Page |133 In the case of loneliness interventions, scientific integrity of the findings), all of the reviews essentially confirmed between 1970 and 2009 (to include and the findings of previous reviews that extend the time interval reviewed social skills training and group-based qualitatively in prior research), and had interventions can succeed in reducing to measure loneliness quantitatively. lgneliness. Is ius ce nets jastned, or Fifty-two intervention studies for is this a case in which prior conclusions lenelinees evel auc tnelusian ecitena have been perpetuated in the manner Th, © : 2 of ; ; . ese studies were divided into three Kuhn describes? . : categories based on the experimental To combat bias favoring results design used to assess the effects of the that confirm dominant theories, some intervention. Twelve studies used a scientists have argued that specific study single group pre-post design in which criteria should be met to warrant an loneliness among participants was evaluation of the effectiveness of the assessed at baseline and again after intervention (18). These criteria include exposure to the intervention. The single random assignment of study participants pre-post design is weak in terms of to receive the intervention, evidence that measuring the effectiveness of an the intervention is more effective than no intervention, however, because intervention, findings that are replicated individuals who have high scores on a by at least one independent research loneliness measure on one occasion are group, and results that are published in likely to score less extremely on a peer-reviewed journals. Previous second occasion even if no intervention reviewers of loneliness interventions had occurred. Said differently, people have, in fact, placed a premium on whose measurements suggest they are randomized trials that contrast a group very lonely at one point in time, on randomly selected to receive the average, appear to be less lonely when intervention