (Dunkle et al., 2004) and whose effects mark not only individuals but society in general (Henning, & Klesges, 2003).
As a result of this growing concern, research into intimate partner violence has gradually become more imperative, and numerous studies have now been conducted on different aspects of the phenomenon.However, the study of psychological abuse has generally been regarded as being of secondary importance, it being treated as a complementary aspect of the study of physical abuse.In relation to the studies that reflect this complementarity, some researchers (e.g.Henning, & Klesges, 2003) conclude that this approach to the study of psychological abuse is due largely to the fact that the studies' authors believed that it always accompanies physical violence rather than existing on its own.This con-ception of psychological abuse is one of the reasons why less research has been carried out to determine the incidence of psychological abuse in couples.However, more recent studies indicate that this type of abuse is very common (e.g.Ulla et al., 2009), and it has also been shown to have serious physical and psychological consequences and to be an important predictor of physical violence (e.g.Follingstad, Rutledge, Berg, Hause, & Polek, 1990;Hennings, & Klesges, 2003).
Despite the apparent importance of psychological abuse as a component of intimate partner violence, there is currently no consensus when it comes to differentiating and characterizing this phenomenon.In fact, it is defined in a wide variety of ways.Interestingly, when these definitions were drawn up the fact that psychological abuse constitutes a set of behaviours was not really taken into account, and the emphasis was therefore laid on the aspect of harassment.Furthermore, these definitions focus mainly on the consequences of psychological abuse and the harm it causes, with the accent on reactions rather than actions.For example, Loring (1994) defines psychological abuse as an active process in which one partner belittles and destroys the other partner's identity, while according to Tolman (1992) it is a set of behaviours intended to harm the other person's welfare by undermining their self-esteem, their sense of control and security.Shepard and Campbell (1992) conceived of it as a set of violent behaviours designed to terrorize the victim.Finally, Marshall (1992) sees psychological abuse as the anales de psicología, 2014, vol. 30, nº 3 (octubre) effect of an act and not the act in itself, as well as being the result of everyday intra-and interpersonal processes that can take on many guises, ranging from those that are more serious and threatening to others in the form of games, jokes or tokens of love.In addition, it is based on everyday communication and interaction between partners, which gradually undermines the psychological, emotional and behavioural competence of the victim.

Classification of psychological abuse strategies in intimate partner relationships
In some studies the way in which abusive strategies were classified led to psychological abuse being conceptualized as a unifactorial construct (e.g.Calvete, Corral, & Estevez, 2005;O'Leary, & Curley, 1986).However, in most cases it has been considered as a set of components or factors and, therefore, as a multifactorial construct (Kelly, 2004).At all events, different studies have produced different classifications that reflect the strategies which supposedly define the scope of psychological abuse in couple relationships.Some of these studies are cited below.The premise of our analysis of these studies is that the phenomenon of psychological abuse can be conceptualized using a psychosocial approach in which the strategies of abuse that define it may be either indirect, influencing the victim's immediate context, or direct, where the intention is to influence the victim's emotions, cognition and behaviour.
Although there are published studies that examine the differentiation of psychological abuse there is still no general consensus regarding the whole set of behaviours or forms of psychological abuse that characterize the phenomenon (Kelly, 2004).Consequently, professionals working in this field are forced to use different categorizations or complementary models to make more complex assessments.The solution to this problem of the differentiation of psychological abuse lies in a taxonomy that groups the different strategies of psychological abuse and provides functional definitions for each of them.Borjesson, Aarons and Dunn (2003) noted the need for research on the severity and frequency of psychological abuse, as well as the effects of chronic exposure to it.Studies of the severity of psychological abuse can be broadly classified into two types: those that have presented indices of severity of psychological abuse as their main findings and those that have focused on exploring how it affects the health of victims.With respect to the former, it is noteworthy that relatively little scientific research has sought to establish the degree of severity of psychological abuse strategies.Valdez-Santiago et al. (2006), working with this parameter, described the development of a scale for measuring intimate partner violence that included a severity index.This index was based on evaluation by a panel of experts of the severity of each of the behaviours listed in the categories.Accordingly, a value was established for each of the behaviours, which was reflected in the description of each category.This value increases with the frequency of the type of abusive behaviour in question.

Severity of psychological abuse strategies
As regards research focused on how psychological abuse affects women's health, mention should be made of the study by Aguilar and Nightingale (1996), in which they describe the link between emotional abuse that seeks to control the other person and the latter's levels of self-esteem.Their results show that women who have been physically abused do not suffer from lower levels of self-esteem.Other studies highlight the relationship between psychological abuse and an increased probability of showing signs of depression, suffering from anxiety and post-traumatic stress, having suicidal thoughts (Pico-Alfonso et al., 2006), develop-ing a chronic illness, suffering from a mental disorder or physical injury, and being more prone to substance abuse (Coker et al., 2002).In addition, it has been shown that psychological abuse is the most effective predictor of separation or divorce (Gorten, Berns, Jacobson, & Gottman, 1997), and also that strategies of abuse that disturb the victim's emotions are those most closely linked to the desire to break off an abusive relationship (Arias, & Pape, 1999;Dutton, Goofman, & Bennett, 1999;Henning, & Klesges, 2003).
Several authors have highlighted the need for more indepth studies on the severity of the strategies that shape psychological abuse, not least because such research could provide interesting and more precise data with which to develop a valid screening tool for detecting this type of abuse.Consequently, and taking into account the abovementioned literature, the aim of the present study was to provide a new definition of psychological abuse in intimate partner relationships, defining and classifying its components (the goal of phase 1) and evaluating and ranking them according to their severity (the goal of phase 2).In order to achieve these theoretical and practical goals the study was divided into two phases.
Phase 1: Proposal of a taxonomy of psychological abuse strategies in intimate partner relationships.
The aim here was to propose an integrated classification of factors or components of psychological abuse by developing a new taxonomy containing the psychological abuse strategies linked to adult intimate partner violence against women, as well as their functional definitions.By explaining precisely the behaviour corresponding to each strategy the new taxonomy would help define the boundaries between this and other types of abuse.

Method Procedure
The development of the new classification of abuse strategies drew on three sources: 1) a review of studies, taxonomies and instruments for measuring psychological abuse, found in the main scientific databases; 2) direct work with witnesses and victims who have suffered psychological abuse; and 3) reading and consultation with health experts on the subject of intimate partner violence.
Following the search for relevant material we selected those scientific papers whose goal was to define and measure the phenomenon of psychological abuse in the context of intimate partner relationships.The aim here was to identify the classifications and elements of abuse contained in these publications so that they could be systematically classified and analysed.Subsequently, four members of the research group (with previous experience in the study of the different types of psychological abuse) extracted, pooled and integrated factors of abuse identified in the whole body of reviewed material, the aim being to reach a consensus on the new categorization of abuse strategies.To this end the four researchers classified each of the strategies into groups of categories and sub-categories until they reached a consensus on the taxonomy presented in this paper.The kappa value was 0.84 (for the symmetric measures, p <0.1), demonstrating a high degree of consensus.

Results
The result of this first phase was a proposed taxonomy of psychological abuse strategies in adult intimate partner violence against women.This taxonomy incorporates a broad spectrum of behaviours ranging from the most obvious to the subtlest.Its terms and definitions were chosen on the basis of a conception of psychological violence focused on the abusive action in itself; they do not go on to define the consequences of these behaviours for victims (see Table 1).
The proposed taxonomy contains six categories, each with several sub-categories and classified according to four areas of abusive behaviour: emotion, cognition, behaviour and context.From a psychosocial perspective this means that psychological abuse affects the victim both directly (effect on emotion, cognition and behaviour) and indirectly (through the immediate context).
The first three categories deal with those strategies of abuse that affect the victim's context, and hence they are also described as indirect strategies of abuse.The three remaining categories focus more directly on the sufferer and are therefore classified as direct strategies of abuse.These actions affect the emotional, cognitive and behavioural processes of the victim.In the category covering emotions (emotional abuse) those actions whose intent is particularly to affect the victim's feelings and emotions are grouped together.
The type of abuse that undermines cognitive processes (imposition of one's own thinking) is covered by the category and sub-categories that contain all those acts whose object is to discredit the ideas of the other person, in addition to rejecting any interpretation of the relationship as abusive.
Lastly, the taxonomy considers strategies that most directly affect the victim's behaviour (imposition of a subservient role), those where the goal is to dominate the victim, demanding full-time dedication to satisfying the abuser's demands.
In all cases, each abusive strategy is accompanied by a functional definition.This enables a better differential analysis of each strategy, as well as better differentiation of the phenomenon of psychological abuse as a whole.

Sexual coercion:
Forcing the other person to perform or watch unwelcome sexual acts.

Control-debilitation of physical and mental health:
Imposing a pattern of conduct that undermines the physical and mental health of the other person (through exhaustion, sleep restriction, physical abuse, incitement to drug abuse) or preventing the other person from treating health problems by consulting professionals and using standard treatments.

EMOTIONAL ABUSE:
Actions intended to influence the feelings and emotions of the other person, in order to manipulate that person and gain her submission.4.1.Self-interested activation of positive emotions: Offering calculated expressions of love or making pleasant promises to the other person, following abuse, in order to obtain her forgiveness and forestall distancing or separation.Also, offering forgiveness or more lenient treatment than usual for some misdemeanour that the other person has been accused of.

Intimidation and threats:
Threatening the other person to make her aware of the physical, psychological or other harm that may come to her, her children or the people around her if she does not do as she is told.

Redefinition of reality:
Rejecting the other person's perception of a problem situation, minimizing or denying any personal responsibility and reinterpreting it in a self-interested way.

Self-interested idealization of the bond of dependence:
Inducing or reinforcing heavily the other person's belief in the importance of mutual dependence, where the bond between the partners is perpetual and indestructible, and where commitment must be total to overcome all difficulties.
Behaviour 6. IMPOSITION OF A SUBSERVIENT ROLE: Imposing one's authority on the other person, forcing her into a subservient role at the service of the demands and whims of the abuser.
anales de psicología, 2014, vol.30, nº 3 (octubre) Phase 2: Validation and hierarchy of psychological abuse strategies in intimate partner relationships The proposed taxonomy, established in the previous phase, was submitted for evaluation by a group of experts on intimate partner violence.The aim here was to validate the contents of this taxonomy and rate the degree of severity of each of the strategies it describes.

Method Design
The procedure chosen to meet these objectives was the Delphi method, one which has proven useful in situations where individual judgments can be combined in order to examine a field or phenomenon that is not fully understood or where there is no general consensus (Polit, & Hungler, 1999).The Delphi method has been applied, for example, in the field of elder abuse (Daly, & Jogerst, 2005) and abuse in the workplace, or mobbing (Rodríguez-Carballeira, Escartín, Visauta, Porrúa, & Martín-Peña, 2010).
In the case of the present study the choice of the Delphi method was justified by the complexity inherent in the study of psychological abuse in intimate partner relationships, particularly when it comes to establishing where its boundaries lie.Moreover, this method ensures the anonymity of respondents, yields a statistically interpretable group response, and allows interaction through controlled feedback (e.g.Adler, & Zliglio, 1996).Since the participants do not interact directly it is easier to avoid biases generated by knowledge of the experts' identities and the pressure that this may cause (Lindstone, & Turoff, 1975).Furthermore, this method assigns the same role to all the participants in the decisionmaking process, regardless of any geographical restrictions that may exist (Geist, 2010).
In light of the above it was decided that the Delphi method could help establish the content and construct validity of the strategies that characterize psychological abuse in intimate partner relationships.

Participants
Initially, 69 Spanish specialists were invited to join a panel of experts, although the sample was finally narrowed down to 32, all of whom participated voluntarily and anonymously.This drop-out (De Leeuw, 2001) was due to the fact that some of the selected participants did not complete all the tasks proposed for this study.In other words, only those who carried out all the proposed activities were included in the study.The choice of participants was initially based on a non-probabilistic sample in which consideration was given to those professionals who were easily contacted and who represented both the professional (65.62%) and ac-ademic and scientific worlds (34.38%).The characteristics of the final sample of 32 experts were as follows: there were 26 women and 6 men; 30 were from the field of psychology, 1 from social work and 1 from social education; 21 were recognized for their professional practice in the field of violence against women, and 11 for their academic and research experience in this area.
Two criteria were used to select the group of experts: -Their professional experience in the field of prevention and treatment had to cover not only women abused by their partners but also aggressors.-The number of relevant publications or studies in which they had participated and which dealt with this issue, as well as their years of clinical experience: the latter had to be at least five consecutive years.

Procedure
The starting point was the taxonomy of psychological abuse strategies in intimate partner relationships that was developed in Phase 1 of this study (see Table 1).Subsequently, two lists were drawn up and given to the panel of experts.The characteristics and purpose of these lists were as follows: List 1: this contained the names of the categories and sub-categories of psychological abuse strategies and their functional definitions.The goal here was to gather the experts' opinions on the strategies of abuse and the respective operational definitions.The information was presented in tabular form, with a box adjacent to each category and subcategory in which they could indicate their rating.The rating scale in each case was 0 to 100.
List 2: This list included the same strategies and operational definitions, but this time accompanied by the following statistical data: the first rating given by the same expert, as well as the mean, the standard deviation, and the maximum and minimum ratings found in the initial ratings of all 32 experts.The rating scale was again 0 to 100 and the distribution of ratings was governed by the same rules as in List 1.
An "Observations" box was included in both the first and second list, and here participants could offer explanations or make suggestions about their ratings or the content of the taxonomy.
The specialists and the working group communicated by email, a procedure also used in the Delphi study by Hagen et al. (2008), among others.After initial contact, and once the panel candidates had been informed of the details of the study and had agreed to participate, they were emailed List 1.After their answers had been analysed as required, the second list was then sent out.In both cases the list in question was accompanied by another document that explained how to perform the evaluation.
In the first round the experts were each asked to rank each group of categories and sub-categories by distributing 100 points as they deemed appropriate (i.e. according to the anales de psicología, 2014, vol.30, nº 3 (octubre) weight they considered each one carried within its group).After receiving the first set of completed lists with the corresponding ratings we calculated the most representative statistics for the data obtained (Hagen et al., 2008).Next, each expert was once again asked to evaluate and rank the categories and sub-categories, this time taking into account the statistical data obtained in the first round.This included the option of repeating the same ratings.
Responses were analysed using SPSS 15.0 software.Quantitative analysis provided descriptive statistics (mean, standard deviation, maximum score and minimum score) for each of the categories and sub-categories that were ranked by the panel of experts.The Kolmogorov-Smirnov test was applied to verify whether the results fitted a normal distribution, and they were also scanned for Spearman's ρ correlations between the two scores.In addition, Wilcoxon's Z test was used to verify the consensus reached by the panel of experts.

Results
The goal of this phase was to analyse the severity of each of the psychological abuse strategies in intimate partner relationships in relation to the phenomenon as a whole.To this end, the mean ratings from the second round were taken as a benchmark.Means were obtained for each category and sub-category of psychological abuse, as reflected in Figure 1, which shows the corresponding values and presents the categories and sub-categories in rank order.The data provided by the second round of evaluation were used because statistical tests showed there were no significant differences between the means of the first and second ratings.Therefore, there was no need for a third round.
It can be seen in Figure 1 that the category of emotional abuse was considered the most severe ( = 25.03), with the highest ratings being given to the sub-categories referring to behaviours intended either to intimidate the other person (intimidation and threats → = 18.23) or to denigrate or insult the reputation and dignity of the other person (contempt, humiliation and rejection as a person → = 19.95).
The category judged to be the second most severe was control of personal life ( = 19.06).As regards its sub-categories the most serious was considered to be forcing the partner to perform or watch unwelcome sexual acts (sexual coercion → = 22.41), followed by the strategy that imposes some kind of behaviour leading to physical and mental exhaustion, and, lastly, by strategies that prevent the other person from receiving professional healthcare or imposing behaviour that can affect the other person's physical well-being (control and debilitation of mental and physical health → = 22,56).
In third position were those strategies that fall into the category of isolation ( = 18.09), especially those referring to abusive actions that seek to separate or distance partners from their families (isolation from the family → = 29.44),their friends and the circle of people in their social environment (isolation from friends and social support network → = 26.75).
Within the category of imposition of one's own thinking ( = 13.34), which comes fourth in the hierarchy, the subcategory that obtained the highest rating refers to abusers' attempts to distort their partners' perception of the problem situation (redefinition of reality → = 34.48).
The fifth ranked category was that referring to control and manipulation of information ( = 12.78).Here, the sub-category given the highest rating concerns forms of abuse aimed at forcing the other person to keep the abuse secret, as well as obstructing any attempt to find help (concealment of abuse → = 59.22).Lastly, the category considered to be the least severe or serious of the six was imposition of a subservient role ( =11.69).Regarding the distribution of the second-round ratings, most of the categories present a symmetrical profile, since the values obtained were within the range of -2 to 2. For further details, see Table 2.  Following these initial analyses the two ratings obtained through the Delphi procedure were compared using Spearman's correlations, since the results of the Kolmogorov-Smirnov test had shown that the data were not normally distributed (p < .05).All correlations (Pearson's rho) were found to be significant at .05 (bilateral), and most of them were over .80.This indicates that the ratings proposed by the panel of experts did not vary much between the first and second rounds, thus confirming a notable interdependence (see Table 2).
To compare the two means obtained for each of the categories and sub-categories a non-parametric test was applied to two related samples (Wilcoxon's Z).The comparison of means showed there was no significant variation in the judgments or assessments made by the experts with regard to the two scores, which were very similar (see the eighth column of Table 2).

Discussion
This study proposes a categorization of psychological abuse strategies that appear in adult intimate partner violence against women.The classification includes the categories and sub-categories of abuse strategies that generate a process of psychological violence when applied systematically and continuously.It was drawn up from a psychosocial perspective and is intended to include behaviours ranging from the most obvious to the subtlest.The classification also includes the functional definition of each of the strategies, thereby contributing to a more precise understanding and better differentiation of the phenomenon of psychological abuse in intimate partner relationships.The panel of experts considered the proposed taxonomy, including the functional definitions, to be adequate.In some cases they suggested a few minor changes and these were introduced later.
The results obtained here are in line with the studies conducted by Borjesson et al. (2003), Hegarty, Bush and Sheehon (2005) and Hegarty, Sheehon and Shonfeld (1999), in which the existence and relevance of the factor referred to as emotional abuse is defended.However, the taxonomy presented here differs in that emotional abuse includes a greater number of abusive strategies that affect the victim's emotional processes, thus giving this important aspect a broader definition within the boundaries of psychological abuse.
The categories covering the control and manipulation of information and control of personal life bear a certain connection to those overt actions (Marshall, 1999) whose aim is to supervise the behaviour of women, as well as to the abuse of control exercised by the abuser over his victim, a category reflected in the work of Pitzner and Drummond (1997).There is also a link to the category of control established by Vázquez et al., (2008), to that of external control proposed by Walker (1985), to Ward's environment control (2000), and to the category of isolation control activity suggested by Wolfson (2002).The category of isolation finds its counterpart in the isolating category put forward by Jones et al. (2005) and Marshall (1999), as well as in the description of isolation proposed by Sonkin, Martin and Walker (1985) and Vázquez et al. (2008), in Kasian and Painter's isolation and control (1992), in Tolman's domination/isolation (1989), and in Walker's social isolation (1979).The strategy of imposing one's own thinking has been less widely considered by other researchers and their classifications cover less ground than the category proposed here.There is, nonetheless, a certain similarity to the category that Sonkin et al. (1985) labels mental degradation and the category that Ward (2000) calls minimizing/blaming.As for the imposition of a subservient role, the latter author has a category called performance orientated.With regard to our sixth category, it should be noted that the experts' justification for ranking it at the bottom of the hierarchy, and considering it less severe, was that the behaviours it refers to were already reflected in the other five categories.Therefore, it was felt that this sixth category offered no new information.What it does reflect, however, and here the experts agreed, is the objective pursued by aggressors when they abuse their partners.
In terms of the structuring of psychological abuse the results of this study are comparable to those obtained in other areas of abuse by Martín-Peña, Rodríguez-Carballeira, Escartín, Porrúa and Winkel (2010) and Rodríguez-Carballeira et al. (2010).The first of these studies focuses on the violence of persecution exercised by a terrorist group, while the second examines the context of work (mobbing or workplace bullying).Although the contexts are different, and despite there being observable differences in the components of abuse and its objective, an analogous classification of psychological abuse strategies emerges in all three cases.This suggests that psychological abuse may be a phenomenon with certain features that are common to different contexts.
Another aim of the present research was to gauge the degree of severity or gravity of each of the components of psychological abuse in adult intimate partner violence against women, and then to present these components or abusive strategies in a hierarchy.This hierarchy was derived from the opinions of a panel of experts, who considered those strategies that attack a person's emotional processes to be the most severe, followed by those that seek to control and isolate the victim.Strategies through which abusers impose their own thinking, those used to seek control and manipulate information and, lastly, those that impose a subservient role were given much lower ratings and were ranked further down the hierarchy.
Within the wide range of emotional abuse strategies special mention should be made of the importance given to those in the sub-category of contempt, humiliation and rejection as a person.This sub-category contains all those actions aimed at expressing contempt for and rejection of the other person through insults, slurs, taunts, ridicule, defamation, slander and similar tactics, which include both verbal and nonverbal behaviour.In line with this aspect, there are various studies that highlight the gravity of this way of expressing psychological abuse.For example, Goetz et al. (2005) found that men who insult their partners are also physically violent, and in this context the authors describe a series of behaviours used in an attempt to gain control over the other person.The strategies that best predict controlling behaviour and the use of physical violence by the aggressor were labelled by these authors as derogating value as a partner/mental competency.Other research has similarly shown that certain behaviour deemed abusive, such as insults, criticism, ridicule and denigration, is closely related to the onset of physical abuse (Murphy, & Hoover, 1999;Sullivan, Parisian, & Davidson, 1991).Likewise, if verbal abuse appears at the beginning of an abusive relationship, the chances are that abuse will continue (Schumacher, & Leonard, 2005).
The hierarchy of strategies in the proposed taxonomy shows both similarities and contrasts with a previously reported hierarchy of mobbing strategies derived from experts (Rodríguez-Carballeira et al., 2010) and laypersons (Escartín, Rodríguez-Carballeira, Zapf, Porrúa, & Martín-Peña, 2009), as well as with a hierarchy of violence of persecution strategies (Martín-Peña et al., 2010.)In all three cases, the category of emotional abuse was considered the most severe, and within it the highest ratings were given to the sub-categories concerning strategies of discredit, humiliation, intimidation and threats.There are, however, a number of interesting differences with respect to the hierarchy of mobbing strategies.For example, the types of psychological abuse related to context assume greater importance in intimate partner violence, which may be explained by the fact that the goal is domination of the victim, with whom the abuser shares an intimate relationship.By contrast, the purpose of mobbing is basically to exclude the worker in a work context that is usually more difficult to control.
In addition to its role in training professionals from different disciplines, the taxonomy proposed here may also contribute to a more precise understanding and definition of psychological abuse in adult intimate partner violence anales de psicología, 2014, vol.30, nº 3 (octubre) against women, with practical implications for both health and legal practice.The taxonomy could also be employed as a stand-alone assessment guide.For instance, professionals could use it to detect the presence of psychological abuse strategies in an abusive intimate partner relationship, and, thanks to the accompanying operational definitions, it could help them to identify more precisely the specific behaviours or attitudes that are clear examples of each of the abuse strategies under consideration.This is clearly an innovative contribution in this area of research.Similarly, the hierarchy offers the possibility of evaluating the impact these behaviours have on victims' health, since by ordering the strategies according to their severity, professionals can better assess how much harm is done depending on the type of abusive behaviour.This information is useful when planning a treatment programme for the persons involved.
One of the main limitations of this study is that the panel of experts was not randomly selected but, rather, chosen from the most accessible pool once the preliminary criteria had been established.In addition, the lack of optimum diversity in the sample may have generated a degree of bias that affected the drift of the experts' responses.In a similar vein, the fact that the Delphi method uses information transmitted by a group may have favoured a certain amount of social pressure within this group.This could have led to a degree of conformity when responding individually.
It is also acknowledged that the construct of psychological abuse as part of intimate partner violence is a complex issue, one that is closely related to the context in which it occurs and, therefore, to the culture of that social environment.This aspect should obviously be taken into consideration when interpreting the results of this study and for un-derstanding the phenomenon, and it follows that it would be worth conducting further research in which cultural variables carry more weight.For example, one could carry out comparative cross-cultural studies in order to examine variations in how the phenomenon is defined and to discover what kind of psychological abuse strategies are most frequent in each cultural setting (for a comparison of this kind within the mobbing domain, see Escartín, Zapf, Arrieta, & Rodríguez-Carballeira, 2011).
In addition to cross-cultural comparisons it would be interesting to contrast the psychological abuse that arises in intimate partner relationships with that in other contexts.In this regard, it is worth mentioning the study by Ward (2000) that analyses the similarities between psychological abuse strategies observed in couples and those applied by manipulative or coercive groups.A study along similar lines was conducted by Wolfson (2002), who, with the aim of establishing a scale of psychological abuse, compares the control exercised through intimate partner violence with that exerted by cults.Finally, it would also be interesting to analyse whether the taxonomy of psychological abuse strategies that is presented here is able to capture the psychological abuse that occurs in intimate partner violence involving adolescents and young people, and whether the severity of each of the strategies is the same as in the case of adult partner violence.
4.3.Contempt for, humiliation and rejection of the other person: Attacking the other person by showing an attitude of contempt and rejection through insults, slurs, taunts, ridicule, defamation, slander and other defamatory tactics.4.4.Contempt for roles: Disregarding or offending the reputation and dignity of the other person as a partner, as carer and educator of the couple's children or as a working professional.4.5.Manipulation of blame: Making the other person feel guilty for some attitude, behaviour or omission the perpetrator is accused of, and unilaterally blaming the victim for it.4.6.Disregard for the other person's emotions and ideas: Belittling or showing indifference to the other person's feelings, emotions or ideas, even going so far as to forbid their expression.Cognition 5. IMPOSITION OF ONE'S OWN THINKING: Discrediting the other person's ideas and imposing one's own point of view.5.1.Denigration of critical thinking: Discrediting and rejecting the other person's reasoning when it does not agree with that of the abuser.

Figure 1 .
Figure 1.Hierarchy of the categories and sub-categories of psychological abuse in intimate partner relationships.

Table 1 .
Taxonomy of the strategies of psychological abuse in intimate partner relationships.Seeking the maximum seclusion of the other person at home, even seeking to distance the other person from her social environment by forcing a change of address.2.CONTROL AND MANIPULATION OF INFORMATION:Filtering and manipulating the information that reaches the other person, forcing her to keep quiet about abuse and restricting any search for new information on the matter.2.1.Manipulation of information: Self-centredly manipulating information addressed to the other person, or withholding information as the abuser sees fit.
2.2.Concealing abuse:Forcing the other person to keep quiet about abuse and forbidding any attempt to seek outside help or information on the matter.3.CONTROL OF PERSONAL LIFE:Keeping the other person under surveillance and maintaining control over activities of everyday life, the children and shared belongings.3.1.Control over/abuse of finances:Taking unilateral decisions on the management and use of goods, money and debts in joint ownership, leading to serious excesses.3.2.Control over children:Unilaterally imposing decisions affecting the children's lives and development (education, nutrition, relationships) 3.3.Control over everyday activities and use of time: Scrutinizing the other person's activities and use of time, attempting to make the other person's life revolve as much as possible around the abuser.

Table 2 .
Descriptive statistics of the categories and sub-categories of psychological abuse in intimate partner relationships.