Book reviews

TOATES, F. Animal Behaviour: A Systems Approach. Chichester: John Wiley and Sons. 1980. Pp. 299. Hardback £17. Paper back £6.40 ISBN 0 471 27724.

DICKINSON, A. Contemporary Animal Learning Theory. Cambridge: University Press. 1980. Pp. 177. Hardback £12.50. Paperback £3.95. ISBN 0 521 23469 7.     

Attributions of Responsibility for Obesity: Narrative Communication Reduces Reactive Counterarguing Among Liberals

Informed by theory and research on attributions and narrative persuasion, we compared the effectiveness of narrative and nonnarrative messages in changing attributions of responsibility for causes and solutions related to obesity in the United States. We randomly assigned 500 adults to view one of three messages (narrative, evidence, and a hybrid of the two) emphasizing environmental causes of obesity, or a no‐exposure control condition. The narrative condition increased the belief that societal actors (government, employers) are responsible for addressing obesity, but only among liberals. This pattern of results was partially explained by the success of the narrative condition in reducing reactive counterarguing, relative to the evidence condition, among liberals. We conclude with theoretical and practical implications.     

The Convergent Validity of Four Burnout Measures in a Chinese Sample: A Confirmatory Factor‐Analytic Approach

The purpose of this paper is to study the convergent validity of the four most widely used burnout measures in a sample of Chinese nurses (N= 717). First, Structural Equation Modeling was used to investigate the factor structure of scores produced by the Maslach Burnout Inventory‐General Survey (MBI‐GS), the Burnout Measure (BM), the Shirom‐Melamed Burnout Measure (SMBM), and the Oldenburg Burnout Inventory (OLBI). Next, several competing models were tested to investigate the convergent validity of these four burnout instruments. The final results suggest that burnout is best conceived of as a multidimensional construct consisting of exhaustion and withdrawal, which are two related but conceptually distinct aspects. In addition, positively phrased items should be dropped from burnout measures for they constitute a separate factor that is considered to be an artifact.