Causal and Prescriptive Statements in Published Social Work Research Articles

Purpose:

Social work researchers are frequently interested in identifying and recommending effective practices and policies to improve outcomes for vulnerable populations. Treatment recommendations should require that some type of causal connection between the intervention and outcome be empirically established, and most agree that controlled researcher manipulation of variables is an essential step to establishing causation. Prior research in other fields, however, indicates that researchers draw causal conclusions and make treatment recommendations based on non-intervention research methodologies that do not meet criteria for establishing causality. The purpose of this study was to examine the extent to which causal and prescriptive statements are expressed in non-intervention research articles published in social work journals. The specific research questions guiding this study were: 1) What are the proportion of non-intervention empirical studies in which causal or prescriptive language is used? 2) Has the use of causal or prescriptive statements increased over the past decade?

Method:

The sample included all articles that reported empirical, non-intervention studies published in 2001 and 2011 in five social work journals: Social Work Research, Social Work, Research on Social Work Practice, Journal of Social Service Research, and Social Service Review. All articles were independently screened for inclusion by two trained coders. Once the sample was selected, two coders independently searched the abstract and discussion sections of each article for sentences where the author(s) used causal language and/or made recommendations for social work practice (i.e., prescriptive statements). A third reviewer reviewed discrepancies in coding. The proportion of causal and prescriptive statements and change between the two time points was calculated within and across all journals.

Results:

Of the 164 empirical, non-intervention studies included, causal language was used in 73% of the articles in 2001 and 72% of the articles in 2011. While the proportion of causal statements remained similar across years, a slight increase in prescriptive statements, from 57% in 2001 to 62% in 2011, was observed. Variation in proportion and trends in the use of causal and prescriptive statements varied across journals.

Conclusions and Implications:

The use of causal and prescriptive statements in the reporting of non-intervention studies is prevalent in social work research and the use of prescriptive statements has increased slightly in the last ten years. Moreover, it is likely that our findings are underestimated as we included nonrandomized and within group intervention studies in the “appropriate causal language” intervention research category; these types of studies have low internal validity and the use of causal language is more likely to be inappropriate. The use of causal and prescriptive statements by social work researchers based on research methodologies that do not warrant causal conclusions are at best misleading and, at worst, can be dangerous and lead to ineffective or inappropriate interventions being adopted. Findings indicate the need for more vigilance on the part of authors, peer reviewers and editors to ensure that appropriate conclusions and recommendations are being made in research articles.