The Notion of Causal Impact in Evidence-Based Social Work: An Introduction to the Special Issue on What Works?

Causality and Causal Inference in Social Work

“…The first was whether it is desirable for social workers to identify cause and effect. It is desirable if we believe social work to be an applied, empirically grounded social and cultural science aiming at both causal explanation and interpretative understanding (Otto & Ziegler, 2008, p. 273), one that includes elements of logical positivism and postmodernist social constructivism. It is also desirable if the foundation of our profession is to change the lives of our clients for the better.…”

“…The attempt to examine effects of childhood sexual abuse using a phenomenological approach is especially noteworthy because the focus on interpretative understanding or verstehen is often seen as a rejection of causal understanding (cf. Otto & Ziegler, 2008). …”

“…Otto and Ziegler (2008) note that RCTs are considered the best form of evidence of practice effectiveness (McNeece & Thyer, 2004) and, therefore, of causality. “These designs serve to control or cancel out and differences that are effects of other Events (Z) to assess whether Event X (cause)—as independent variable—is nonspuriously conjunct with Event Y (effect) in the context of a controlled ceteris paribus condition” (Otto & Ziegler, 2008, p. 275).…”

“…Otto and Ziegler (2008) note that RCTs are considered the best form of evidence of practice effectiveness (McNeece & Thyer, 2004) and, therefore, of causality. “These designs serve to control or cancel out and differences that are effects of other Events (Z) to assess whether Event X (cause)—as independent variable—is nonspuriously conjunct with Event Y (effect) in the context of a controlled ceteris paribus condition” (Otto & Ziegler, 2008, p. 275). They further argue that the criteria of using the RCT design to determine causal connections between an intervention and its outcomes can hardly be applied to qualitative research such as ethnographic studies or deep hermeneutical interviews (Otto & Ziegler, 2008, p. 275).…”

“…“These designs serve to control or cancel out and differences that are effects of other Events (Z) to assess whether Event X (cause)—as independent variable—is nonspuriously conjunct with Event Y (effect) in the context of a controlled ceteris paribus condition” (Otto & Ziegler, 2008, p. 275). They further argue that the criteria of using the RCT design to determine causal connections between an intervention and its outcomes can hardly be applied to qualitative research such as ethnographic studies or deep hermeneutical interviews (Otto & Ziegler, 2008, p. 275). Consequently, qualitative studies are placed on a lower rank of evidence of causality (McNeece & Thyer, 2004), and below what Cook and Campbell (1979) considered as the minimum interpretable design necessary and adequate for drawing valid conclusions about the effectiveness of treatments (Otto & Ziegler, 2008, p. 275).…”