 |
 |
 |
| Saturday, 15 January 2005: 8:00 AM-9:45 AM | |||
| Hibiscus B (Hyatt Regency Miami) | |||
| Causal Analysis With Or Without A Control Group: Using Binary Logistic Regression To Investigate What Interventions Work And In What Circumstances | |||
| Roundtable/Workshop Submitter(s)s: | Mansoor A. F. Kazi, PhD, University of Huddersfield John Varlow, MSc, Calderdale & Huddersfield NHS Trust | ||
| Format: | Workshop | ||
| Abstract Text: Purpose: The workshop will illustrate a new approach in causal analysis that can be used with or without a control group. Drawn from epidemiology, the forward-conditional binary logistic regression method can be used to determine what interventions work and in what circumstances. To date, empirical practice approaches have tended to focus on outcomes to evaluate the effectiveness of human service interventions. Randomised controlled trials (RCTs) help to identify interventions that are effective and to establish a causal connection between the intervention and its effects. However, in controlling for extraneous variables, RCTs tend not to explain why the intervention may not have worked with a minority of service users. Binary logistic regression models help to deal with this limitation in investigating the conditions in which an intervention is likely to be effective, even where such random control groups are not possible. First, the sample is divided into two groups---those that improved against a desired outcome, and those that did not. Second, the variables that may be influencing the outcome are identified through bivariate analysis, other research findings, and/or practice wisdom. Third, these selected variables are entered in a forward-conditional model. The variables that are actually influencing the outcome are retained in the equation, and those that are significant provide an exponential beta which is interpreted as an odds ratio, indicating the odds of the intervention achieving the outcome where the significant factor(s) may be present. Pedagogical techniques. First, the methodology will be introduced, and then actual databases from evaluations of social work programmes will be used to demonstrate the utility of the methodology in investigating the circumstances where particular interventions are more likely to be effective. The audience will be invited to take part in an interactive process of data analysis and discourse using real databases from evaluations of social work interventions in a variety of service-user settings. | |||
See more of Workshop
See more of Celebrating a Decade of SSWR (January 13 - 16, 2005)