Editorial

“So what harm would be done if evaluators ignored values? In one sense, no harm would be done, because evaluations would still have values implicit in them. Values inevitably permeate the selection of independent and dependent variables, the choice of questions and stakeholders, and the social and political context from which many evaluations arise. Evaluators cannot avoid values even if they try. But in another sense, real harm is done if evaluators deal with values naively or poorly through their implicit choices.” Shadish, 1994, p. 35

The evaluation community is increasingly talking about values in evaluation, and the need to be more explicit about values. In these two special issues of the Evaluation Journal of Australasia, we aim to extend the discussion on values in evaluation and recognise the challenges of dealing with values. The papers in these issues show some of the choices evaluators have made, both in how they conceptualise values and how they engage with values in evaluation, in very different contexts. In this issue, the paper by Caitlin Blaser Mapitsa, Aisha Ali and Linda Khumalo considers the role of social and political values in establishing national evaluation systems. In the praxis papers, three authors reflect on their experiences working in a variety of evaluation contexts as they engage with stakeholder values. David Week discusses how multiple values perspectives can create insights and encourages evaluators to be open to seeing other cultures and viewpoints as valid. Robyn Thomas Pitts focuses on the extent to which responsiveness should be centred as a key value for evaluators. Aileen Reid contemplates negotiating and surfacing stakeholder values in a context in which clients may not understand the benefits of engaging an evaluator.

The papers in this special issue reflect ongoing discussions on the role of values in evaluation. Some evaluation theorists have been writing about values since the 1970s, encouraging evaluators to pay attention to values, and to consider the questions of ‘whose values’ and ‘which values’ should be prioritised in evaluation (Greene, 1997; House & Howe, 1999; MacDonald, 1974; Mertens, 2009; Schwandt, 1989; Stake et al., 1997).

One question that appears to be missing from the discussion on values in evaluation is – what do we mean when we talk about values? The word ‘values’ is used to label a broad range of concepts such as attitudes, needs, social norms, personal goals, opinions, or personality traits (Hitlin & Piliavin, 2004; House & Howe, 1999; Sagiv & Roccas, 2017). Having many different and unarticulated concepts of values can make it hard to discuss values, even within the same professional field (Pattison & Pill, 2004). This problem is not unique to evaluation. Research has found a wide range of meanings associated with the word ‘values’ in fields such as health care (Fulford, 2011; Giacomini, Hurley, Gold, Smith, & Abelson, 2001), and public policy (Fischer & Gottweis, 2012; Marchetti & Marchetti, 2016).

The divergent concepts of values across disciplines reflect the broader history of Western social science. In the mid-20th century, there was a widespread belief that social science should be value-free (Douglas, 2009; House & Howe, 1999). The ideal of value-free social science rested on the fact/value dichotomy, the idea that facts and values were different types of concepts and should be kept separate (Gorski, 2017; Putnam, 2002; Scriven, 1972). The fact/value dichotomy in turn rested on philosophers’ arguments that values were personal, subjective, emotional and therefore irrational (Foot, 1995; Marchetti & Marchetti, 2016; Putnam, 2002). As the evaluation field developed through the twentieth century, it was influenced by these broader trends in social science (Schwandt, 1997; Shadish, Cook, & Leviton, 1991). Many evaluators were therefore taught that values are a problem for evaluation, without being aware of the philosophical arguments that underpinned the ideal of value-free social science (for a detailed critique see Scriven, 2012).

Over the last fifty years, assumptions about values as subjective, personal, emotional and irrational have been challenged and refuted, based on research in social psychology and philosophy. Philosophers now recognise that facts, values, subjectivity, rationality, emotion and objectivity are entangled in our social world and cannot be easily untangled. As a result, the ideal of value-free social science is widely rejected (Douglas, 2009; Elliott, 2017b; Longino, 1990; Putnam, 2002; Scriven, 1972, 2012). This transition away from value-free social science has important implications for evaluation (House & Howe, 1999; Schwandt, 2002). Many evaluators now reject the value-free ideal and seek to acknowledge values in evaluation (Abma, 2005; Davidson, 2015; Greene, 2015). There are many choices to be made about how to conduct evaluation following the collapse of the fact/value dichotomy (House & Howe, 1999).

An important first choice for dealing with values in evaluation may be to choose a suitable concept of values, one that reflects the current research on values and the transition from the concepts that underpinned the fact/value dichotomy. Some literature in social science and evaluation uses the term ‘values’ interchangeably with ‘what people value’ or ‘value judgements’ – “what is right, better or ought to be” (Scriven, 1966). For example, one textbook gives the following example of a value, that “government should provide individuals with a decent standard of living in retirement” (Mark, Henry, & Julnes, 2000, p. 41). In recent decades, research in social psychology and philosophy has moved to understanding values as a system for meaning-making, a set of abstract motivational guides that influence how people interpret the world and make choices about action (Hitlin & Piliavin, 2004; Rescher, 2017; Sagiv & Roccas, 2017). Therefore, to update the previous example, the idea that “government should provide individuals with a decent standard of living in retirement” could be described as a policy priority built on the values of compassion, equality and justice. This can be seen as a shift in the meaning of the term ‘values’ – from conceptualising values as advocacy of valued ends, and toward values as the motivational reasons underlying the positions that are advocated.

Contemporary research has also identified many kinds of values. Each kind of values is about some aspect of our social world and how we decide what is good ¹ (Sadler, 1997; Sayer, 2011).