Use of Causal Language in Studies on the Relationship between Spiritually-Based Treatments and Substance Abuse and Relapse Prevention

Abstract

The main goal of scientific research is to explain what causes a phenomenon. However, only well-controlled studies guarantee sufficient internal validity to support causal explanations (i.e., experimental and some quasi-experimental designs). The use of causal claims in non-experimental studies can mislead readers into assuming a cause–effect relationship when alternative explanations have not been ruled out, undermining the principle of scientific rigor and the credibility of scientific findings. Although spiritual practices form part of some interventions for health and behavioral problems, their effectiveness cannot often be assessed via experimental methodology. This paper assesses the validity of causal inferences in published non-experimental studies, and more specifically in studies on the relationship between spiritually based treatments and substance abuse improvement and relapse prevention. We conducted a systematic review using Scopus, Pubmed, and several databases included in ProQuest, for the period 2015 to 2020. Out of 16 studies selected, six studies (37.5%) used correct language in the title, abstract, and discussion sections; 10 studies (68.8%) used tendentious or incorrect language in at least one section. Spiritually based treatments show promising results in some health improvement outcomes. Most studies show transparency when reporting results. However, researchers should be careful not to make causal assertions unless the internal validity of the research is sound.

The “Why” in Mental Health, Stigma, and Addictive Behaviors: Causal Inferences in Applied Settings

“…In applied areas where experimentation is not feasible or ethical, correlational studies are used to accumulate scientific knowledge from evidence. However, inadequate causal inferences from correlational data appear in various fields, such as nursing [8], epidemiology [9], and nutrition [10]; in substance abuse treatment [11] and antidepressant treatments [7]; or in the effect of cannabis on the development of psychosis [12]. In these studies, authors often publish conclusions using causal language in the absence of experimental methodology.…”