Why We Should Stop Talking About Objectivity and Subjectivity in Social Work

In debates about knowledge in social work, the terms ‘objectivity’ and ‘subjectivity’ are frequently used with varying degrees of positive and negative connotations. We argue that the terms have become so ambiguous that they should be avoided. In its place, we suggest focusing on the individual attributes associated with objectivity and subjectivity and consider how the desirable attributes can be strengthened and the undesirable ones avoided. This division differs significantly from the typical objective/ subjective division. We examine three key social work issues: the contribution of empirical research, dealing with dissent and the role of the personal. When the attributes of objectivity and subjectivity are examined in detail, it becomes apparent that they vary in how desirable and how feasible they are. A more precise use of language makes it easier to see the contributions of values, bias and power in social work policy and practice and reduce the risks of people over-claiming the reliability and neutrality of their assertions.


Introduction
Social workers have long debated whether their practice is an art or a science. In that debate, some see objectivity as something to aspire to while others consider it conflicts with the empathic, context-specific, interpretive understanding that informs their practice. From a study of the use of the term in the literature, our conclusion is that it has acquired so many meanings that it obscures more than illuminates the serious issues at the heart of debates on how to improve services to those who need social work help. In its place, we suggest focusing on the individual attributes associated with objectivity and subjectivity and consider how the desirable attributes can be strengthened and the undesirable ones avoided. This division differs significantly from the typical objective/subjective division.
As a rough generalisation, the 'objective' ideal is more favoured by managers, policy makers and academics, while those involved in direct work give more value to their personal skills and understanding. Consequently, many practitioners share the experience of those reported in Hardesty's (2015) study of child-protection workers in Mississippi that gives a vivid picture of how an agency that privileges the achievement of objectivity over the other skills of social workers influences their sense of what matters, and leads them to feel that they are living in two worlds. On the one hand, they are encouraged to be empathic-to use their imagination and emotions to try and feel as well as they can how the parents, the child, the wider family members might be experiencing the situation. This plays a major part in the development of their understanding of the case and their planning of what needs to be done, as it should, and as several studies have reported does indeed happen (Munro, 1999;Bartelink et al., 2015). However, when it comes to writing up their work, especially when it is writing it up for an outsider such as a judge, they are expected to be 'objective'-that is, to seek to remove all trace of the worker as the observer and interpreter of the family. This attempt to remove or hide the subjective elements of practice suggests that they are seen as second-class, inferior to objective knowledge. Indeed, many of the reforms in recent decades have sought to eliminate or minimise the contribution of the individual worker. Performance indicators prescribing how long a task should take, assessment frameworks prescribing what information should be gathered and decision aids that calculate a conclusion on the information entered are all reducing the autonomy of the worker. There is a growing trend to define 'evidence-based practice' as practice that uses an evaluated method of intervention instead of its original meaning in evidence-based medicine and evidence-based social work where it referred to a practitioner drawing upon the best evidence from research as well as their clinical expertise and the user's preferences in deciding what to do, namely making an expert judgement (Sackett, 1996;Gambrill, 1999). This shift has the effect of downplaying the role of individual reasoning.
We are not opposed to such reform efforts in themselves but will argue that undertaking them using a simplistic notion of objectivity is harmful for the development of social work effectiveness. As an alternative, we offer a more nuanced approach that recognises the strengths and weaknesses in the range of our ways of reasoning.
Because of the authors' experience, the practice references in this article are to children and family social work, but the key arguments apply to all branches of social work.
We begin by exploring the meanings ascribed to the terms and offer the example of perception as a paradigm example of objectivity, contrasting this with a typical example of social work practice. This identifies the key attributes, both desirable and undesirable, associated with the terms 'objectivity' and 'subjectivity'.
We then examine how each attribute can or should be achieved in social work using three key issues to illustrate the details: the contribution of empirical research, dealing with dissent and the role of the personal. This is potentially a mammoth task so we only begin to sketch it out. The aim is to show that being more precise in what we want to achieve or avoid is more constructive than using one blanket, ill-defined notion to do many different jobs. It both allows more attention to strengthening reasoning that is typically seen as subjective and also reduces the problem of claims to objectivity obscuring implicit values, bias or use of power.

Defining objectivity and subjectivity
The concepts of objectivity and subjectivity are typically defined relative to each other: if something is not objective, then it is subjective. They have a surprisingly varied history, as shown in Daston and Galison's (2007) study. Although objectivity is now strongly associated with science, Daston and Gallison date its emergence as a scientific term as late as the mid-nineteenth century, long after the scientific revolution of the seventeenth century. It emerged in tandem with the concept of subjectivity and a variety of different meanings of objectivity have been contrasted with a variety of different meanings of subjectivity, not always one-to-one. Discussions of objectivity focus not just on what it offers, but what it guards us against, namely subjectivity. One example is 'to be objective is to aspire to knowledge that bears no trace of the knowerknowledge unmarked by prejudice or skill, fantasy or judgment, wishing or striving' (Daston and Galison, 2007, p. 17).
To clarify the meaning of 'objectivity' and to illustrate its appeal, we take as an example the proposition 'there is a table in this room'. We offer the example of sensory perception not just because it is simple, but because the paradigm of sensory perception seems to be central to the power and attraction of the concept of objectivity, and it helps us identify its desirable attributes.
There is a reliable process for establishing the truth of the proposition 'there is a table in this room'. It is possible to specify the rules to follow to reach the correct conclusion-use your eyes. The process is transparent-we can say how we checked the truth. It is impersonalanyone who understands the language of the proposition and has reasonable eyesight can check its truth. It is value-free-the observer's personal values or preferences do not influence their observation. If someone disagrees with the majority view, it is generally possible to find sound reasons to exclude their view: there is something wrong with their eyesight; they do not understand the language. Overall, there is a reliable, rule-following, impersonal, value-free and transparent way of establishing the truth and so there is convergence of opinion with means of legitimately discrediting the view of anyone who disagrees. We can feel confident that the proposition is true and others will agree with us.
These attributes of reliability, convergence, impersonality, value freedom, rule following and transparency seem to capture what is aspired to when objectivity is sought in social work. It is easy to understand their potential benefits. Reliable knowledge contributes to more accurate assessments and more effective services. Convergence and legitimate ways of excluding dissenting voices are reassuring when dealing with judgements and decisions with far-reaching consequences for service users: conclusions can be reached that are 'right' to everyone and not imposed by one person or group on another, such as by the powerful social worker on the service user. Impersonality makes service users less vulnerable to the biases, values or preferences of the particular social worker they are dealing with. Rule following makes it possible to teach others how to reach the correct conclusion. It also helps to make practice more transparent, allowing others to check the reasoning process. This is particularly valuable now when it has become a major requirement for professionals to account for their actions not just to service users, but also to society generally and central government funders in particular (Power, 1997).
To what extent does or can social work practice share these attributes of perception? We offer an example of a typical social work relationship with a family to explore this.
The scenario is a social worker working with a family where a child has suffered significant harm from a physical injury inflicted by the father and is now on a child-protection plan (i.e. services are working with the family with the goal of creating sufficient safety for the child to be able to stay at home). We give them all names to highlight the lack of impersonality in this scenario. The social worker is Jane, the mother Sharon, the father Phil and the one-year-old baby girl is Mia. Jane has developed a good relationship with Sharon, whom she likes, but a more distant one with Phil, whom she finds a little frightening. He claims that the injury occurred when he was chastising Mia for misbehaviour and he did not intend to cause such an injury-that is, it was accidental. Jane hopes that it will be possible for the child to stay with her parents. She uses her knowledge of abusive behaviour and strategies for changing behaviour as well as her empathic understanding of life in this family as she seeks to help them identify the problems that provoke Phil's anger and ways in which he could manage his anger less destructively and discipline Mia in less harmful ways.
This scenario exemplifies an approach to practice that does not score very well against the attributes of objectivity listed above. Neither research nor her empathic understanding provides 'reliable' understanding in the sense of being certain, especially about the father's motivation when he hit Mia. Her work is rule-following in a limited sense in that there are procedures for managing cases (Department for Education, 2015) but these are high-level outlines of how to proceed. Considerable autonomy is exercised by each worker in deciding how to work in detail. Jane is not interchangeable with any other social worker: Sharon and Phil have formed a relationship with her personally and another worker might have acted differently in many respects. The work is not valuefree: the law embodies the principle that children should be with their birth family if possible and Jane is influenced by her own view on the importance of the birth family in deciding how to weight the principle in this case. Some degree of transparency is required: her reasoning will be scrutinised as the case proceeds by family members, her supervisor, other professionals who know the family and ultimately by the courts if she decides the child is not safe at home and seeks power to remove her. But this transparency need not lead to convergence of opinion. Professionals and parents often disagree on what is best for the child. In contrast, when disagreements on perception arise, the transparency of the process shows how disagreement has arisen, and hence allows its resolution by seeing who went wrong and how, including the option of excluding one view because it does not conform with the protocols. But Jane's colleagues may just think that her empathetic account of Sharon is wrong. They may challenge it, and they may all come to agree after all. But that will not be the result of applying common professional standards. And, if they do not agree, there is no basis for excluding Jane's view on the grounds that it is subjective, lacking in objectivity. They just disagree. Transparency means that her decision has been explained, by setting out her reasons. Other people-her colleagues or family members-might well have reached a different conclusion for their different reasons.
This picture of practice is a long way from the paradigm of objectivity. Should we feel worried or ashamed by this and aspire to increasing the objectivity of practice, or is some of what subjectivity brings to the work both unavoidable and desirable?
Why is objectivity seen as desirable in social work?
While there have been numerous debates in the philosophy of science about the precise meaning of the terms 'objectivity' and 'subjectivity', two features are constant. First, claiming objectivity entails a claim about the accuracy or reliability of the knowledge and, in this context, subjectivity is seen as fallible. Second, claiming objectivity entails a claim that the listener/reader can trust the findings (Reiss and Sprenger, 2014).
Both of these claims carry weight in social work. Reliable knowledge can contribute to more effective services. The interest in evidence-based practice and the encouragement to monitor the effectiveness of interventions inherent in the English inspection framework both illustrate the drive to improve results.
Trust is also important, since society gives social workers the responsibility to intervene in the lives of individuals and families. When society feels little trust in them, then social workers find their judgements and decisions being questioned and they need to rely on the endorsement of other professionals (doctors, psychologists) when presenting a case in court. Also, when parents do not feel trust in the social workers' expertise, they may fear asking for help in case they are misunderstood. Increasing objectivity in practice is seen as one route for building trust. It is how this can or cannot be achieved that we examine in the following three sections.

The contribution of empirical research
Empirical research is seen by many as the primary source of objective knowledge. There is a growing tendency to equate 'evidence' with the findings of research, as in the phrase 'evidence-based policy and practice'. This is in contrast to the everyday use of the term, where 'evidence' refers to the range of information that is called upon to justify a conclusion. In this sense, a social worker draws on information obtained from numerous sources, such as direct observation, reports from others, past history, laws and procedures. We have two major concerns about current practices around using empirical research. First, its reliability, value freedom and the contribution it can make are frequently exaggerated. Second, it is being given high status while improving the reliability of other sources of evidence is being neglected.
On the first point, we agree wholeheartedly that research is valuable. Research methods aim to minimise the effects of bias so that we can have some confidence in the conclusions. However, whether in the natural or social sciences, they cannot claim to produce infallible knowledge. In some areas of research, this point can be seen as just a philosophical nicety, while the conclusions are treated in practice as 'true'. In research of relevance to social work, it is rare to find such consensus or robust conclusions. Littell's (2005) review of studies of the effectiveness of multi-systemic therapy illustrates the complexity of research of the social world. She reports that there were variations in the overall conclusion on whether it was effective or not and, within those reporting some effectiveness, there was no consistency in the variables that showed improvement. Such variability is not surprising when studying human behaviour, but it illustrates the point that we do not have a body of research that gives us universal laws but more nuanced indications that some variables tend to be influenced by some other factors on some occasions. This point indicates the need to treat research findings with caution but not to dismiss them-we are right to have some confidence in the findings.
The value freedom of research is also problematic. Values enter science at several points. They influence the choice of research problem and methods, what starting assumptions are allowed, what modelling strategies are used, what evidence is gathered, the acceptance of a theory or hypothesis as an adequate answer, and the proliferation and application of the results (Reiss and Sprenger, 2014).
The influence of values can be seen in social work research. While most would agree that human behaviour is influenced by both internal and external factors, values affect the choice of targets of research and of interventions. Over the decades, social workers have, to varying degrees, been interested in both social action to change the environment and individual work to change individuals and their contexts.
Values also influence the content of specific interventions. For example, the extent to which it is believed that the wider family and community should share in the task of raising children influences efforts to improve the care of children. Hopman et al. analyse the assumptions in the Triple P parenting programme, asking 'what kinds of values are being passed on to parents and children participating in these interventions?' (Hopman et al., 2014(Hopman et al., , p. 1358. They report that 'a sense of belonging' is not highlighted in Triple P and the approach does not highlight the value of the wider social network in resolving parenting issues. In contrast, the Signs of Safety approach to child welfare does emphasise the role of the wider family and the naturally occurring social networks as a crucial resource to improve the care of children (Turnell, 2012).
The difference is perhaps linked to the context in which the interventions were developed. While both originate in Australia, Signs of Safety was influenced strongly by traditional child-raising practices of Aboriginal culture while Triple P was more influenced by the white culture.
Research does not eradicate the personal. Researchers' socio-temporal location influences their perception of the social world and the way problems are framed. This has been well explored by feminist critiques of male research (Oakley, 2000) and within social work, where there are concerns about the class and ethnicity of researchers compared with those of the subjects of research (Rubin and Babbie, 2016).
Research is associated with the objective attribute of reliability but this is a desirable quality of all social work reasoning. It is an aspiration both for the information used and the way in which it is used in reasoning to an action. Ideally, the 'right' people get the 'right' service. Its relevance cuts across the standard objective/subjective distinction. We are concerned with the reliability of information used in reasoning, however it is collected-that is, by direct observation, by expert judgement, by reports from others or from research findings. An emphasis on seeking reliable knowledge through empirical research can distract focus from the importance of improving the reliability of information, however obtained.
In social work practice, the importance of checking information for accuracy is recognised with numerous ways of doing so, such as finding more examples of the relevant behaviour, obtaining the views of others and critically reflecting on your reasoning, usually with the help of a supervisor or colleagues. However, its importance is often overlooked when an agency is under pressure when workers are pushed to complete tasks quickly. Eradicating poor work conditions can be a far more effective way of improving accuracy than eradicating 'subjectivity'.
Expert judgement is also essential in engaging with families and in applying guidance and procedures to specific cases. Yet, again, work conditions can fail to nurture the development of more reliable expert judgement by reducing the continuity of contact so workers do not have the feedback to know whether their judgements were accurate or not.
Reliability can be threatened by the undesirable attributes often associated with subjectivity-bias, prejudice, self-interest, personal involvement and whimsicality. Unfortunately, these can operate at an unconscious level so they are hard to eradicate. Racial prejudice, for example, may be consciously considered abhorrent and yet studies reveal that it can influence professional practice unconsciously. In Cleaver and Freeman's (1995) study, workers were given vignettes and asked to assess the level of danger for the child. When the vignette included the additional information that the mother was West Indian, the danger was rated higher than when this information was omitted.
Moreover, the danger of the undesirable attributes associated with subjectivity still exists in research. Research methods aim to minimise the intrusion of these undesirable attributes but it is increasingly recognised that their success is partial. For example, medical journals now require authors to make a 'competing interests declaration' because, for example, funding sources might bias the reported findings. Evaluations of interventions conducted by those who have a commercial interest in the product are treated with more caution than those by independent researchers. Transparency in research is dependent upon comprehensive and honest accounts of the studies being available. Articles summarising research will contain selective material to limit the size, but fuller details should be made available in research reports or on individuals' web pages. There is a growing literature on how research findings can be incorrectly communicated or misunderstood (Sterne et al., 2001;Ioannidis, 2005;Mazar and Ariely, 2015). Greenberg (2009), for example, shows how citations of research can lead to major distortions in later references to that research: 'Through distortions in its social use that include bias, amplification, and invention, citation can be used to generate information cascades resulting in unfounded authority of claims' (p. 1).
One final point to make about research is that, although valuable, it can make only a limited contribution to the final decision on what action to take, since many other factors-case-specific, moral, financial or legal-may need to be considered. In addition, deciding whether a study is relevant to the problem you are dealing with is complex: 'You are often told that you will need to judge that your context is sufficiently like the one of the RCT. "Sufficiently like" is, however, a challenging judgement to make when working in the social world' (Munro et al., 2017, p. 7).

Dealing with dissent
A claim to objectivity can imply a claim that anyone who is dissenting has no valid grounds to do so. To return to the simple perception example, if someone denies that there is a table in the room, we can generally find a reason for dismissing their opinion, such as their having poor eyesight. Can social workers draw on 'objective knowledge' to dismiss dissenting views from service users?
When a worker and a parent conflict on what is best for a child, how is this to be resolved? An assessment of the parenting is crucial in the case scenario of Jane and the parents, and is central to child and family social work generally: it forms the basis of decisions about what services, if any, a family needs. The task carries major consequences for children, young people and their families-no help may be offered although wanted, services may be offered or, in extreme cases, legal proceedings Objectivity and Subjectivity 419 may be started with the aim of breaking up the family and removing the children.
When it comes to defining 'good' parenting, research has shown that there is considerable variation in parenting beliefs and practices linked to factors such as socio-economic status, religion, race and ethnicity (Garcia Coll et al., 1995). For example, spanking as a disciplinary method, co-sleeping and adult nudity in front of children are all seen as undesirable by some and culturally normal by others (Craig et al., 2000).
'Good' parenting can be defined as, among other things, parenting that contributes to helping the child develop into a 'good' person. Views on this vary too. Gender differences around the world are particularly noticeable, with some cultures wanting well-educated, independent women and others wanting well-behaved women who acknowledge the authority of the men in their lives. Governments' worries about poor parenting have varied over time. In England in 1900, the primary worry was about children's physical health as so many young men were found to be unfit to fight in the Boer War. In 1997, the new Labour government was clear that its main concern was 'education, education, education'.
There are differences also in how parents would balance the dimensions of a 'good' outcome when achieving on one goal seems to reduce the chances of achieving on another; for example, achieving good educational outcomes can at times have an adverse impact on mental health.
This seems an area where knowledge with all the attributes of objectivity would be useful but, even if available, it would not resolve the differences in goals.
In reality, although there is considerable diversity over time and between cultures, in any one society, there will be many occasions on which services users and social workers agree on ends but the disagreement is about how to achieve them. Jane and Mia's parents may agree that chastisement is necessary, but disagree on the best ways of doing so. In this particular scenario, Jane can take the simple option of saying the English law bans physical chastisement of the level practised by the father but she can also draw on research to challenge the view that it is the most effective means of disciplining a child. Research has a valuable role to play in giving weight to social workers' judgements, but it does not amount to a total knockout of any dissenting opinion.
Disagreements are hard to deal with. Discussions can illuminate the thinking behind the rival views and efforts to achieve consensus in social work can lead to agreement. However, there are times when social workers need to exercise authority in over-riding parents and justify it by referring to their expert knowledge. This requires them to have a good knowledge of the relevant research and to keep up to date with developments.

The role of the personal
Impersonality is valued in seeking objectivity because it is contrasted with people bringing personal baggage-values, feelings, preferences-to the reasoning process and thereby distorting it. However, social work is a human service and there are limits to how impersonal it can be. The personal enters in two main areas: in the relationships between workers and service users, and in making judgements.

The importance of relationships
Social work generally involves having some direct contact with services users and this involves some relationship. This is not an option, but an aspect of people talking to each other-we react to each other automatically and to a large degree unconsciously. In the case scenario, Jane is not just any social worker to the parents; she is a specific personality.
The Munro Review of Child Protection (2011) highlighted the way that many of the reforms that were designed to formalise practice more have reduced the individual contribution and so have ignored or undervalued the importance of the interactions between social workers and service users. As a consequence of the efforts to formalise practice, work conditions had been modified in many places to reduce continuity of contact, with service users being moved along a conveyor belt of tasks within the formalised process.
Organisational cultures have also changed in deleterious ways. Close involvement has two key consequences for organisations. First, it can introduce bias, as the worker's relationship with individual users influences their reasoning. For instance, feeling sympathy for a neglectful mother can lead to an underestimate of the harmful effect of the neglect on the child (Horwath and Tarr, 2015). Supervision and discussions with colleagues are ways in which this bias can be identified and rectified. However, the priority given to supervision is diminished in organisations that focus on meeting performance indicators and, even where it occurs, it can focus on process issues more than on substantive review of the reasoning about a case (Turner-Daly and Jack, 2014).
Second, close involvement has a psychological cost for social workers who are exposed to intense forms of human misery, anger and anxiety. The importance of organisational support was commonly accepted in the past but is less prioritised nowadays. Several studies have shown the extent to which social workers rely on conversations with colleagues to manage the cognitive and emotional dimensions of their work (Helm, 2013;Saltiel, 2016) but reforms such as hot desking can reduce the opportunities to do so.
Rather than seeing impersonality as the desirable goal, it would be more constructive to seek ways of reducing the bias that can be part of the necessarily personal relationship between social workers and service users.

Rule following and making judgements
Recent decades show a steady increase in the rules and guidance being provided to social workers, so it raises the question as to whether the ideal state is to turn social work into a wholly rule-bound activity so that whoever followed the rules would reach the same conclusion and take the same actions in relation to service users-that is, the contribution of the individual could be eradicated. Concerns for social justice and equity make this seem desirable. There is something disturbing about service users being at the mercy of the particular worker they meet. However, there is also something disturbing about thinking that a helping profession can be reduced to a set of rules.
This is an area where the cultural preference for explicit logical processes is, we think, having a damaging effect on the development of social work practice. It is damaging in that rule-bound services are failing to provide requisite variety (Ashby, 1991) to meet the range of needs of service users. Timescales for completing assessments, for example, are an over-rigid interpretation of the principle that timeliness matters and drift is harmful to children. One child may need urgent assessment so the timescale is too long and another might be in a very complicated situation where a satisfactory assessment cannot be achieved in the specified time. Judgement is needed to determine how to apply the principle of timeliness in any particular case.
The admiration for rules is also damaging by focusing reform efforts on creating more rules rather than focusing on how to help social workers make more accurate judgements. Yet this is an area where a growing body of research provides a wealth of ideas (Gigerenzer and Edwards, 2003;Klein, 2009;Kahneman, 2011). There is no algorithm leading to the 'right' answer, but there are better and worse ways of deliberating.
There is reason to be concerned about the variety of differences in social work judgements, such as about the level of risk a child is exposed to (Regehr et al., 2010;Bartelink et al., 2014;Fleming et al., 2014). However, the extent to which greater consensus can be achieved through objective measures is debatable. In some jurisdictions, actuarial riskassessment tools are being used to turn the process into a rule-following one. These typically claim to be objective-for example, 'The actuarial risk assessment can be described as an objective classification tool that estimates the likelihood of future harm' (Mendoza et al., 2016, p. 2). Yet, the construction of such tools involved several judgements about what data and methods to use. Their data-set is biased because it is drawn from administrative data of families known to child-protection services-a group where low-income and ethnic minority families are overrepresented. These may be referred more frequently for reasons unrelated to being more abusive or neglectful than the population in general. The point here is not to reject such approaches to improving risk assessment, but to question what exactly is claimed-and what can justifiably be claimed-by describing them as objective.
Ironically, such risk-assessment tools have a deleterious effect on another of the desired attributes of objectivity: transparency. A decision is reached where not even the social worker, and certainly not the family, is aware of how it was computed. Even though the rules to explain it are available, such an explanation would require extensive study and, in reality, it means workers are using the tool blindly and families do not understand the way in which major decisions are made about them.
Rules can be created in areas of work where they are otherwise not appropriate as a mechanism for deflecting blame. Hood et al. (2001) have analysed how the rise of a blame culture has led to more 'protocolisation', setting out a set of rules to follow so that the individual and organisation can defend themselves against any adverse outcome by saying 'we followed the protocol'. Such a defence has been used in social work when a child dies: 'procedures were followed.' The view that rules are the ideal and relying on judgement is secondbest misunderstands why rules are inappropriate in some cases: it is not a weakness in our current understanding, but a feature of the complex world in which we live. If social work is to provide a personal service, social workers need to be able to take account of unique features of the case rather than taking the procrustean approach of mutilating service users to fit their system. Moreover, the use of rules, even where appropriate, often relies first on the use of individual expertise. A social worker has to decide which rules apply in a given situation. In the inquiry into the care of Victoria Climbié (Laming, 2003), the social worker and her supervisor were criticised for not following procedure. But a closer reading of the text shows that the mistake they made was about how to classify the case: they managed it as one requiring family support not as a childprotection investigation. They failed to follow the right child-protection procedures because they had decided that they were irrelevant in this case. That decision was not rules-based, but resulted from a mistaken expert judgement.
The central principle is that rules need to be judged on whether they achieve their intended aim in terms of helping service users and the conclusion must be that they have only a limited role in the complex scenario of working with service users to solve problems.

Conclusion
Intuition and empathy are central to how humans reason about the world and how social workers seek to help those who use their services. Our analytic skill strengthens our understanding but does not replace their central role. Efforts to make social work more objective are seeking to create a bigger role for analytic reasoning. This is desirable, but the theme of this article has been to criticise current efforts both for over-claiming what analytic processes can achieve and under-valuing intuitive and empathic skills.
There are several sources of motivation for seeking to make social work use more explicit analytic reasoning, some of which are more praiseworthy than others. First, there is the desire to use our logical skills to strengthen our ability to offer effective help to service users. Second, some degree of standardisation looks desirable to reduce the element of chance in the service any user experiences, with social workers operating in very different ways. Third, there is an understandable but worrying need to manage the high level of anxiety in a service that deals with high-risk scenarios in a very punitive social climate. Providing a protocol for practice can reduce individual accountability and offer the defence that due process has been followed. And, finally, there is the prosaic factor that work within the technical rational paradigm that underpins the search for objectivity is good for a university career.
When the attributes of objectivity are explored, it becomes apparent that they vary in how desirable or feasible they are. Even when empirical research can be well evidenced and reliable, it is not alone sufficient to make decisions on what to do in social work. Conflicting values, political views and preferences can create situations of conflict between social workers and users. Some judgement has to be made about whose opinion carries most weight. In some scenarios (e.g. when a child is at risk of suffering significant harm), social workers will use their authority to decide that their expert opinion outweighs the family. In others, the users' views are paramount. Values permeate the work and processes that appear objective can be based on implicit values. Rule following is feasible with simple tasks but fails to handle the variety in the complex worlds of services users. Moreover, judgement is needed in deciding what rule applies in a particular situation. Some aspects of practice can be made transparent but can then mislead because it is such a limited picture of the work. And, finally, the impersonality associated with objectivity seems out of reach in the human activity of engaging with another to understand and help them with their problems, but strategies can reduce the risk of bias.
The overall conclusion from this exploration is that talk of 'objectivity' is highly ambiguous and misleading. The simplicity of the objective/subjective divide is distorting and detracts attention for the need to be as concerned about the presence of attributes such as bias, self-interest and unreliability in 'objective' research as in the practices typically labelled as subjective. Often, a claim to be objective reflects only one or two of the attributes such as when 'reliability' is claimed meaning 'inter-rater agreement of a small group of trained personnel'-not the attribute of accuracy that is associated in the common-sense meaning of reliable. Talk of objectivity can also mislead in that it refers to some layer of practice such as saying that performance indicators collect objective data but fails to discuss the values and theories that underpin the choice of indicators.
We conclude with two suggestions. Instead of talking of objectivity and subjectivity, we should be more precise and state which of their attributes we are referring to. And all of us need to be more cautious and critical in our use of knowledge. The solid appearance provided by objectivity needs to be replaced with a hesitant acceptance of knowledge claims as possibly the best available at present.