Risk factors in the work of the social worker: summary of a research on external user violence

In Italy, especially in recent years, the phenomena of violence against social workers by external users have assumed worrying dimensions, as in many other countries. Over time, numerous researches have shown that the phenomenon has become a real alarm in the professional community and that the wide spread reached by violence against this category of workers is connected with the levels of precariousness of social policies and the slow but constant weakening of the networks of social services supporting people in difficulty. Starting from these premises and from the more general assumption that at the centre of the professional action of the assistant there is the relationship with the user, a new research study was conducted with the aim of monitoring the phenomena of violence against social workers residing in the Region of Sicily, their incidence and types of consequences. The research, following a memorandum of understanding between the University of Enna 'Kore' and Croas (Regional Council of the Order of Social Workers of Sicily), was conducted by means of an anonymous information questionnaire submitted online to a population of 4736 social workers registered with the Order of Social Workers of Sicily. The data were collected in February and March 2023 and processed in the following month of April. 32.4%, that is, 1542 operators, of whom 941 registered in the A register section (61%) and 601 registered in the B register section (39%) participated by answering the questionnaire in full. Closed-ended questions were used, which are known to fairly facilitate the interpretation of participants' answers, on the basis of which a comparative analysis can be carried out and, where this was not the case, the data was modified and converted to the form and context necessary to allow for correct interpretation (number, scale, graph). The results showed that approximately 50 per cent of the respondents had experienced at least one incident of physical/verbal violence in the course of their work.

when the need is includable among those for which there are still resources available to be used to design change with the person (Campanini 2020).
In this way, however, the responsibilities transferred to the social service can be very onerous. The professional takes it upon himself to 'deny' rights, without the citizen being able to fully understand the real and serious contraction of resources made available by the State. Services and professions are affected by the risks of an even more general loss of meaning, due on the one hand to the occurrence of social phenomena and problems, to the meeting/clash between different cultures and ethnic groups, to the increasingly dramatic gap between variety, dynamism of needs and limitedness, difficulty of responses; on the other hand to the rise of myths, such as the race for ever more driven technologies and economic competition, which seem to create new and deeper problems (Sanfelici 2017).
It is often in services that social micro-conflicts, the incompatibilities of different interests and needs, the rights of individuals to protect and the defence needs of society pose dilemmas, the meaning and solution of which go far beyond the technical or organisational level. It seems that the professional, if left alone, cannot bear the weight and responsibility of difficult, delicate, risky choices. In the face of such social changes, the professional often finds himself the single, sometimes sole, interlocutor for the citizen in communicating to him the scarcity of available resources, exposed to continuous uncertainty and alone to the increasingly widespread social inequalities (Cortigiani and Marchetti 2020). At the centre of professional action is the relationship with the user, which, at times, can be charged with aggression. Pittaluga (2003) defines the professional social worker as a "trusted stranger"; the latter, in fact, is identified by the user as an external reference point to whom, for various reasons, the user comes to bring a need to which the stranger must respond. The basis of the help is in the relationship of trust that the professional will try to initiate. Trust between citizen and social worker is built slowly and is a ground to work on. It is a work of reciprocal openness that encourages a process of introspection and increases awareness of the risk and limit of expert knowledge; these risks are distributed to varying degrees among people and are strongly conditioned by economic factors, meaning that the user initiates a relationship with the professional that will be developed according to the models tested in the development of relational maturity (Biffi and Pasini 2022).
The social worker as a helping professional must therefore consolidate his or her competences with regard to some distinctive and peculiar elements of his or her profession. Aware that it is on those that the credibility of a profession under attack from so many points of view can be played out, but firm on the front of certain values and inspiring principles: dignity of the person, freedom to self-determine and to exercise indispensable human rights, equality among men, solidarity. The values of social service find concreteness in professional attitudes that sometimes already constitute components of the individual's personality, and which the professional is then socialised into during training, and then become an affective and motivational component in professional action over time. Among the main ones we can identify the attitude of acceptance of the person as a whole, the attitude of listening and confidentiality regarding what is shared within the helping relationship, the attitude of tolerance and flexibility to change and to the plurality of possible interlocutors. It is thus observed that 'help is not simply giving (Raineri and Corradini 2022). It is an exchange, a negotiation of views, goals, tasks. Greater possibilities for self-determination will be the outcome of a process done together. Social workers continue, therefore, to invest in empowerment, in the potential of individuals to activate their individual and network resources to get out of the state of need in which they may sometimes find themselves at some point in their existence (Balducci and Tre Re 2016). The social worker, therefore, adapts his or her actions to external goals, goals that are not his or her own, goals given by the very fact of exercising his or her professional activity. Neve (2008) argues that the work activity of the social worker is "something more and different from a simple job or employment". Dominelli (2015) understands power as a negotiated force of interaction, i.e. he recognises that in the relationship both parties in play have the potential to act a role of power and that no one can be without it. He attributes to the social worker the duty to consider himself in a position of dialectical parity and at the same time attributes to him the role of empowering the user to exercise his own capacity to act, despite his condition of discomfort, giving him greater potential for change.

The violent and aggressive component in human interaction: theoretical contributions
Simmel (2018) highlights the 'normality' character of conflictual relationships: the author defines these as forms of human association that 'are indistinguishable from other relationships'. Hostility and conflict, even to the extreme of violence, are to be considered as real social ties and forms of relationship, the more intense the greater the equality and intimacy between the parties. Elias (1988), in his work 'The Civilisation Process', analyses the instruments that lead to the control of violence in the modern age. Parallel to the political transformations that give rise to the establishment of the modern state, multiple changes take place from a psychological and social point of view, functional to the development of the control and emotional repression of human beings. According to Elias, with modernity, western nations begin a process of pacification, made possible by the 'civilisation of drives' into standardised and predictable, universally accepted behaviour. Violence is thus relegated to precise spatial and temporal enclaves, considered legitimate only if exercised by specific figures expressly authorised by state power -as in the case of the police -or if framed as a manifestation of the masses, in times of crisis or upheaval. (Farmer 2004: pp. 305-325), recalls the concept of 'structural violence' coined by the Norwegian sociologist Johan Galtung in the late 1960s, according to which there are three types of violence: a) direct or physical violence; b) cultural or symbolic violence; c) structural violence, the latter determined by social structures and the action of political and economic institutions that act on the lives of individuals. The concept was first taken up by Gilligan (1997), who described structural violence as 'the high rates of death and disability suffered by those who occupy the lowest strata of society, in contrast to the relatively lower rates of those above'. In turn, (Farmer 2004: pp. 305-325) reworks (Galtung's 1969: pp. 167-191) concepts through a new definition, according to which: structural violence indicates violence exercised systematically -that is, indirectly -by anyone belonging to a certain social order. In short, the concept of structural violence aims to inform the study of the social mechanisms of oppression. The latter, which appears to be 'a given', 'nobody's fault', acts on individuals in different ways, depending on their place in the social order. Farmer (2004) invites ethnographers to focus their attention on the mechanisms of structural violence and its material manifestations, i.e. on everything that is 'ethnographically visible. This approach makes it possible to grasp how inequalities are structured or historically legitimised by social and cultural processes, and how they are an expression of a particular political and economic order -as in the case of racism, poverty, the gender gap, sexism. Despite the criticism, the concept of structural violence reworked by Farmer has been widely appreciated, as it places the question of power and social inequalities at the centre of ethnographic research, with particular emphasis on the social inscription [of violence] in bodies. Scheper-Hughes and Bourgois (1992) propose to 'conceive of violence as operating along a continuum, from physical aggression to symbolic violence, to rutinary forms of everyday violence, including chronic, historically integrated structural violence whose visibility is obscured by cultural hegemonies'. According to American anthropologists, the interest in violence lies not only in its 'physicality', i.e. its material manifestations, but more properly in the social and cultural dimensions that give it power and meaning. The concept of the violence continuum aims to break down traditional distinctions between public and private, visible and invisible, legitimate and illegitimate forms of violence. Bourdieu et al. (1999) use the term 'symbolic violence' to define 'gentle and often invisible' violence, which is not recognised as such, but rather confused with something else. Understanding this peculiar form of domination requires overcoming the sharp distinction between compulsion and consent, between external imposition and internal impulse. Bourdieu et al. (1999) invite us to recognise the ways in which 'the dominated' often fully adhere to forms of domination, producing and reproducing them.
The topic of aggression implies a broad concept, developed under the lens of different disciplines. The literature on the subject of aggression offers numerous contributions from as many disciplines intent on studying the phenomenon of human violence. Gallino (2004) defines the term 'violence' as an 'extreme form of material aggression', intending to emphasise the extreme and concrete act of an individual to the detriment of others. It is, however, considered appropriate to start from the author's definition of 'aggression': "act, behaviour or action of an individual or collectivity consciously directed at harming, subduing, belittling, physically or psychically injuring another individual or collectivity in an arbitrary or illegitimate manner from the point of view of the victim and of the social system of which he or she is part". The term opens up to broad meanings in which the most varied acts of abuse by individuals or groups of individuals to the detriment of others or the contexts to which they belong can be included. It is therefore problematic to provide an unambiguous definition of aggression. Sanza (1999) defines aggression as a very complex phenomenon, in which biological, mental and, last but not least, environmental aspects intervene. It is even more difficult to attempt to draw a line between aggression and violence.
One could say that aggressive behaviour is acts intended to cause physical or moral harm to other subjects, while violent behaviour is acts deliberately intended to harm, physically or morally, other subjects. It is therefore possible to identify a differentiation between aggression and violence in the will of the perpetrator. Adhering to this view and wanting to refer to it to explain the aggression against operators working in the social services system, these facts could, therefore, be understood as the wide variety of acts carried out for the purpose of harm by those who interact with the same operators. There are various reasons that can lead individuals to the development of aggressive actions. According to Nardi and Recupero (1999), the theories that explain why certain people act or why violent acts are more likely to occur in certain circumstances than in others are of various matrixes and natures and concern: organicist theories, ethological theories, frustration/aggressiveness theories, social learning theories, sociological-environmental theories, psychoanalytically oriented theories, and finally the theories specifically concerned with people with 'borderline states. Organicist theories, thanks to neuroscience studies, have been able to locate aggression in the brain area of the limbic system, the fundamental seat of all emotions. Today, it is known that motor action and reactivity to the aggressive stimulus are controlled by neuromediators and hormones. Alcohol and psychotropic substances (psychiatric drugs and narcotics) interfere with the action of neurotransmitters, influencing, either by enhancing or depressing, the control over the aggressive stimulus. Lorenz (2015) distinguished between aggression directed at members of different species and intraspecific aggression, towards individuals of one's own species, with the animal fighting to hunt or to define its territory, respectively. He identified aggression expressly in the second of the two instincts as a tool for the preservation and organisation of the species in that it allows confrontation between members of the same species for the purpose of improving one's condition in the group to which one belongs.
Aggression is, therefore, considered by ethologists, for animal species and in the human species, an adaptive response and an ineradicable instinct that can be less harmful if redirected into ritual type behaviour that mitigates its destructive potential. The situations that can favour aggression have been identified by ethologists in the condition in which the subject experiences physical, psychic or social suffering. The individual in such a case will attempt to take immediate action to put an end to it, through defensive mechanisms, and will manifest needs including that of having containment, of being near loved ones and of being respected in their physical and psychic dignity. The frustration/aggressiveness theory (Dollard 2011), elaborated at Yale University since 1939, considers aggression a response that always develops in the face of the frustration of needs, expectations and expectations.
Thinking back to this eventuality in the social services, the condition of frustration of one's expectations is an eventuality that has many opportunities to occur if operators and those who work with them are not careful to avoid its occurrence. The Social Services act within social policies and local political dynamics that can often see them, not surprisingly, overloaded with potential and/or competences in the eyes of the citizens, without this being followed by the congruous and indispensable effective allocation of economic and personnel resources. The factors of anger and aggression from further frustration that falls back on previous conditions of serious suffering or lack must in no way be underestimated when in the encounter with the services people see their expectations first solicited and then disappointed. It is therefore very important to set up, from the very beginning of the relationship with the user, a clear and as transparent as possible communication on the actual opportunities that the service itself may or may not offer the user. The social learning theory (Bandura 1977) traces aggressive behaviour back to the implementation of models learnt through observation and imitation by others, which are considered all the more appreciable the more they allow advantages to be gained.
The sources that provide the behavioural models in the complex modern society can be the family, the social reference group and so on. Bandura (1977) emphasises the responsibilities of society in modifying people's living conditions, e.g. bureaucratisation, urbanisation, mechanisation, which, if taken to extremes, can lead the individual to internalise aggressive behaviour. Psychoanalytic theories are numerous and articulate and refer to the theoretical assumptions of numerous authors. They all start from the assumption that drive energy, the death instinct, if not adequately integrated with the vital energies of Eros, the forces of love, can generate destructive or self-destructive behaviour. There are also specific theoretical insights that deal with investigating 'borderline states' to understand what happens in the face of fragile individual personalities, in which there is a fragmentation of the inner personality structure and difficulties in maintaining the delicate balance of mental functioning. In particular, in the face of borderline personalities, social services should, first of all, work in an integrated and interdisciplinary manner, rather than arguing about mutual competences. Bronfenbrenner (1979) in particular studies the individual in interaction with his environment from which he receives important and continuous influences. He conceives the social context as a complex of four structures embedded in each other: the microsystem, the mesosystem, the exosystem and the macrosystem. The first level is that of the individual with his personal characteristics and relationships. The second is composed of several interacting microsystems. The area of the exosystem relates to the relationships between different organisational contexts in which individuals do not participate directly, but are affected. Finally, the macrosystem is understood as the superstructural context with influence on all the preceding ones, an example of which could be the complex of social policies. Fenoglio et. al. (2012) conduct an interesting reflection that should always guide the actions of those who work in social services: "knowing that one is in a network in an operational context capable of respecting and enhancing, centrality and importance of each one's function and role, being able to count on an organisation that is well anchored in the social context to which it belongs, which in turn is a network point aware of the strengths and weaknesses present in the territory and of the many important synergies that can be activated, is a further guarantee of the capacity to welcome and manage in a correct and respectful manner those situations of aggressiveness and high problematic nature that can sometimes be associated with conditions of lack and deep psycho-social distress". They emphasise the positivity of the helping relationship and valorise the strength of the social network as a possible protective context in which to distribute the inevitable contradictions and social conflicts that contemporary life contexts bring into people's lives, especially in the lives of fragile people.
A further theoretical contribution can be found in Sicora (2013), according to whom social workers do not have sufficient skills in their theoretical-practical background to be prepared to deal with the possible aggression of users. It also hypothesises that the practitioner seeks protection in the accumulated professional experience, in the organisation to which he belongs or in the professional community of reference, thus contributing to aggravate the ambivalence embodied in the institutional mandate that requires the practitioner to simultaneously perform functions of help and control. In such ambivalence, the helping relationship will easily be polluted by misunderstandings and prejudices, and the more general possibility of helping the person to find an answer to the need that led him or her to interact with the Service (at his or her own request or at the request of the Institution) may easily be compromised. The institution can be experienced by the user as the "monster", deaf and blind to people's needs, which has, among its various cogs, the operator, who participates in spreading similar coldness and indifference. The user who develops such a conviction should have the opportunity to establish a helping relationship in which the professional can clarify his or her professional mandate.
If such a clarification does not take place, the person may manifest his or her powerlessness to communicate his or her demands in the position of inferiority that they hold through aggressive behaviour. According to Kolnic-Acker (2008), it is therefore necessary for the practitioner to be aware of the inherent contradiction connected to his/her professional mandate and to act in unison with his/her organisation in order to be able to recognise ambiguous situations and to do everything possible to resolve them and prevent them from arising again. Studies, research and victim testimonies have shown that various violence and assaults against social workers tend to take place within one of two possible contexts: the home visit and the institutional setting. The home visit constitutes a context of greater exposure to the risk of aggression. The home context represents a particular setting for conducting a professional interview, even if the practitioner does not go there to carry out control functions, but as in this case pursues social development aims. He/she may find hostility and distrust, in this case even expulsion on the part of the user who, unaware of his/her discomfort, intends to hide it and deny him/herself (Bini and Peruzzi 2016). In spite of the competence in putting in place behaviours to prevent the risk of aggression, there are still unforeseen events that could not be foreseen or controlled, such as, for example, the entry of other people that the social worker was certain were not present. Perhaps it is precisely because of this unpredictability that this context remains an indispensable place for the professional social worker, because it is often only in this context that it is possible to grasp fundamental elements for evaluation and for reading aspects of family dynamics that the user does not reveal elsewhere. Quite different, on the other hand, are the aggressive and violent acts in the institutional offices of the social services: the violent and aggressive type is directed both at the operator and at the structure itself, through the destruction of office furniture, computer equipment, and so on (Giribaldi 2011).

Legal Profiles of Countering Violence in the Workplace
In order to effectively combat violent and aggressive actions in the helping professions, Italy has enacted Law no. 113 of 14 August 2020 (Provisions on safety for health and social-health professions in the exercise of their functions), which provides for administrative and criminal sanctions. The repressive intent of the Italian legislator is expressed in particular in Articles 4, 5, 6 and 9. Article 4 of the law intervenes on Article 583quater of the Italian Criminal Code, which is extended and supplemented by including personnel exercising a health or social-health profession (including auxiliary activities) on a par with a public official; serious personal injuries are punished with a penalty ranging from 4 to 10 years in prison. Very serious injuries are punished with 8 to 16 years' imprisonment.
The repressive instrument is strengthened by Article 5: among the aggravating circumstances of the offence is the new paragraph 11-octies of Article 61 of the criminal code (having acted with violence or threat). Article 6 provides for ex officio prosecution (without a complaint by the offended person) or the obligation to prosecute acts of physical violence against health and social care personnel. In the event that the act committed does not fall under the offence of injury, threat, harassment or similar, Article 9 provides for an administrative sanction of between EUR 500 and EUR 5,000 for violent, insulting, offensive or harassing conduct acted against healthcare or social-healthcare professional staff, as well as anyone who performs auxiliary care, healthcare assistance or rescue activities (Official Gazette of the Italian Republic No. 224 of 2020). In Italy, given the increase in cases of violence or attempted violence, as well as assaults, against social workers, the National Order of Social Workers has drawn up a Vademecum to prevent and address the risk of violence against aid professionals. It is recognised that violence and harassment in the world of work are unacceptable and incompatible with decent work. The main new features of the Convention include the definition of violence and harassment in the world of work, the identification of protected persons, the specification of work areas of application and the identification of preventive and countermeasures to be taken, as well as the persons responsible for their implementation. The Convention requires States to take an inclusive, integrated and gender-sensitive approach to the prevention and elimination of violence and harassment in the world of work, according to the different roles and functions played by governments, employers, workers and their organisations, taking into account the diversity of the nature and scope of their respective responsibilities. The Convention and Recommendation have three main pillars: 1. protection and prevention; 2. enforcement and redress mechanisms; 3. guidance, training and awareness-raising. .

Methodology
The statistical survey was conducted by means of an anonymous questionnaire submitted in February and March 2023 in online mode through the official website of the CROAS (Regional Council of the Order of Social Workers of Sicily) to a population of 4736 social workers; of these, 1542 (32.4%) individuals participated by answering the questionnaire. Questionnaires, as is well known, can be used to collect routine or infrequent data and for specialised studies.
The response rate (i.e. the percentage of selected individuals completing the survey) is a key parameter and helps to understand the validity of the survey and the sources of 'nonresponse' error; to maximise response rates, questionnaires are designed to be as simple and clear as possible (Edwards and Cantor, 1991;Sudman et al., 2000: pp. 211-236) with targeted sections and questions. The statistical analysis was therefore based on a sample of 1542 social workers, of whom 941 were registered in Register A (61%) and 601 in Register B (39%).
Closed-ended questions were used in the questionnaire, which give the respondents' answers a fair interpretation, on the basis of which a comparative analysis can be carried out and, where this was not the case, the data was modified and converted to the form and context necessary to allow for correct interpretation (number, scale, graph). The focus of the statistical analysis in this research was to observe the phenomenon of violence against social workers in the course of their work. In the first stage of descriptive analysis also called 'preprocessing' the raw data was cleaned and organised for the next stage of the processing.
During preparation, the raw data was rigorously checked for errors. The purpose of this phase was in fact to eliminate bad data (redundant, incomplete or incorrect) and to start selecting quality data, also eliminating questionnaire variables/information not of interest for the research under investigation. In the second phase, the data were translated into the language comprehensible to the systems that were used for the purposes of statistical analysis (R-Excel), creating a database with common values and information for the individual variables and calculating some position indices (mean and median, extreme sup. and inf.) and decoding useful for the synthesis of some information concerning the statistical distribution of the original data. In the third phase, the data were actually processed for interpretation, providing results that were then transformed into user-usable information through the use of graphical representations (pie, histogram, funnel) and tables. In the last phase of the analysis, the correlation between some main variables and the frequency distribution of the phenomenon under investigation was studied. Correlation is a statistical index, i.e. a number, which varies between -1 and +1.
The closer its value is to its extremes, the stronger the correlation between two variables will be, and if the values of one variable are directly associated with the values of the other variable, there is a positive correlation; if the x and y values of two variables are in inverse relationship then the correlation is negative; when there is no relationship between the values of the two variables, the correlation is zero. Once the main variables with a significant correlation between them had been identified, a regression between them was carried out to further the research at an inferential level because, while correlation is a useful tool for understanding the relationship between two quantities, regression also helps to understand how one quantity is affected by another. In correlation analysis, the two quantities are considered symmetrically, whereas in regression analysis, one is assumed to depend on the other, non-symmetrically (New Palgrave: A Dictionary of Economics "Regression and correlation analysis", Lindley 1987: pp. 120-123).
In particular, the generalised linear regression model (GLM) used for the case study at hand is the Logit model; this is a logistic/statistical model that configures the probability of an event occurring by making the log-odds for the event a linear combination of one or more independent variables. In regression analysis, logistic regression consists of estimating the parameters of a logistic model (the coefficients of the linear combination). Formally, in binary logistic regression there is a single binary dependent variable, coded by an indicator variable, in which the two values are labelled '0' and '1', while the independent variables can each be a binary variable (two classes, coded by an indicator variable) or a continuous variable (any real value). The corresponding probability of the value labelled '1' can vary between 0 and 1; the function that converts log-odds into probabilities is precisely the logistic function, hence the name (Generalized Linear Models; McCullagh and Nelder 1989). As far as the frequency distribution is concerned, this indicates the way in which the different modes of a character are distributed in the statistical units that make up the collective under study, and the analysis, also inferential here, involved the comparison of a distribution adapted using Kernel Density Estimation (Annals of Statistics "Kernel density estimation via diffusion"; Botev, Grotowski and Kroese 2010: pp. 2916 -2957) and a lognormal distribution with mean and variance parameters extrapolated from the data sampled in the questionnaire.

Results
In order to analyse the phenomenon of violence against social workers, it was appropriate to start by processing a series of significant data from the questionnaire under review and appropriate statistical indices. The data processing was carried out in April 2023. The age range of the sample surveyed is between 21 and 70 years; the average age of the respondents calculated as the weighted average with coefficient (weight) of each age group is 46.5 years; the median is 48.5 years. The sample analysed is fairly gender unequal: only 5.7% of the respondents are male.
The steps followed for the purpose of the analysis are listed below in bullet points: -respondents who have experienced physical/verbal violence themselves in the course of their work; -respondents divided by age group who have experienced physical/verbal violence in the workplace; -respondents broken down by experience in the field who have reported experiencing violence; -respondents who are aware of the legal information and safety provisions for health professionals; -respondents who are victims of violence who are aware of law no. 113 but have not reported the phenomenon; -frequency of psychological problems and frequency of the phenomenon 'violence' related to the interpretation of aggressive behaviour; -correlation analysis, regression and study of the distribution of the phenomenon 'number of violence suffered'.
As a first step, the sample was divided among the interviewed staff who had experienced physical/verbal violence themselves in the course of their work independently of working hours.
As depicted in Figure 1, 47% of the interviewed staff had experienced violence. Thus, 721 social workers out of the total sample analysed (1542) were victims of violence, almost one in every two respondents.
When the sample is standardised by age, it is found that there are: no.23 respondents with an average age of 24 who have experienced violence (28.75%); no.83 respondents with an average age of 30.5 who have experienced violence (36.8%); no.66 respondents with an average age of 36.5 who have experienced violence (34%); no.71 respondents with an average age of 42. 5 who had experienced violence (43.8%); no.122 respondents with an average age of 48.5 who had experienced violence (50.8%); no.173 respondents with an average age of 54.5 who had experienced violence (54.5%); no.154 respondents with an average age of 60.5 who had experienced violence (57.9%); no.28 respondents with an average age of 66.5 who had experienced violence (50%); no. 2 respondents with an average age of 70 years who have experienced violence (50%).

53% 47%
Not suffered violence Suffered violence  Figure 2 shows the frequency in percentage order of the phenomenon of violence suffered in the course of work against the interviewed staff, broken down by age group. The age group most affected is the 53-63 age group, in fact, 57.9 per cent of respondents in this age group confirmed having experienced at least one episode of physical/verbal violence in the course of their work.

Figure 2. Frequency of violence suffered in the sector by age group
The phenomenon of violence occurs several times during the work experience, the same respondent, in fact, confirms having experienced physical/verbal violence several times during work. Therefore, in the third step the number of violent incidents in the course of work experience and the reporting of this phenomenon to the judicial authorities was investigated. The results, which can be seen in Figure 3, show how, in the face of a substantial number of violent incidents, the propensity of the respondents is not to report the phenomenon. The sample shows: approximately 168 violent incidents suffered by respondents with experience in the sector between 0-2 years and only 33 reports of the phenomenon; 271 violent incidents suffered by those with experience in the sector between 3-6 years and only 12 reports of the latter; 333 violent incidents suffered by those with experience in the sector between 7-10 years and 35 reports of the latter; 136 violent incidents suffered by those with experience in the sector between 11-14 years and only 8 reports of  Age Range  INF  SUP  Tot individuals N_Violence  21-27 years  21  27  80  23  28-33 years  28  33  225  83  34-39 years  34  39  194  66  40-45 years  40  45  162  71  46-51 years  46  51  240  122  52-57 years  52  57  317  173  58-63 years  58  63  266 154 64-69 years 64 69 56 28 > 70 years 2 1 the latter; 134 violence suffered by those with experience in the field between 15-18 years and 10 complaints; 69 violence suffered by those with experience in the field between 19-22 years and 7 complaints; 125 violence suffered by those with experience in the field between 23-26 years and 4 complaints; and 674 violence suffered by staff with experience in the field for more than 26 years. The 'low' number of complaints could be due to the unfamiliarity of the legal information that guarantees protection for the interviewed staff in the exercise of their profession, therefore, as a fourth step, the interviewed sample was analysed on the basis of 2 main parameters ( Figure 4): a) knowledge and reading of the legal information Vademecum and therefore the form to report violence and assaults suffered and/or assisted; b) knowledge of law no. 113 of 14 August 2020 entitled 'Provisions on safety for health and social health professions in the exercise of their functions'. From this analysis it emerges that, while 75% of respondents are aware of the existence of the Legal Information Vademecum, less than 40% of staff have actually read it and are aware of the laws that protect them. Moreover, less than 50% of the sample interviewed are informed about Law No. 113, which provides the provisions for the protection and prevention of the phenomenon; therefore, they have no knowledge of how to safeguard their person in their work.  As a fifth step, the interviewees who were victims of violence and who were aware of law no. 113 for the protection of the profession decided to report the phenomenon were analysed. The results, as visible in Figure 5, show that of 721 respondents, who experienced physical/verbal violence themselves, only 319 respondents were aware of the law to protect their interests and of these almost half (137) chose to report the phenomenon to the judicial authorities. The reason why, despite having experienced physical violence and being aware of law no. 113, the interviewed staff still chose not to report the phenomenon, should be sought in the interviewees' individual interpretation of the concept of violence. As a sixth step the sample was analysed according to the interpretation of the phenomenon 'violence', and from this analysis, carried out on the whole sample, the following emerges ( Figure 6). -58%, of the respondents, while not justifying the phenomenon, would not take legal action; -24% of respondents would inform their employer, reserving the right to take legal action; -16% of the respondents would justify the perpetrator as being in a state of need/disorder; -3% of the respondents, on the other hand, would do nothing, because they find no protection from the employer/regulation. As can be seen from Figure no. 7: -75% of those who state that they do not feel sufficient protection from their employer turn out to be later victims of violence and, at the same time, they are also those who present more psychological problems following the event; -following this, the cluster most affected by both violence (61%) and psychological repercussions is that of those who have a general tendency to justify the aggression, as they state that the perpetrator is in a disturbed condition; -the 45% of the cluster most present in the sample surveyed, i.e. the cluster of respondents who, while not justifying the phenomenon, would not take legal action, are victims of aggression and, in 21% of cases, undergo psychological treatment following the event; 24% 58% 16% 3% I do not justify and proceed/proceed to inform the employer, reserving the right to take legal action I think that no form of violence can be acted upon regardless of the person's condition I justify the person as being in a state of need/disturbance I would do nothing because I find no protection from my employer 0% 20% 40% 60% 80%

Victims frequency
Psychological issues frequency I would do nothing because I find no protection from my employer I justify the person as being in a state of need/disturbance -the cluster least affected by both violence (37%) and psychological consequences (21%) is that of those who would proceed to inform their employer, reserving the right to take legal action.The last step of the analysis concerns the study of the correlation between the main quantitative variables acquired from the questionnaire that explain the phenomenon of violence and its consequences. In statistics, correlation is a relationship between two variables such that each value of the former corresponds to a value of the latter, following a certain regularity. Correlation does not depend on a cause-effect relationship, but on the tendency of one variable to change as a function of another and is calculated as the ratio of the covariance of the two variables to the product of the two individual variances. In formulae: = The variance of a variable "X" is defined as: = ' *( − ) 2 "N" being the sample number and μ the arithmetic mean. The covariance between two variables X and Y is instead calculated as: The results of the analysis (Figure no. 8) show the correlations between the following variables: violence suffered (Victim), number of violence suffered (N_Violence); event occurred during working hours (Working Hours); psychological repercussions due to the phenomenon (Psychology); change of working style due to the phenomenon (Change_Style); reported event (Complaint); knowledge of law no. 113 which provides the provisions for the protection and prevention of the phenomenon (Law) and years of experience in the field (Experience).

Figure no. 8. Correlation of main variables
The main results obtained show that: (a) there is a high positive correlation between psychological problems and the number of violent incidents (ρ = 0.4) and therefore, as one would expect, the greater the number of violent incidents the more likely one is to experience psychological repercussions; b) there is a high positive correlation both between change in work style and being or not being a victim, and between change in work style and psychological problems (ρ = 0.4), and thus the more violence or psychological repercussions an individual experiences the more he or she changes his or her work style for subsequent patients; (c) there is a moderate positive correlation between reporting and psychological problems (ρ = 0.3), and thus if individuals experience repercussions as a result of the act they have suffered, the more often they report. In addition to measuring the intensity of the link between the aforementioned variables, we are also interested in ascertaining how one of them (dependent variable) varies with the other (independent variable(s)), by finding an appropriate analytical function that summarises this relationship (GLM). This inferential analysis takes the name of generalised linear regression, as the object of study will be a dependent variable (Change in work style) and two explanatory variables (Victim and Psychology) in binary format (0;1). The GLM used for our analysis is presented as an expansion of the linear equation, in formulae: = + + + where: y: log (odds) of the dependent variable (Style Change); β_0: intercept of y, i.e. the value of y when x_1, x_2 are 0. β_1, β_2: regression coefficients representing the change of y with respect to the change of one unit of x_1, x_2; x_1, x_2: predictor variables (Victim and Psychology); ε: outlier component  The significance level α selected is 0.05. Therefore, the correlation coefficient is considered statistically significant when p-value < 0.05 and the linear regression produced has p-values of less than this threshold so that all parameters are statistically significant, in formulae the equation: y= -2,8411 + 1,6743 +1,2532 The results of the equation show that: (a) if β_1= 0 (the social worker has never been a victim of violence) and β_2 = 0 (the social worker has never experienced psychological repercussions in the course of work Beta Estimate p-value Significance -2,8411 <2e-16 *** 1,6743 <2e-16 *** 1,2532 <2e-16 *** performance), then: y = -2.8411; this value shows that the probability of such an individual to change work style is 5%. b) if β_1= 1 and if β_2= 1 the social worker experienced both violence in the course of work and psychological repercussions, then: y = 0.08636; this value shows that the probability of such an individual to change work style is 52%. Table 3 shows the results of the GLM model in the 4 different case studies. Finally, an analysis of the distribution of the number of violent incidents was carried out, depicting in the same graph (Figure 9): the histogram obtained from the sample data; a continuous probability distribution fictitious (blue line); a log-normal distribution with mean (µ) and standard deviation (σ) equal to the sample data (red line).

Figure no. 9 Number of Violences: Fit Histogram vs Log-Normal
As can be seen from the graph, the Log-Normal distribution fits the sample probability distribution of the number of violent incidents experienced by a social worker in the course of employment fairly well.

Conclusions
The phenomenon of violence against social workers seems to be very frequent and of particular concern. The aim of this research work was to show that in the course of their work about 50% of the respondents have experienced at least one episode of physical/verbal violence. Consistent with the expectations, the results show that as the years of experience in the sector increase, this phenomenon increases and the episodes of violence are not single but add up to dozens of incidents for a single person, just think of the total number of episodes of violence reported by those with more than 26 years of experience in the sector (674). The phenomenon of violence causes various psychological repercussions for those practising the profession and, as seen in the regression study, about one in two respondents would change their working style (52%) and thus change the way they practise their profession towards their next patients. Respondents do not always confirm that they are aware of the legal provisions protecting their profession (less than 50% of the social workers interviewed are aware of them) and in this respect they do not feel sufficiently protected in their profession; more often than not, the latter are the ones who suffer the greatest psychological repercussions by not externalising the incident. The overall number of reports of the incident seems to be ephemeral over the total number of incidents reported by the interviewees, in fact, out of a total sum of 1775 incidents reported by the interviewed staff only 135 times the incident was reported, therefore the judicial authorities are only aware of about 7.5% of the incidents that actually occurred.
The results suggest that it might be useful to pay more attention to this phenomenon, which is so widespread in this sector, and to try to limit it with more information from a legal point of view in order to be able to protect the staff concerned. This study could be a starting point to make staff aware of the incident and try to limit its occurrence. Future research should monitor, also on a national scale and with more significant samples, the trend of violence against social workers in the context of their work.