An Experimental Template for Case Study Research

Methods are usually classified as either "experimental" or "observational," a dichotomy that has governed social science research for centuries. By implication, this dichotomization precludes a consideration of experimental strategies in case study work. Yet, we argue that one gains purchase on the tasks of research design by integrating the criteria traditionally applied to experimental work to all research in the social sciences—including case study work, the focus of this article. Experimental research designs aim to achieve variation through time and across space while maintaining ceteris paribus assumptions, thus maximizing leverage into the fundamental problem of causal inference. We propose to capture these multiple criteria in a four-fold typology: (1) A Dynamic comparison mirrors laboratory experimentation through the use of both temporal and spatial variation; (2) A Longitudinal comparison employs temporal variation; (3) A Spatial comparison exploits variation through space; and (4) A Counterfactual comparison relies on imagined comparison. All comparison case study research designs can be slotted into one of these four categories. Moreover, the typology illustrates in a concise fashion the ways in which case study research designs attempt to mimic the virtues of experimental design and the degree to which they succeed. The classic experiment, with manipulated treatment and randomized control, thus provides a useful template for discussion about methodological issues in experimental and observational contexts.     

The review panel process: an algorithm for the conditional release of insanity acquittees

The release of insanity acquittees requires making informed decisions regarding both the presence and severity of an individuals' mental illness and the dangerousness of these individuals. This study evaluated the usefulness of employing structured assessments of mental health and violence risk factors in the conditional release decision-making process. All persons found Not Guilty by Reason of Insanity at East Louisiana Mental Health System, Forensic Division who underwent a review panel between July 1, 1997 and July 1, 1999 were included in this study. The Classification and Regression Tree analysis was utilized to arrive at cutpoints that would optimize the predictive ability of the decision tree analysis. The results indicated that the Community Outpatient Treatment Readiness Profile score was the strongest predictor -- all patients receiving a score of 62 or greater on this scale were recommended to remain at the facility. When women were recommended for release, it was to civil facilities and with moderate levels of symptoms. For males with moderate symptoms, low PCL-R scores were associated with recommendations for release, whereas high scores were associated with recommendations for continued commitment. Our data suggests that algorithms may be useful to governing bodies when making release decisions.     

A cluster analysis of people on Community Treatment Orders in Victoria, Australia

This paper explores the clinical, social and demographic characteristics of 164 people on Community Treatment Orders (CTOs) in one area mental health service in Victoria, Australia. The results of an exploratory cluster analysis are presented to address the question of whether people on Community Treatment Orders can be categorised into statistically reliable, qualitatively distinct groupings. The data are presented in the context of key stakeholder perspectives on the current use and purpose of CTOs. Three stable clusters emerged and each potentially reflects how social dimensions, as well as clinical issues, influence decision making regarding the implementation of CTOs. These findings are important in the context of policy and practice in Victoria, where the use of CTOs is common practice, and orders are generally made for a 12 month period. The potential for improved targeting of CTOs and more specific treatment planning is identified.     

Clothing damage analysis in alleged sexual assaults--the need for a systematic approach

Clothing damage analysis is an integral part of the examinations carried out in sexual assault type cases. This analysis can be used to corroborate different versions of events and is at its most powerful in elucidating false allegation cases and consent cases. The purpose of this study was to determine to what extent people with varying levels of forensic awareness, experience and training could correctly carry out damage analysis. Two participant groups were asked to take part in this study. Group A ('forensic group') comprised participants at a forensic science conference, and Group B ('student group') comprised students undertaking a degree course in Forensic Science. Each group was given a practical workshop consisting of a lecture outlining common fabric types and general features observed in different damage types. Each participant was subsequently shown 25 pieces of 'damage' and asked to identify both the type of fabric construction (knit or weave) and the type of damage (cut, tear, rip, wear and tear). The ability to identify fabric construction and damage types varied within the two groups studied and across the groups. The forensic group performed better both in fabric and damage assessment than the student group. This paper suggests a systematic approach to clothing damage analysis to maximise the benefits that can be obtained from this area of forensic science and to minimise the subjectivity within the field.     

The transfer of touch DNA from hands to glass, fabric and wood

The transfer of DNA from hands to objects by holding or touching has been examined in the past. The main purpose of this study was to examine the variation in the amount of DNA transferred from hands to glass, fabric and wood. The study involved 300 volunteers (100 for glass, 100 for fabric and 100 for wood) 50% of which were male and 50% female. The volunteers held the material for 60 s. The DNA was recovered from the objects using a minitape lift, quantified using the Quantifiler kit assay, extracted using a ‘Qiagen^® QIAamp DNA mini kit’ and amplified using the AmpFlSTR^® SGM Plus™ Amplification Kit at 28 cycles. The results show that using ANOVA there was a significant difference (F = 8.2, p < 0.05) between the three object types in the amount of DNA recovered. In terms of DNA transfer and recovery, wood gave the best yield, followed by fabric and then glass. The likelihood of success of obtaining a profile indicative of the holder was approximately 9% for glass samples, 23% for fabric and 36% for wood. There was no significant difference between the amount of DNA transferred by male or female volunteers. In this study good shedder status, as defined by obtaining useful profiles of 6 or more alleles, is estimated at approximately 22% of the population. The phenomenon of secondary transfer was observed when mixed DNA profiles were obtained but the incidence was low at approximately 10% of the total number of samples. DNA profiles corresponding to more than one person were found on objects which had been touched by only one volunteer. Although secondary transfer is possible the profiles obtained from touched objects are more likely to be as a result of primary transfer rather than a secondary source.     

Short-Term Forces and Partisanship

One of the most intriguing aspects of the debate regarding the persistence of party identification is that proponents of different schools of thought have each managed to use the same quasi-experimental data and similar state of the art techniques to defend their point of view. In this article we argue that this debate cannot be resolved with quasi-experimental data alone and propose another method that we believe can help us triangulate in on the correct answer: experimentation. Two experiments are performed and analyzed. The first tests the hypothesis that party identification is updated in response to the vote choice; the second tests the hypothesis that candidate evaluations influence party choices. The results of our experiments provide some additional support for the traditional conception of partisanship as the unmoved mover of American politics.