Network intervention and social workers’ personal support networks: a survey in the Italian context

Luigi Tronca

Università degli Studi di Verona, Italy

Sandro Stanzani

Università degli Studi di Verona, Italy

CORRESPONDENCE:

Luigi Tronca

e-mail: luigi.tronca@univr.it

Abstract

This article presents the results of a study conducted in Veneto, an Italian region, on a sample of social workers. The research examines the link between the characteristics of social workers’ everyday personal support networks and their propensity to use the network intervention methodology in their professional practice. The network intervention methodology is a set of techniques that enable social workers to help people taken into care by social services by activating and/or reactivating their daily support networks. Therefore, this paper seeks to understand whether the characteristics of the daily support networks that social workers personally have affect the trust they place in the networks of the people they assist in overcoming their situations of need. The research used Personal Network Analysis to identify the characteristics of social workers’ personal support networks and a series of questions to record their propensity to use the network intervention methodology in their professional practice. The survey shows that about one-third of social workers place network intervention among their most frequently performed activities. Positive links are found between the size of social workers’ personal support networks and the importance of network intervention, and between the ego-centric density and the tendency to adopt a bonding approach to network intervention.

Keywords

Social work, Network intervention, Personal Network Analysis, Personal support networks.

Introduction

Network intervention is a social service methodology that emerged in Canada in the second half of the 1900s. It is a helping relationship that draws on social relationships as resources for supporting people’s well-being. The social work professional is responsible for guiding the social work recipient within his or her relationship networks so that these can provide the support of which they are capable (Valente, 2012). Such relational guidance by the social worker can activate latent nodes (Robins et al., 2023) and promote the generation of new relational contexts. This methodology requires social workers’ mastery of relational concepts and tools, and there is now an articulate literature on this subject. The work we are proposing questions the degree to which the methodology of network intervention is widespread among social workers registered in Veneto, a region of Italy, to understand how relationships between facilitators, participants, and/or other stakeholders can impact this form of intervention.

In this regard, we conducted an exploratory survey of social workers in a northern Italian region (Veneto). In addition, by analyzing respondents’ opinions, we sought to observe how network intervention skills combine with other professional skills of the social worker and to identify which professional culture best accommodates the methodological skills of network intervention. Network intervention, as will be seen in the next section, is a specific social work methodology. Finally, the work set a specific research question to be tested. Using social network analysis tools, an attempt was made to determine whether a link exists between the structure of personal and professional support networks and the social worker’s attitudes toward network intervention methodologies, and whether they favor their adoption in professional practice.

Theoretical background

The definition of social work and the methods and techniques used for social intervention have evolved in correspondence with changes in social policies and cultural processes concerning the semantics of well-being.

For much of the 20th century, in the modern West, subjective well-being was linked to the individual’s ability to cope with environmental challenges and achieve their goals (Parsons, 1972). Consequently, the social worker’s intervention has been interpreted as a therapeutic intervention, centered on individual functioning. According to this perspective, the professional’s action aims to unlock the individual and foster his or her coping skills. The social worker must provide the individual with the skills to read reality, identify the resources to be mobilized to cope with environmental challenges, and thus carry out their tasks and achieve their goals.

The intervention methodology is called case work (Richmond, 1922; Perlman, 1957). In this case, the social worker operates according to a clinical-therapeutic approach (Lasala, 2022), he/she works on the case and implements predominantly helping actions rather than helping relationships, since he/she activates towards the case in order to transfer resources to it unidirectionally: material resources, such as economic subsidies or rights to access services, but also strategies with well-defined tasks and objectives, on which the person must focus his or her action and undertake a pathway out of the problem (Reid & Epstein, 1972).

In summary, the professional’s helping action fosters the person’s empowerment and facilitates coping capacity. During the 1900s, a systemic-collective approach was used alongside the individual clinical approach. Social workers continue to observe the case from an individual perspective but place the case in the social context and are responsible for coordinating access to the service system. There is a shift in perspective from case work to case management. The case manager is asked to observe the service user in the relevant social context and manage his or her access to complex resources and services to address multiple needs.

This holistic view of social work gave rise to two variants: a more distinctly technical-managerial nature and a critical-emancipatory nature. The latter had a particular development around the 1960s-70s, coinciding with the youth protest season. The approach aims to promote subjective well-being by observing the person in his or her social environment and trying to identify and eliminate possible structural causes of lack of well-being (Moreau, 1979). It is a more political approach that sees social workers in an advocacy role vis-à-vis political and economic institutions. According to this approach, the social worker continues to observe the case from an individual perspective and manage access to resources and services. However, the social worker is also charged with unraveling and erecting a bank in the face of sociocultural structures that place the individual at a disadvantage, thus mitigating the effect of discrimination acting at the cultural, structural, and systemic levels, generating individual distress. Some observe a shift from individual blame to system blame (Thompson, 1992), which aligns with the shift from case work to case management perspectives.

The 1960s-70s, an era of post-industrialism and major youth protests, also saw cultural processes at work that changed the welfare culture and further influenced social work methodology. There is a move away from rationalistic and individualistic, task-oriented conceptions of welfare. Conceptions of well-being become more complex, the importance of biopsychic aspects is emphasized, and emotional and relational components are valued, centered on the sensitivity and reflection of the individual. What matters is the search for an adequate relationship with oneself with a view to the realization of one’s authenticity.

Having abandoned the behaviorist and cognitivist psychological perspective, social work also makes way for other psychological currents, from psychoanalysis to the relational systemic approach (Watzlawick & Weakland, 1977) to the Rogerian approach (Rogers, 1951). It loses relevance to the task-centered approach and, with the rise of the Palo Alto school and systemic therapy, the idea emerges that treating social distress must involve the network of relationships in which the subject with problems is involved. It is no longer a single subject that has coping problems. However, the entire network of his or her daily life relationships simultaneously generates and causes distress. The same network can also be a resource for fostering well-being and working to treat subjective distress.

These are years in which social movements are lobbying for the abolition of total institutions of care that generate discrimination and distress. In such a cultural climate, social work methodology opens a robust reflection that continues to shift the focus from case work to case management and community case management (Cooper, 1989; Guay, 1987; Payne, 1986) and eventually to care management (Challis & Davies, 1986), developing the methodological tools for network intervention (Maguire, 1983) and community care (Bulmer, 1987). It is no accident that the conception and first experimentation of network intervention occurred with Brodeur and Rousseau (1984) in a psychiatric institution in Quebec. That experience symbolically marked the shift from the institutional systemic realm of care to that of primary social relations networks or the extended community. It contributed to a reciprocal revision of social intervention. A well-established conception of social work as an action of help by the social worker who, as helper, provides resources and services to enable the help recipient to accomplish the task was attenuated. A reciprocal relationship perspective, in which the social worker who, as a helper, pays attention to the contribution that the person receiving help can bring to the helping process has increased in the methodology of helping (Carkhuff, 1987).

Some authors have emphasized that the change process is not only the result of the resources the person receiving help may have, but also of making their motivation effective and achieving their goals. There has been a strong insistence on relational aspects, and some believe that motivation for change is an «interpersonal process» (Miller & Rollnick, 2002). Thus, in addition to the case management of the clinical approach and the political pressure and advocacy actions suggested by the critical approach, a methodological perspective that promotes network and community intervention has gained ground, leading some to talk about «relational social work» methodology (Folgheraiter, 2004; Folgheraiter & Raineri, 2012).

In this paper, we intend to study further the conditions that can foster social workers’ adoption of network intervention (Besson, 1995; Brodeur & Rousseau, 1984; Maguire, 1983; Sanicola, 1996).

Among the strengths of network intervention are the tendency, after getting to know their networks, to empower help recipients (Maguire, 1983; Seed, 1990), which can imply saving economic resources necessary to deliver services to the person, and an anti-oppressive orientation (Dominelli, 2002) in delivering supportive interventions. As Sanicola (1996) reminds us, the professional who practices network intervention is in the position of participating in the primary network — that is, the network of family relationships, friendships, etc. — of the person assisted, being at the same time placed in the secondary network, consisting of the professionals who must help him or her. The social worker who uses network intervention fosters the taking charge of the person assisted, placing him or her in a helping project that makes him or her essentially autonomous from the secondary network (Sanicola, 1996). «This is not just a technique, but a strategy for action that goes beyond the logic of state intervention, where services are activated based on a right. This strategy also goes beyond the logic of the market, which offers goods and services in return for payment. The network worker is not seen as someone who must act «in the place of», nor as a «therapist» or «activist», but rather as a support for reflection on collective problems and a stimulus for a movement of transformation within the network» (translated from Sanicola, 1996, p. 51).

In summary, we can say that network intervention consists, on the part of the social worker, of reactivating existing ties and creating new ones to mobilize some social support toward the person being cared for. Social support is closely related to well-being and health (Meo, 1999; Song et al., 2011). The purpose of network intervention is to sustainably improve the well-being of the assisted person by fostering the structuring of an enduring network of support around them. As Song et al. (2011, p. 118) note, social support is «the aid — the supply of tangible or intangible resources —individuals gain from their network members». Among the network actions directly related to strengthening social support towards the help recipient, we can find the reactivation of latent support ties, alleviating the caregiver’s support burden, encouraging its redistribution within the network, and promoting self- and mutual aid groups (Sanicola, 2009). The focus on social support is of particular importance in the Italian context, which is characterized, as Sanicola (2009) notes, by the following relational aspects: «a persistence of primary solidarity, with a significant permanence of the family «node»; the emergence of the phenomenon of secondary solidarity, which manifests itself through the spread of associations and cooperatives; a pervasive system of sectoral services, which gives an institutional character to our country’s welfare mode» (translated from Sanicola 2009, p. 50).

We would like to clarify that, in this article, we use the term network intervention to refer to one of the intervention methods available to social workers and generally covered in their training. The research team did not provide any training on network intervention to the social workers involved in this study. The option of not answering questions was always guaranteed, so social workers unfamiliar with the social intervention methodology known as «network intervention» could avoid answering questions about it.

Method

The research from which we draw the data presented here is entitled «The relevance of social networks in the social work profession in Veneto» and was carried out by Sandro Stanzani and Luigi Tronca of the Department of Human Sciences at the University of Verona. This research was carried out in collaboration with the Order of Social Workers of Veneto (Italy). This eminently exploratory survey provided, among other things, information on the relationships between existing collaboration networks among professionals and, where applicable, on how they apply network intervention methodology. The survey was conducted on a sample of individuals registered with the Order of Social Workers of Veneto.

The reference population consisted of 2,970 individuals as of October 22, 2019, i.e., before data collection began. The sample was constructed, and data were collected by the research company Demetra opinioni.net s.r.l., which is based in Venice-Mestre (Italy)¹. This research company was responsible for:

• extracting cases from the list of members provided by the Order of Social Workers of Veneto
• administering a questionnaire online using the CAWI (Computer Assisted Web Interviewing) methodology.

The research company proceeded by randomly extracting a sample of 1,000 cases, proportionally stratified according to the gender of the population (females = 93.0%; males = 7.0%). Data collection began on October 24, 2019. Each selected subject was sent a dedicated link to complete the questionnaire, with a deadline of November 15, 2019.

Given the low number of responses — only 239 complete interviews had been received by the planned closing date — it was decided to reissue the questionnaire and extend the deadline to November 30, 2019. By the closing date for data collection, 536 interviews had been completed: females = 93.1%; males = 5.4%; no response = 1.5%². The average age of the respondents is 42.17 years (SD = 10.78, min = 23, max = 72, n = 510), and there are no respondents who are not working. All respondents have a university degree in social work. Social workers account for 92.7% of the sample, with the remaining 7.3% engaged in social welfare or educational work (n = 534). Because this paper is particularly interested in learning more about the conditions that can foster network intervention among social workers, from now on, we will consider exclusively those among the respondents who work as social workers, that is, the 92.7% we mentioned just above, i.e., 495 respondents.

As is well known, social support can be effectively investigated using the research tools of Social Network Analysis (Fernández-Peña et al., 2022; Song et al., 2011). In Italy, there is a well-established tradition of using Personal Network Analysis (Bidart & Degenne, 2005; Di Nicola et al., 2011a; Tronca, 2012; Wellman, 2007a, 2007b;) to study social support in national samples of the resident population, with face-to-face administration of the questionnaire (Di Nicola et al., 2011b) or through the CAWI strategy (Secondulfo et al., 2017; Tronca, 2019). It was therefore decided to adopt Personal (or Ego-centric) Network Analysis in this survey to study the respondents’ social support. In Personal Network Analysis, the personal networks of single subjects are reconstructed and analyzed.

In line with national research conducted in Italy in 2016 (Secondulfo et al., 2017) and in 2018 (Tronca, 2019) by the Household consumption monitoring centre of the Department of Human Sciences at the University of Verona, a name generator (Bidart & Charbonneau, 2011) was used to identify the personal support networks of the individuals interviewed. This allowed the production of a list of names (alters) for each respondent (ego). As in the case of the Household consumption monitoring centre’s national surveys, the maximum number of people who could be indicated was six. This choice aligns with what was estimated from a previous survey, also conducted in Italy, but without a maximum limit on the number of indications, because it was conducted face to face (Di Nicola et al., 2011b)³. For each of the alters indicated, attributive information was then collected using three name interpreters for gender, age, and educational qualification. This information made it possible not only to analyze the composition of social workers’ support networks but also to study their levels of homophily⁴. An additional name interpreter was used to identify the social circle to which the alters belonged (Simmel, 1908). Finally, to investigate the morphological characteristics of these personal support networks, a name interrelator (Tronca, 2013) was administered to those who had indicated at least one alter, which made it possible to know whether the ego reciprocated the potential support and whether there were direct links of potential support between the alters. Ultimately, about support ties, this survey made it possible to reconstruct a direct graph for each respondent. We would like to clarify that the reference to these surveys previously conducted on samples of the general population residing in Italy — and, therefore, not on the population of social workers — serves solely as background information to justify the decision to set the maximum number of alters each interviewee could indicate at six.

In our analysis, we will first present the personal and professional profile variables for social workers. We will then present the characteristics of personal support networks. We will then outline the methodology for network intervention, beginning with its culture and structure. Finally, we will discuss the link between social workers’ everyday personal support networks and their propensity to use the network intervention methodology.

In addition to presenting univariate and bivariate analyses, the article also proposes multivariate analyses. Principal component analysis will be used to investigate the professional cultures of social workers, based on the importance attributed to different professional skills and the frequency with which these skills are used in the profession. We will therefore examine how skills in network intervention methodology match the others considered in the research. Multiple linear regression models will be used to determine whether and to what extent variables describing social workers’ personal support networks influence, net of certain personal characteristics and aspects of their work, their propensity to use the network intervention methodology, and which structural characteristics they recognize as most important for this strategy for working with people taken into care by social services.

Results

Personal and professional profile variables

As anticipated, the analyses we will conduct will only cover respondents in the social work profession (n = 495). 94.5% are female (n = 487). The average age of the respondents is 42.90 years (SD = 10.51, min = 24, max = 72, n = 470). The average length of work experience is 16.97 years (SD = 11.00, min = 0, max = 42, n = 474)⁵. All respondents except one have Italian citizenship (n = 488). The vast majority of social workers, 74.7%, say they work as social workers in the field, with direct interventions on users (n = 478). In contrast, the others work with other functions (coordinators of practitioners or services, etc.).

Respondents could indicate more than one issue when asked about the social field in which they practice their profession. The answers provided by individual respondents should, of course, be interpreted as their perceptions, not as objective information about their work. Table 1 presents the valid percentages of indications concerning each intervention area the research considers. The three main focus areas where respondents work are older adults, family, and adults in need, with more than a quarter of respondents engaged. This is followed by maternity and childhood, disability, youth, immigration, mental health, women who have experienced abuse, and addiction, with values between about 10% and 20%.

Table 1

Different social fields of intervention (n = 492)

	%
Older adults	45.5
Family	42.7
Adults in distress	32.7
Maternity, childhood	21.7
Disability	21.1
Youth	13.8
Immigration	11.8
Mental health	11.4
Women who have experienced abuse	10.8
Addiction area	10.2
Employment integration	8.3
Adult justice	5.3
Juvenile criminal	2.2
Hospital social work	1.6
Other	5.5

Three independent variables will serve as a reference point for subsequent analyses. First, gender differences are considered, even though in Italy, and in the sample of this research, the vast majority of social workers are female. However, we want to examine whether gender influences how personal and professional support networks are structured, as we believe it is worth verifying whether there is a different way to consider and implement network interventions. Years of experience as a social worker may be another variable that influences professional culture and the place of network intervention methodology within it. The third independent variable, for the same reasons, will be continuously observed in the bi- and multivariate analyses: the practice of the profession in the field. In subsequent analyses, other independent variables that influence the concepts being analyzed may also be considered.

Personal support networks

Since we want to understand whether there are any links between social workers’ personal support networks and their propensity to use network intervention methodology in their professional practice, we will now focus on their personal support networks. The data collection on personal support networks will focus on the everyday nature of the support, not on the reason (personal, professional, or other) this support is provided, because we want to know whether the tendency to rely on the activation or reactivation of networks to help other people is in any way linked to the personal and everyday experience that social workers have of their own support networks.

We begin the presentation of data on the personal support networks of the social workers interviewed by introducing the text of the name generator used: «In everyday life, we sometimes have to deal with needs or solve problems. Very often, it is essential to rely on the help and support of the people we are close to. Could you tell me how many (up to 6) are and what characteristics have the people who you can rely on for support or help when you need it in your daily life?». Table 2 presents the data collected using this tool and relating to the size of personal support networks.

Table 2

The size of personal support network

Total number of alters mentioned
1,817
Distribution of alters
Min	0
Max	6
Mean	3.99
SD	1.58
Number of alters (%)
0	1.8
1	4.4
2	11.2
3	22.0
4	21.5
5	14.3
6	24.8
Total	100.0
n	455

Overall, more than 1,800 alters were mentioned. The average number of people that the social workers interviewed can rely on is 3.99, and only 1.8% of them have no one to turn to in case of need in their daily lives. It should be noted, in order to have a very general comparison with the Italian context, that in 2018, the average number recorded, using the same name generator, for a sample of residents in Italy was 2.99 people (SD = 1.69, n = 1,942; Tronca, 2019)⁶. While the gender of social workers has no connection with this variable, nor is there any relationship with whether they work in the field or administrative roles within the service, we find a weak negative relationship with the number of years they have worked as social workers⁷.

Table 3

The characteristics of alters: gender

Average number of male and female alters in personal networks with at least one alter
Males
Min	0
Max	5
Mean	1.30
SD	0.94
Females
Min	0
Max	6
Mean	2.74
SD	1.31
Average percentage of male and female alters in personal support networks with at least one alter
Males
Min	0.00
Max	100.00
Mean	32.67
SD	23.62
Females
Min	0.00
Max	100.00
Mean	67.33
SD	23.62
n	421

Let us now turn to the information on alters collected with name interpreters. Radicalizing what is a well-known trend among the broader national population, personal support networks are populated mainly by women, who account for an average of 67.33% (Table 3) (at the national level and among the general population, this incidence was 53.38% in 2018, see Tronca, 2019). We note that males tend to indicate males more than females — an average of 1.74 vs. 1.28⁸ — and each gender is more prevalent in the support network of the subjects belonging to it: for males, alters males account for an average of 43.33% (female alters, obviously, for 56.67%), compared to 32.38% recorded for females⁹. Neither the number of years spent working as a social worker nor whether or not one works in the field affects the presence of males or females in the support networks of social workers.

The average age of alters is close to 49, slightly lower than that of the sample of adult residents at the national level, which is 52.07 (Tronca, 2019) (Table 4). The gender of respondents has no relation to the average age of alters. A positive correlation emerges with the years spent working as a social worker. However, this correlation disappears when controlled for the age of respondents, which appears to be positively correlated with that of alters. The relationship with working in the field — average age of alters 48.06 years — or with organizational functions in the service — 50.84¹⁰ — is also primarily attributable to the correlation between this variable and the age of respondents: 41.97 vs. 47.91 years¹¹.

Table 4

The characteristics of alters: age

Average age of alters in personal support networks with at least one alter
Min	28.75
Max	72.00
Mean	48.64
SD	8.17
n	368

Table 5

The characteristics of alters: educational qualification

Number of alters with the following educational qualifications (at least one alter in the network): None/Elementary school diploma/Middle school diploma		Percentage of alters with the following educational qualifications (at least one alter in the network): None/Elementary school diploma/Middle school diploma
Min	0	Min	0,00
Max	4	Max	100,00
Mean	0.46	Mean	11.15
SD	0.74	SD	19.28
Number of alters with the following qualifications (at least one alter in the network): Professional qualification diploma/High school diploma		Percentage of alters with the following qualifications (at least one alter in the network): Professional qualification diploma/High school diploma
Min	0	Min	0.00
Max	5	Max	100.00
Mean	1.27	Mean	31.50
SD	1.14	SD	28.80
Number of alters with the following educational qualifications (at least one alter in the network): Non-university tertiary diploma/University diploma/University degree		Percentage of alters with the following educational qualifications (at least one alter in the network): Non-university tertiary diploma/University diploma/University degree
Min	0	Min	0.00
Max	6	Max	100.00
Mean	2.11	Mean	52.65
SD	1.12	SD	33.32
Number of alters with the following qualifications (at least one alter in the network): Master’s Degree Level I or II/Specialization diploma/PhD		Percentage of alters with the following qualifications (at least one alter in the network): Master’s Degree Level I or II/Specialization diploma/PhD
Min	0	Min	0.00
Max	6	Max	100.00
Mean	0.20	Mean	4.70
SD	0.58	SD	13.29
n	410	n	410

Table 6

The social circles to which alters belong

Number of relatives (at least one alter in the network)		Percentage of relatives (at least one alter in the network)
Min	0	Min	0.00
Max	6	Max	100.00
Mean	1.48	Mean	34.89
SD	1.38	SD	31.01
Number of friends (at least one alter in the network)		Percentage of friends (at least one alter in the network)
Min	0	Min	0.00
Max	5	Max	100.00
Mean	0.99	Mean	23.04
SD	1.13	SD	26.31
Number of neighbors (at least one alter in the network)		Percentage of neighbors (at least one alter in the network)
Min	0	Min	0.00
Max	2	Max	33.33
Mean	0.03	Mean	0.63
SD	0.18	SD	3.98
Number of colleagues who are social workers (at least one alter in the network)		Percentage of colleagues who are social workers (at least one alter in the network)
Min	0	Min	0.00
Max	6	Max	100.00
Mean	0.69	Mean	18.25
SD	0.93	SD	25.48
Number of colleagues who are not social workers (at least one alter in the network)		Percentage of colleagues who are not social workers (at least one alter in the network)
Min	0	Min	0.00
Max	5	Max	100.00
Mean	0.59	Mean	15.86
SD	0.94	SD	25.33
Number of members with the respondent within a third sector association (at least one alter in the network)		Percentage of members with the respondent within a third sector association (at least one alter in the network)
Min	0	Min	0.00
Max	4	Max	100.00
Mean	0.06	Mean	1.59
SD	0.33	SD	9.05
Number of others (at least one alter in the network)		Percentage of others (at least one alter in the network)
Min	0	Min	0.00
Max	2	Max	100.00
Mean	0.22	Mean	5.74
SD	0.46	SD	13.34
n	423	n	423

The summary data on educational qualifications of the alters shown in Table 5 were obtained by reaggregating the categories originally used to record this information for each of the alters indicated: (1) No educational qualifications, elementary school diploma, middle school diploma; (2) 2-3 year upper secondary school professional qualification diploma (level II) that does not allow enrollment in university, 4-5 year upper secondary school diploma/secondary education diploma (level II) that allows enrollment in university; (3) Non-university tertiary diploma (academy of fine arts, conservatory, etc.), university diploma/degree (all types); (4) First or second level master’s degree, specialization diploma, PhD.

The most common educational qualifications overall, and therefore, on average, the most impactful (52.65%), are those in group 3, which we consider high-level. It should be noted, however, that at the national level, the most impactful qualifications are those in group 2 (43.09%, see Tronca, 2019). No difference emerges in the gender of the respondents. Even if we move on to the variables more closely related to professional experience that we are considering for these cross-references, no relationship emerges with the educational qualifications of the alters.

Table 6 presents information on the presence and incidence of different social circles in the personal support network of respondents. The question to obtain this information was as follows: «Is this person for you (indicate the type of characteristic you consider most relevant, if the person has more than one)». The question was followed by these options: relative, friend, neighbor, work colleague who is a social worker, work colleague who is not a social worker, member with the respondent within a third sector association, and other¹². It should be noted immediately that the circle with the most significant impact on the personal support network is the circle of relatives, with an average value of around 35%, followed by the circle of friends, with just over 23%. While the latter value is substantially in line with the national value recorded in 2018, which is 19.08%, the value relating to the circle of relatives is much lower, as the figure recorded in 2018 was 71.97% (for an average total number of relatives equal to 2.27, see Tronca, 2019).

The circle of relatives has less than half the impact on the personal support network for the daily lives of social workers than the standard national population of residents. What we find very interesting is the high incidence of work colleagues, particularly social workers, which exceeds 18%. Overall, the incidence of the circle of colleagues (social workers or otherwise) is 34.11%, which is essentially equivalent to that of relatives¹³.

The gender of respondents does not affect these variables, while the number of years spent working as a social worker correlates negatively with the number of alters who are relatives¹⁴ and their incidence on the personal support network¹⁵ and positively with the incidence of non-social worker colleagues¹⁶. Two relationships, which are difficult to interpret, emerge between the number¹⁷ and percentage incidence¹⁸ of the residual «circle» of others and whether or not one works in the field, in the sense that slightly higher average values are identified among those who work in organizational roles in services: 0.29 vs. 0.18 and 7.76% vs. 4.63%. One aspect that we find very interesting overall and will be explored in more detail later is the absence of relationships with the circle of social worker colleagues, which therefore seems to be a solid constant in the daily support experience of the professionals interviewed.

We conclude our analysis of social workers’ personal and daily support networks by examining the level of cohesion within these networks. To do this, respondents were asked the following name interrelator question: «And who do you think these people can count on, in case of need?». For each person mentioned, they were asked to «Answer by indicating, for each person you mentioned, whether they can rely on each of the others you mentioned and on you».

As can be seen from the name interrelator text, this question allowed us to record, for each respondent, the outdegree of support ties concerning the indicated alters, which has an average value of 4.00 (SD = 1.52, min = 0, max = 6, n = 422). Thanks to the outdegree, it was also possible to determine the average level of reciprocity of the personal support networks of social workers, which is 98.59% (SD = 7.80, min = 0.00, max = 100.00, n = 422). The respondents’ gender does not impact these two variables, while the number of years they have worked as social workers has a weak negative correlation with the outdegree¹⁹. Whether or not they work in the field has no impact on the two variables we are considering.

In addition to studying the reciprocity of help, the name interrelator naturally made it possible to identify, for each respondent, the first-degree zone of the personal support network, reconstructing it through a directed graph. To study its level of cohesion and, therefore, of closure/openness, we used the measure of ego-centric density, which focuses on direct support links between alters, since it is possible to take for granted, also in consideration of the very high reciprocity identified, those for and from the ego. It should be noted that the density of a graph is obtained by calculating the ratio between the number of ties activated and the number of ties that can potentially be activated, and therefore varies between 0 and 1 (Wasserman & Faust, 1994). Ego-centric density, which varies between 0 and 1, is calculated by removing the ego and all its ties from the graph (Scott, 2000). This measure was determined for all social workers who indicated at least two alters and provided valid values (yes/no) in response to questions about potential support ties between them.

Table 7

Ego-centric density of the personal support network (network with at least two alters)

Min	0.00
Max	1.00
Media	0.55
Dev. Std.	0.36
n	311

As shown in Table 7, the average ego-centric density and, therefore, the average level of cohesion of social workers’ personal support networks is relatively high, especially when compared to the national average in 2018, which was 0.38 (SD = 0.38, n = 1,007; Tronca, 2019). The social workers interviewed thus tend to have a support network that is more bonding, i.e., more community-based, than that recorded at the national level in Italy. The gender of social workers does not correlate with this structural characteristic of their support networks. Neither the number of years they have been working in this profession nor the fact that they work in the field or in an organizational capacity within the service correlates with the level of cohesion of their personal support networks.

Culture of network intervention

Coming to consider the methodology of network intervention, it is worth investigating what the professional culture of social workers is and how the method of network intervention fits in.

Table 8 presents the descriptive analysis of the response to the question «In your opinion, for a social worker, how important is each of the following areas of expertise?». The answer was given on a scale from 0 to 10, and the mean values obtained are listed in the table in descending order for the importance given to the different skills. As can be seen, with a mean value of 8.89, competencies in network intervention methodology come right after those related to need assessment. Therefore, the sample’s professional culture is oriented toward using network intervention methodologies. The finding is also reflected in the question aimed at assessing to what extent, from 0 to 10, the respondent feels in his work that he applies different skills (Table 9). Again, in this second case, skills related to network intervention come in second place after those of need assessment.

Table 8

Ranking of the importance attributed to different professional competencies (range 0-10)

Competencies ranking	Mean	SD	Min	Max	n
Competencies in need assessment	9.33	0.92	5	10	494
Competencies in network intervention methodology	8.89	1.16	4	10	495
Competencies in service design	8.62	1.37	2	10	494
Management competencies	8.35	1.45	1	10	492
Competencies in counseling	8.28	1.49	2	10	493
Legal competencies	8.19	1.53	0	10	494
Psychological competencies	7.90	1.40	0	10	494
Sociological competencies	7.71	1.49	3	10	494
Pedagogical competencies	7.24	1.72	0	10	494

Table 9

Ranking related to how much they happen to use, in social work, different professional skills (range 0-10)

Competencies ranking	Mean	SD	Min	Max	n
Competencies in need assessment	8.89	1.31	1	10	492
Competencies in network intervention methodology	8.14	1.66	1	10	492
Competencies in counseling	7.48	1.99	0	10	490
Psychological competencies	7.48	1.65	0	10	490
Legal competencies	7.44	1.94	0	10	492
Management competencies	7.29	2.44	0	10	489
Competencies in service design	6.96	2.54	0	10	492
Sociological competencies	6.64	1.97	0	10	491
Pedagogical competencies	6.28	2.43	0	10	491

To understand professional culture, it is appropriate to investigate respondents’ opinions and observe other skills associated with the network intervention methodology. The principal component analysis of the importance of different professional competencies shows two components (Table 10) among which the different skills are polarized, except for the one related to counseling, which is equally distributed between the two. In one component, theoretical skills (legal, psychological, sociological, and pedagogical) are placed. On the other hand, skills more related to social intervention and service organization are favored by the respondents, and among them, skills in network intervention methodology, which is associated with skills in need assessment, service design, and management.

From this second component emerges a professional culture that sees the social worker not in a counseling role in the context of an individual helping relationship, but instead, engaged in relational guidance of the aid recipient. In this case, it is a different way of interpreting the counseling function in the helping relationship. Help does not come exclusively from the supportive function the social worker performs. The real actors of help are the primary, secondary, and institutional networks. These can perform the task of caring for the citizen in need. The social worker is in charge of guiding the citizen within these networks, on the one hand by orienting them in their journey within these networks, and on the other hand, through the skills of network intervention as well as those of service system management and design, by mobilizing and managing the primary relational resources.

It also emerges that the concern for network intervention methodology is strongly practical. The analysis we have just conducted shows that the network intervention methodology, far from being considered a specific theoretical area of social service, is presented with a cultural valence that is primarily practical-applicative.

Table 10

Principal component analysis of the importance attributed to different professional competencies

	Component
Competencies	1	2
Legal competencies	0.692	0.219
Psychological competencies	0.855	0.119
Sociological competencies	0.707	0.294
Pedagogical competencies	0.747	0.218
Competencies in network intervention methodology	0.383	0.693
Competencies in counseling	0.499	0.470
Competencies in need assessment	0.286	0.708
Competencies in service design	0.088	0.862
Management competencies	0.208	0.785
% of variance	31.053	30.659

Note: Rotation Method: Varimax with Kaiser Normalization, rotation converged in 3 iterations; KMO = 0.847; Bartlett’s Test of Sphericity Approx. Chi-Square = 1799.669, df = 36, p < .0001; n = 488.

A similar principal component analysis asked respondents how much they use professional skills daily (Table 11). The analysis of the results still reveals two components, among which the different skills are polarized, except in this case, the skills related to need assessment, which is evenly distributed between the two components, highlighting that, in practice, need assessment is an activity that does not «fit» with one particular mode of intervention rather than another.

The two components are delineated as follows. In one area, theoretical skills (legal, psychological, sociological, and pedagogical) are gathered, and counseling skills are added. This finding leads to the interpretation of a culture of social work that considers the professional as a therapist whose intervention, guided by a set of theoretical skills, through the activity of counseling, realizes direct help, almost a kind of socio-psycho-pedagogical therapy, for the citizen user.

The second component, on the other hand, sees gathered together with the skills of network intervention more practical/operational skills such as planning and management skills that configure social intervention not as help brought by the helping professional but by the social networks around them that need to be managed and coordinated by the professional, so that they can express their latent potential to support the person in need.

This second principal component analysis devoted to the social worker’s competencies confirms a perspective in which two ways of considering help are placed in front of each other: one in which the profession places itself in a therapeutic position of direct help from the social worker, while the other seems to privilege, as the leading actor of help, the social network in which the citizen are embedded.

Table 11

Principal component analysis of how much, in social work, different professional skills are used

	Component
Competencies	1	2
Legal competencies	0.656	0.100
Psychological competencies	0.745	0.075
Sociological competencies	0.662	0.292
Pedagogical competencies	0.719	0.042
Competencies in network intervention methodology	0.445	0.628
Competencies in counseling	0.606	0.287
Competencies in need assessment	0.433	0.428
Competencies in service design	0.068	0.892
Management competencies	0.080	0.874
% of variance	30.050	25.798

Note: Rotation Method: Varimax with Kaiser Normalization, rotation converged in 3 iterations; KMO = 0.796; Bartlett’s Test of Sphericity Approx. Chi-Square = 1360.277, df = 36, p < .0001; n = 485.

Structure of network intervention

This section focuses on the topic of network intervention in social work. It considers the extent to which the sample analyzed uses network intervention methodology in the performance of their profession, and the areas in which this methodology is most widely used.

Through data analysis, a reflection will then be proposed on the place that network intervention methodology occupies in the professional culture of the interviewees, through a comparison with the importance attributed to other methodologies/modalities of social work. It will also be observed which mode of network intervention is considered most relevant: whether it is an intervention that enhances the bonding function of social ties or the bridging position and function that the network nodes can perform.

We asked the respondents which four activities performed in their profession engage them for more hours. The activities they could choose are presented in descending order of impact in Table 12. Social work interviewing is the most frequent activity. It turns out to be the fundamental basic professional tool that is used by 77.1% of respondents. Other activities are reported by less than half of the respondents, in the following order: teamwork, case management or information and referral services, networking with other institutions, administrative support activities, networking intervention, individual treatment plan, professional writing, home visits, coordination of practitioners, assessment activities, planning and design of services, their coordination, budget management, community work, work with user groups, and research and communication activities.

Table 12

The four activities most frequently done at work (multiple responses)

	A.V.	Percentage of cases (n=472)
Social work interviews	364	77.1%
Teamwork	230	48.7%
Case management or information and referral services	218	46.2%
Network with other agencies, institutions, and subjects	168	35.6%
Administrative support activities	163	34.5%
Network intervention	135	28.6%
Individual treatment plan	107	22.7%
Professional writing	98	20.8%
Home visits	96	20.3%
Coordinate practitioners	68	14.4%
Assessment activities	53	11.2%
Planning	40	8.5%
Service plan	39	8.3%
Coordination of services	32	6.8%
Budget management	30	6.4%
Community work	21	4.4%
Work with user groups	17	3.6%
Research and communication activities	4	0.8%
Other	5	1.1%

Network intervention thus ranks sixth in importance, indicated by 28.6% of respondents. Network intervention is preceded mainly by internal service management and interrelation activities with other institutions. Network intervention seems, therefore, quite relevant as an operational strategy, especially if we consider that it precedes, though not by much, individual planning.

Table 13

Percentage of use of the network intervention methodology by social intervention areas

Social intervention area (no. of respondents)	%
Addiction area (n=50)	40.0
Immigration (n=55)	40.0
Mental health (n=56)	37.5
Hospital social work (n=8)	37.5
Adult justice (n=25)	36.0
Youth (n=67)	34.3
Maternity, childhood (n=102)	30.4
Employment integration (n=40)	30.0
Adults in distress (n=159)	29.6
Family (n=203)	26.6
Women who have experienced abuse (n=50)	26.0
Disability (n=100)	24.0
Older adults (n=213)	21.6
Juvenile criminal (n=11)	18.2

Those who have network intervention as one of their four main activities do so predominantly in the areas of addiction, immigration, mental health, and hospital services, which, however, have too few cases to be considered reliable (Table 13). The areas in which the methodology of network intervention is less frequent are juvenile justice, the respondents on which the percentage is calculated are few, the older adults and disability, which show a lower incidence of 25%.

Social workers who perform this activity do not have significantly different years of service from those who do not. Nor do gender, the distinction between work in the field and work with organizational functions, or the public/private institutional context of work significantly affect the propensity to use the network intervention methodology.

The interviewees were asked to indicate to what extent from 0 to 10 they believe that in order to carry out a good network intervention, for the support of a person taken in charge by the social service, one should «strengthen the ties between all members of the person’s network» and to what extent one should «identify, in particular, some members of the personal network of the person taken in charge by the social service to whom one should recognize an auxiliary function of coordination» (Table 14). This second mode of intervention adopts a perspective that we might call bridging, as it considers specific nodes as bridges to facilitate access to symbolic or material resources by the person needing help. On the other hand, the first mode of intervention employs a perspective that we call bonding, as it believes that the overall quality of relationships within the network enables the person taken into care to receive the resources necessary to improve their well-being (Di Nicola et al., 2011a). An analysis of the data shows that respondents do not take a definite stand for either perspective. With mean values of 8.16 (SD = 1.62) for the «bonding» variant and 8.13 (SD = 1.48) for the «bridging» variant, respondents consider both modes of implementing the network intervention to be fundamental. When calculating the difference between the scores of the two variables, the largest proportion of respondents — 42.4% — assigned identical scores to both modes of intervention. For the remaining participants, scores were 28.9% higher for the bonding variant and 28.7% higher for the bridging variant, respectively (Table 15).

Table 14

Importance of the different structural ways to carry out network intervention

Ways of intervention	Mean	SD	Min	Max	n
Strengthen the ties between all members of the person’s network (bonding)	8,16	1,62	0	10	490
Identify, in particular, some members of the personal network of the person taken in charge by the social service to whom one should recognize an auxiliary function of coordination (bridging)	8,13	1,48	0	10	486

Table 15

Difference in the scores given to bridging and bonding modes (n = 484)

Ways of intervention	%
Bridging > Bonding	28,7
Bridging = Bonding	42,4
Bridging < Bonding	28,9

If we examine the relationships between these last two variables, on the one hand, and gender, years of work as a social worker, working in the field/organizing, the presence of networking among the four main activities, and whether one works in the public or private sector, on the other, no statistically significant relationships emerge.

Personal support network and network intervention

We conclude by presenting the results of some multivariate models, in which we sought to understand whether and how the characteristics of personal support networks affect, net of the independent variables we most frequently used in the paper: (1) on the culture of network intervention — measured with variables relating to the importance attributed to the network intervention methodology and the propensity to use this intervention methodology; (2) on the structure of network intervention, understood as the tendency to create a predominantly bonding or bridging intervention network.

Table 16

Multiple linear regression models: culture of network intervention methodology dependent variables (range 0-10)

		Dependent variables
		Level of importance attributed to the skills of network intervention methodology		Frequency of use of skills of network intervention methodology
Predictors		Model 1	Model 2	Model 1	Model 2
(Constant)	b Std. Error	8.387 (0.373)***	7.940 (0.436)***	7.128 (0.457)***	6.503 (0.536)***
Gender and profile variables concerning work
Gender (female)	b Std. Error ß VIF	0.727 (0.331) 0.128* 1.001	0.677 (0.330) 0.119* 1.005	1.349 (0.406) 0.193** 1.001	1.340 (0.406) 0.192** 1.005
Years of experience as a social worker	b Std. Error ß VIF	0.001 (0.006) 0.008 1.064	0.001 (0.006) 0.014 1.077	0.005 (0.008) 0.039 1.064	0.006 (0.008) 0.050 1.077
Profile concerning work (work in the field)	b Std. Error ß VIF	-0.315 (0.154) -0.123* 1.064	-0.311 (0.154) -0.122* 1.069	-0.183 (0.189) -0.058 1.064	-0.173 (0.189) -0.055 1.069
Personal support network variables
Size of personal support network	b Std. Error ß VIF	-	0.103 (0.048) 0.126* 1.034	-	0.116 (0.059) 0.115 1.034
Ego-centric density of personal support network	b Std. Error ß VIF	-	-0.030 (0.189) -0.009 1.007	-	0.263 (0.232) 0.066 1.007
Percentage of incidence of social workers colleagues in personal support network	b Std. Error ß VIF	-	0.004 (0.003) 0.100 1.035	-	-0.001 (0.003) -0.010 1.035
R2		0.032	0.054	0.043	0.061
Adjusted R2		0.022	0.034	0.033	0.040
Std. Error		1.123	1.116	1.377	1.372
Durbin-Watson		2.042	2.061	1.952	1.978
n		287	287	287	287

Note: The highest code for gender and work profile and the standard errors are shown in parentheses. Predictor entry criterion: enter. ***p < .001, **p < .01, *p < .05.

The models we will present are multiple linear regression models. The first group of predictors consists of gender and the two profile variables related to work that we have mainly used in the paper: the number of years of experience as a social worker and whether or not one works in the field. Dummy variables have been created for both the latter variable and gender.

A second group of predictors contains those most closely related to personal support networks. These include network size, ego-centric density, and the percentage of colleague social workers. At the same time, the reasons for including the first two variables in the multiple linear regression models are quite obvious, as they are two fundamental structural dimensions of networks; the reason for including the last variable in the models is related to its great relevance to social workers’ personal support networks and the fact that it is, curiously, completely independent of the work-related profile variables we have considered.

A model containing the first group of predictors and one containing the first and second groups of predictors was created for each dependent variable. We note that, in order to more accurately assess the ability of personal support networks to influence dependent variables, for each dependent variable, we used cases with valid values on all predictors (both from the first and second groups) in both models dedicated to it.

An initial analysis of multiple linear regression models looked at the importance given to the professional methodology of network intervention as a dependent variable (Table 16). The first model included the independent variables of gender, years of service, and fieldwork in the analysis. The variables relating to gender and field work are significant and affect the increase in the importance attributed to the methodology of network intervention. In particular, being female contributes to increasing the importance of network work, while field work decreases it. The adjusted R² value shows that the model explains 2.2% variance of the dependent variable. In model 2, adding variables relating to the personal support network increases the variance explained by the model to 3.4%. In this second model, in addition to finding the influences seen in model 1 (being female and working in the field), the size of the support network positively impacts the importance attributed to the network methodology. In essence, having a larger personal support network positively impacts the importance attributed to the networking methodology.

Also in Table 16, but in the last two columns, are the two models relating to the dependent variable that present data on the frequency of use of skills of network intervention methodology. Model 1 partly mirrors the results seen in a similar model applied to the variable relating to the level of importance attributed to the skills of network intervention methodology, as it shows a positive incidence on the part of the female gender. Adding network variables increases the explained variance, which rises from 3.3% to 4.0%, although no significant incidence emerges from the network variables.

Table 17

Multiple linear regression models: structure of network intervention methodology dependent variables (range 0-10)

		Dependent variables
		Bonding perspective of network intervention methodology		Bridging perspective of network intervention methodology
Predictors		Model 1	Model 2	Model 1	Model 2
(Constant)	b Std. Error	6.364 (0.498)***	5.841 (0.585)***	6.721 (0.467)***	6.309 (0.549)***
Gender and profile variables concerning work
Gender (female)	b Std. Error ß VIF	1.606 (0.442) 0.212*** 1.001	1.621 (0.440) 0.214*** 1.005	1.211 (0.413) 0.173** 1.001	1.238 (0.413) 0.177** 1.005
Years of experience as a social worker	b Std. Error ß VIF	0.007 (0.008) 0.048 1.069	0.007 (0.008) 0.051 1.082	0.014 (0.008) 0.105 1.072	0.015 (0.008) 0.115 1.087
Profile concerning work (work in the field)	b Std. Error ß VIF	0.237 (0.208) 0.069 1.069	0.230 (0.208) 0.066 1.074	0.143 (0.195) 0.044 1.071	0.148 (0.195) 0.046 1.078
Personal support network variables
Size of personal support network	b Std. Error ß VIF	-	0.046 (0.065) 0.042 1.038	-	0.049 (0.061) 0.048 1.036
Ego-centric density of personal support network	b Std. Error ß VIF	-	0.599 (0.253) 0.138* 1.006	-	0.418 (0.239) 0.103 1.008
Percentage of incidence of social workers colleagues in personal support network	b Std. Error ß VIF	-	0.000 (0.004) -0.006 1.043	-	-0.004 (0.003) -0.064 1.041
R2		0.050	0.070	0.040	0.056
Adjusted R2		0.039	0.050	0.029	0.035
Std. Error		1.497	1.489	1.400	1.395
Durbin-Watson		1.948	1.970	1.970	1.960
n		284	284	281	281

Note: The highest code for gender and work profile and the standard errors are shown in parentheses. Predictor entry criterion: enter. ***p < .001, **p < .01, *p < .05.

Another series of multivariate analyses was carried out, choosing the two approaches to network intervention, bonding and bridging, as dependent variables (Table 17). Analysis of models 1 and 2 relating to the bonding approach shows that female gender has a positive impact on the bonding approach to network intervention. If we then add the variables relating to the personal support network to the profile variables, we see that the variance explained by the model increases from 3.9% to 5.0%, and that this increase depends in a statistically significant way on the ego-centric density of the support network experienced by the respondents. Meanwhile, analysis of the regression models relating to the bridging concept of network intervention (last two columns of Table 15) reveals only the impact of female gender on the dependent variable. In this case, however, in the transition from model 1 to model 2, which also contains variables relating to personal support networks, there is an increase in the proportion of variance explained, from 2.9% to 3.5%. A fascinating finding that emerges from these latest analyses is that a measure of structural closure of support networks, i.e., their ego-centric density, has a positive and exclusive impact on the bonding approach to the provision of network intervention.

Discussion

The analysis of the data, conducted with a focus on the methodology of network intervention in the professional field of social work, allows us to observe how it fits into the Italian context. The methodology of network intervention can be observed both from the point of view of professional practice and from the interviewees’ professional culture.

A specific aspect of this work is exploring the links between network intervention methodology and the characteristics of social workers’ personal support networks. The analysis of personal support networks revealed differences compared to the support networks identified in Italy, using the same tools, in a sample of the adult resident population. The support network of social workers is, in fact, broader and denser. This information should be read together, considering that broader networks should, in theory, make achieving higher density levels more challenging. The support networks of social workers are, therefore, small communities on which social workers can rely daily.

The personal support networks of social workers show a certain tendency toward homophily about gender and age. Unlike the national population, for the social workers interviewed, high-level educational qualifications, i.e., similar to those they possess, in their support network are the more relevant, regardless of their specific professional experience. These characteristics of social workers’ personal support networks contribute to portraying them as homogeneous communities of mutual support. The high level of reciprocity in support relationships emerges not only as a particular feature of egos but also from the significant average level of ego-centric density, which emerges from the survey of direct support relationships between alters.

A factor that further defines this phenomenon is the high incidence of social workers’ colleagues, particularly fellow social workers, in their personal support networks. A comparison of these data with national data shows that, for social workers, colleagues seem to partially replace relatives in terms of providing support. This impression is further confirmed by the number and incidence of relatives in personal support networks decreasing as the number of years spent working as a social worker increases. In contrast, the incidence of colleagues who are not social workers increases. Therefore, years of service do not affect the number and incidence of social worker colleagues in the personal support network. Given the great importance of this circle, we can conclude that a basic culture of mutual aid is strongly connected to this profession.

The survey conducted an in-depth analysis of the impact of the culture of internal network support within the profession on the working methods used, focusing on the methodology of network intervention. By comparing the importance that respondents attribute to different professional skills, including those related to network intervention methodology, and analyzing the data using principal component analysis, two different ways of understanding social work emerged from the empirical evidence. On the one hand, it is seen as an individualized intervention on a case. An intervention that involves the social worker in a direct relationship with the person. The intervention takes on the character of therapeutic counseling based on theoretical skills provided to the social worker by the humanities and social sciences. These theoretical skills are probably the tools for promoting awareness and a change in attitude in the person being assisted. On the other hand, the interviewees represented the profession as a more operational and practical activity that sees the social worker engaged in counseling that helps the person move within the primary and secondary social networks in which they are inserted. At the same time, social workers carry out network interventions by acting as mobilizers and managers of primary and secondary latent network resources in which the client is or may be involved.

Multivariate analysis of the importance attributed to different professional skills shows that today’s representation of the profession reflects the methodological shift in the 1980s. That is, the transition from an idea of social intervention as case work to a relational guidance action of the networks in which the person in need of help is involved. Another issue explored in depth was how, according to the interviewees, intervention within primary support networks should be carried out. The fundamental question posed by the research is whether, for a more effective intervention, it is more appropriate to strengthen the bonds within the primary support network in general, according to a bonding perspective, or whether, according to a bridging perspective, it is better to identify specific nodes capable of facilitating access to resources by the user. The data show that many social workers place the two modes of intervention on the same level, while the remainder are equally divided in their preference for one perspective or the other.

In a third step of the analysis, the research examined the existence of an impact that the personal support networks of social workers themselves may have on the way network intervention is conceived and implemented. The results of multiple linear regressions show that variables such as gender and field work tend to influence the importance attributed to the methodology of network work in the profession. Female respondents and those who work within services, with organizational tasks, tend to attribute greater importance to network intervention. Of these clear causal links, only that with female gender remains if we move on to the frequency with which network intervention is used. However, the size of the personal support network affects the importance of network work. Therefore, social workers’ support networks can transfer the sense of community we have discussed into the perception of a specific social work practice, based primarily on relationships and networks.

Finally, let us move on to network intervention’s bonding and bridging indicators. We observe that this causal power of social workers’ personal support networks complements, in some way, the picture of its effects. We observe that a measure of closure, i.e., bonding of networks, such as ego-centric density, establishes a single positive relationship precisely with a bonding conception of network intervention. The low levels of variance explained by multiple linear regression models indicate that relevant variables related to organizational contexts and relationships with people receiving social services should be considered to better explain the use of the network intervention methodology. In this research, however, we were particularly interested in obtaining initial evidence on the relationship between the personal support networks of social workers themselves and the use of a working method that relies precisely on the activation and reactivation of these networks among the people they assist. The evidence that emerged on this point appears interesting and prompts reflection on the need to further explore this line of research, especially in relation to the conditions of individuals taken into care by social services. As recent literature on social work shows, adopting a network intervention methodology can help empower vulnerable people within the anti-oppressive social service perspective (Dominelli, 2002).

Conclusions

The data show that networking is among the four most frequently performed activities for about one-third of respondents. There are areas of social intervention in which networking is used more frequently. These are primarily areas in which psychological distress is particularly relevant, such as immigration and youth.

Network intervention emerged and became established in social work methodology theory during the 1980s due to a new understanding of well-being. At the time, subjective well-being was considered not only a coping skill but also a state of balance between the individual and their biological, psychological, and social environments. Social work focuses on methods to improve the relationship between people in distress and their social environment, emphasizing network intervention. Forty years after this turning point, we have observed that the culture of network intervention has found a place in the professional practice of social workers in a region of Italy, who view the profession as an emancipatory social practice conducted within social networks, rather than as a profession of individual assistance.

In particular, social workers build up extensive and cohesive personal support networks around themselves, which also positively impact the importance of the professional methodology of network intervention and the morphological structure that such intervention should take.

Regarding the analyses carried out, we must highlight some limitations. The first consists of collecting fewer questionnaires than had been requested. However, it should be noted that the collection of questionnaires on 536 cases, out of a total population of 2,970, seems to be a reasonably satisfactory result, considering that a probabilistic sample was constructed. The second limitation is related to the small number of males involved in the research, which aligns with the proportionally stratified sampling plan concerning gender. This small number may, in some analyses, have influenced the extent of the differences identified. However, it should be noted that the frequent recurrence of relationships with the gender variable, even concerning different topics, seems to support attitudes and opinions that are substantially differentiated concerning gender. A possible improvement in the quality of the research results could come from sampling that overestimates the presence of males. Another limitation is related to the fact that the concrete practice of network intervention deserves more detailed data collection.

Ultimately, this study provides further evidence of the enormous importance of social resources, especially in areas of intervention such as social services. Regarding social problems, practical strategies for utilizing social resources can compensate for any scarcity of other resources, such as economic resources, and, above all, focus on people’s well-being.

Bibliography

Besson, C. (1995). Per intervenire in rete. In L. Sanicola (Ed.), Reti professionali e intervento sociale. Liguori.

Bidart, C., & Charbonneau, J. (2011). How to Generate Personal Networks: Issues and Tools for a Sociological Perspective. Field Methods, 23(3), 266-286.

Bidart, C., & Degenne, A. (2005). Introduction: the Dynamics of Personal Networks. Social Networks, 27(4), 283-287.

Brodeur, C., & Rousseau, R. (Eds.) (1984). L’intervention de réseaux, une pratique nouvelle. Éditions France-Amérique.

Bulmer, M. (1987). The Social Basis of Community Care. Allen & Unwin.

Carkhuff, R. (1987). The Art of Helping. Human Resources Development Press.

Challis, D., & Davis, B. (1986). Case Management in Community Care. Gower Publishing.

Cooper, J. (1989). From Case work to Community Care. British Journal of Social Work, 19, 177-188.

Di Nicola, P., Stanzani, S., & Tronca, L. (2011a). Personal Networks as Social Capital: a Research Strategy to Measure Contents and Forms of Social Support. Italian Sociological Review, 1(1), 1-15.

Di Nicola, P., Stanzani, S., & Tronca, L. (2011b). Capitale sociale e benefici pubblici: reti di prossimità e cicli di vita della famiglia. Sociologia e politiche sociali, 14(1), 23-69.

Dominelli, L. (2002). Anti-Oppressive Social Work: Theory and Practice. Palgrave Macmillan.

Fernández-Peña, R., OvallePerandones, M.-A., MarquésSánchez, P., OrtegoMaté, C., & SerranoFuentes, N. (2022). The use of social network analysis in social support and care: a systematic scoping review. Systematic Reviews, 11(9). https://doi.org/10.1186/s13643021018762

Folgheraiter, F. (2004). Relational social work: Toward networking and societal practices. Jessica Kingsley.

Folgheraiter, F., & Raineri, M.L. (2012). A critical analysis of the social work definition according to the relational paradigm. International Social Work, 55(4), 473-487.

Guay, J. (Ed.) (1987). Manuel québécois de psychologie communautaire. Gaëtan Morin Éditeur.

Lasala, C. (2022). Clinical Social Work with Individuals, Families and Groups: the Healing Power of Relationships. Taylor & Francis.

Maguire, L. (1983). Understanding social networks. Sage.

Meo, A. (1999). Relazioni, reti e «social support». Rassegna Italiana di Sociologia, XXXX(1), 129-158.

Miller, W.R., & Rollnick, S. (2002). Motivational Interviewing. Second Edition. Preparing People for Change. The Guilford Press.

Moreau, M.J. (1979). A Structural Approach to Social Work Practice. Canadian Journal of Social Work Education / Revue canadienne d’éducation en service social, 5(1), 78-94.

Parsons, T. (1972). Definitions of Health and Illness in the Light of American Values and Social Structure. In E.G. Jaco & E. Gartley (Eds.), Patients, Physicans and Illness: A Resource Book in Behavioral Science and Health (pp. 165-187). Collier-Macmillan.

Payne, M. (1986). Social Care in the Community. Macmillan.

Perlman, H.H. (1957). Social Case work: A Problem-solving Process. Chicago University Press.

Reid, W., & Epstein, L. (1972). Task Centered Case work. Columbia University Press.

Richmond, M. (1922). What is Social Case work? Russell.

Robins, G., Lusher, D., Broccatelli, C., Bright, D., Gallagher, C., Karkavandi, M. A., Matous, P., Coutinho, J., Wang, P., Koskinen, J., Roden, B., & Sadewo, G.R.P. (2023). Multilevel network interventions: Goals, actions, and outcomes. Social Networks, 72, 108-120.

Rogers, C.R. (1951). Client-centered therapy. Houghton Mifflin Harcourt.

Sanicola, L. (1996). La contribution de l’intervention de réseau à la prévention. Nouvelles pratiques sociales, 9(2), 49-64.

Sanicola, L. (2009). Dinamiche di rete e lavoro sociale. Un metodo relazionale. Liguori.

Scott, J. (2000). Social Network Analysis: A Handbook (2nd edition). Sage.

Secondulfo, D., Tronca, L., & Migliorati, L. (2017). Primo rapporto dell’Osservatorio sui consumi delle famiglie. Una nuova normalità. FrancoAngeli.

Seed, P. (1990). Introducing network analysis in social work. Jessica Kingsley.

Simmel, G. (1908). Soziologie. Dunker & Humblot.

Song, L., Son, J., & Lin, N. (2011). Social Support. In J. Scott, & P.J. Carrington (Eds.), The Sage Handbook of Social Network Analysis (pp. 116-128), Sage.

Thompson, N. (1992). Anti-Discriminatory Practice. Palgrave Macmillan.

Tronca, L. (Ed.) (2012). Personal network analysis. Monographic issue of Sociologia e Politiche Sociali, 15(2).

Tronca, L. (2013). Sociologia relazionale e social network analysis. Analisi delle strutture sociali. FrancoAngeli.

Tronca, L. (2019). Capitale sociale e consumi: il ruolo dei personal network di sostegno nelle dinamiche di consumo. In L. Tronca, & D. Secondulfo (Eds.), Secondo rapporto dell’Osservatorio sui consumi delle famiglie. Il consolidamento dei nuovi profili di consumo (pp. 89-122). FrancoAngeli.

Valente, T.W. (2012). Network interventions. Science, 337(6090), 49-53.

Wasserman, S., & Faust K. (1994). Social Network Analysis: Methods and Applications. Cambridge University Press.

Watzlawick, P. & Weakland, J.H. (Eds.) (1977). The Interactional View: Studies of the Mental Research Institute, 1965-1974. W.W. Norton.

Wellman, B. (2007a). The Network is Personal: Introduction to a Special Issue of Social Networks. Social Networks, 29(3), 349-356.

Wellman, B. (2007b). Challenges in Collecting Personal Network Data: the Nature of Personal Network Analysis. Field Methods, 19(2), 111-115.

---

1. ¹ The research was funded by the Department of Human Sciences at the University of Verona. The research project, including the questionnaire used, was approved by the Ethics Committee of the Department of Human Sciences at the University of Verona – Repertory no. 562/2019, Protocol no. 226153 of June 24, 2019.

2. ² There were a total of 714 accesses to the questionnaire. However, four individuals completed the questionnaire after the closing date (specifically, between December 1 and 2, 2019); there were 89 voluntary interruptions of valid interviews (up to the point of interruption), and 85 subjects were screened out: 25 did not accept the privacy policy required by the law on privacy protection (GDPR 2016/679) and 60 declared themselves «unemployed». The questionnaire was administered in Italian. For all questions, including the one on gender, the options «Do not know» and/or «No answer» were always provided, which, if indicated, we consider as missing values. Unless expressly indicated, the values commented on are always to be considered net of missing values. The article always reports the number of cases with valid values (n) to which the data presented refer. The data was analyzed using the software IBM SPSS Statistics – Version 23 (Release 23.0.0.0).

3. ³ This research, carried out in 2009, recorded an average size of 4.27 for the personal support network of individuals residing in Italy and aged between 25 and 80 (SD = 2.13; min = 1; max = 14; n = 1,219). Since the CAWI data collection method required a maximum number of responses to be defined, 50% was added to this value, and the final value was rounded up. Obviously, this choice directly affects the size of personal networks and, indirectly, cohesion measures: quite simply, the larger a network is, the less likely it is that all nodes will be connected to each other.

4. ⁴ For space reasons, we will not be able to address the issue of homophilia in these networks directly here. However, the particular configuration of the cross-references that we will present with the attributive characteristics of the alters will allow us to address this issue indirectly, especially regarding the gender and age of the respondents. Regarding educational qualifications and their social status, more generally, the high degree of homogeneity of the reference population makes it uninteresting to conduct a detailed analysis here.

5. ⁵ This variable is, of course, strongly and positively correlated with the age of respondents (Pearson’s r = 0.923, p < .001 (2-tailed), n = 450).

6. ⁶ In December 2020, at the height of the COVID-19 pandemic, the Household consumption monitoring centre once again surveyed personal support networks among a sample of adults residing in Italy. Given the exceptional circumstances surrounding the 2020 data collection, we have chosen to report some of the data collected in 2018.

7. ⁷ Pearson’s r = -0.104, p < .05 (2-tailed), n = 436.

8. ⁸ F(1, 415) = 4.281; p < .05.

9. ⁹ F(1, 415) = 3.928; p < .05.

10. ¹⁰ F(1, 353) = 8.671; p < .01.

11. ¹¹ F(1, 347) = 24.438; p < .001 (listwise about the controlled variable).

12. ¹² This residual category is not strictly speaking a social circle; we will refer to it as such only for the sake of convenience in our presentation.

13. ¹³ To confirm the relevance of this information, we would like to point out that, in the national data of the Household consumption monitoring centre for 2018, the overall average incidence rate of social circles of relatives and friends exceeds 90%, with a tendency towards a 70/20 ratio between the circle of relatives and friends, even for all individual categories of working respondents (public/private sector employees, self-employed workers, atypical workers).

14. ¹⁴ Pearson’s r = -0.156, p < .01 (2-tailed), n = 406.

15. ¹⁵ Pearson’s r = -0.135, p < .01 (2-tailed), n = 406.

16. ¹⁶ Pearson’s r = 0.102, p < .05 (2-tailed), n = 406.

17. ¹⁷ For ANOVA, in the presence of a statistically significant Levene’s test, the Brown-Forsythe test is taken into consideration: Brown-Forsythe test = 4.236, df1 = 1, df2 = 167.168, p < .05, n = 409.

18. ¹⁸ Brown-Forsythe test = 4.200, df1 = 1, df2 = 170.036, p < .05, n = 409.

19. ¹⁹ Pearson’s r = -0.099, p < .05 (2-tailed), n = 404.