This paper challenges the notion that the social sciences require a fundamentally different methodology from the natural sciences due to the interactivity of human kinds. By examining the concept of classificatory looping and its impact on human kinds, the author argues that understanding the causal pathways and utilizing behavioral science can offer reliable generalizations about human kinds. The paper presents examples such as the Buraku of Japan and African Americans to demonstrate how behavioral science can be employed to predict changes in properties of social groups. It posits that the social sciences can operate in a manner similar to the natural sciences by examining generic human tendencies that hold broadly across diverse social contexts. This exploration ultimately supports the unity thesis, emphasizing that social sciences can indeed gain a scientific understanding of human kinds comparable to the knowledge offered by natural sciences. / Master of Arts / This paper explores the idea that social sciences, which study human behavior and societies, can use methods similar to those in natural sciences, which study the natural world. The challenge lies in the fact that human behavior can change based on how people are classified, making it difficult to establish reliable patterns. The author argues that by understanding the reasons behind these changes and applying insights from behavioral science, we can still make valid generalizations about human behavior. Real-life examples, such as the Buraku people of Japan and African Americans, are used to demonstrate how behavioral science can help predict changes in social groups. By focusing on common human tendencies that apply across different social contexts, the paper supports the idea that social sciences can gain valuable insights into human behavior, much like the natural sciences do.

Varje dag utförs riskbedömningar inom rättspsykiatri och kriminalvård. Syftet med denna uppsats är att undersöka det vetenskapliga underlaget för riskbedömningar. Uppsatsen inleds med en kort beskrivning av i vilka sammanhang riskbedömningar används och den traditionella uppfattningen av hur området och instrumenten har utvecklats. Sedan följer en genomgång av olika studier som behandlar bedömningarnas träffsäkerhet generellt samt mellan olika generationer av bedömningar. Därnäst redovisas och framförs kritik mot validiteten i Psychopathy Checklist Revised (PCL-R, Hare, 1991) och dess tillämpning inom riskbedömningar. Avslutningsvis analyseras psykopatibegreppet utifrån Ian Hackings teorier om ”human kinds” och ”loopingeffekter” och det förs en kort diskussion av tänkbara konsekvenser av och farhågor kring användningen av PCL-R och psykopatibegreppet generellt. / Every day, forensic risk assessments are carried out. The aim of this thesis was to evaluate their scientific grounds. Initially, a short description of the contexts in which risk assesments are carried out, as well as a description of the historical development of risk assesment instruments are provided. A summary of evaluations of different types of risk assessments and their predictive validity is presented, followed by a critique on the validity of the Psychopathy Checklist Revised (PCL-R, Hare, 1991) and its use in risk assesments. The psychopathy construct is analysed in relation to Ian Hacking’s theories of human kinds and looping effects and potential risks and consequences of the application of PCL-R and the psychopathy construct are discussed.