Marsh H W, Roche L A
Faculty of Education, University of Western Sydney, Macarthur, Australia.
Res Q Exerc Sport. 1996 Mar;67(1):13-23. doi: 10.1080/02701367.1996.10607921.
Silhouette ratings of actual and ideal body fat (N = 258 high school students) were related to 4 self-concept scales (Body Fat, Physical Appearance, General Physical, Esteem) and objective body composition measures in order to test predictions from actual-ideal discrepancy models. Actual-ideal discrepancy scores were more strongly related to self-concept than actual scores alone, thus supporting the traditional discrepancy model. However, multiple regression analyses demonstrated that more sophisticated actual-ideal discrepancy models that considered discrepancies in either direction (feeling too fat or too thin) did significantly better. Because the traditional discrepancy model is a special case of this more general discrepancy model, the more general model should have broad applicability. Important aspects of this task were (a) that discrepancies were evaluated along a continuum in which the ideal was not one of the endpoints of the continuum, (b) that there was reasonable variation in ideal ratings, and (c) that positive and negative discrepancies were reported.
实际和理想体脂的轮廓评分（N = 258名高中生）与4种自我概念量表（体脂、外貌、总体身体、自尊）以及客观身体成分测量相关，以检验实际-理想差异模型的预测。实际-理想差异得分与自我概念的相关性比单独的实际得分更强，从而支持了传统差异模型。然而，多元回归分析表明，考虑到两个方向（感觉太胖或太瘦）差异的更复杂的实际-理想差异模型表现明显更好。由于传统差异模型是这个更通用差异模型的一个特例，所以更通用的模型应该具有广泛的适用性。这项任务的重要方面包括：（a）差异是沿着一个连续体进行评估的，其中理想状态不是连续体的端点之一；（b）理想评分存在合理的变化；（c）报告了正向和负向差异。
