框架与序列趋势对趋势阻尼的影响研究

发布时间：2018-04-22 12:09

本文选题：判断预测 + 趋势阻尼　；参考：《山东师范大学》2016年硕士论文

【摘要】：预测方法有两类:统计预测和判断预测。与统计预测相比,判断预测往往带有主观性,因此容易出现偏差,其中趋势阻尼最为常见。趋势阻尼指人们在预测有噪音的序列时,往往低估下降序列的趋势,而高估上升序列的趋势,即低估序列的陡峭性。已有研究探讨了序列趋势对趋势阻尼的影响,发现下降序列的阻尼大于上升序列。但是,已有研究采用的实验材料多是越多越好的数据,人们可能将上升序列感知为收益,而将下降序列感知为损失。也即是已有研究可能混淆了收益—损失框架对趋势阻尼的影响。而研究表明框架会对趋势阻尼产生影响。据此,本研究通过两个实验,系统探讨框架和序列趋势对趋势阻尼的影响及原因。实验1选取可控事件(公司盈亏)探讨框架和序列趋势对趋势阻尼的影响。对这种影响的原因存在两种可能的解释:乐观偏见和扭转期望。根据这两种解释,本研究假设:在收益框架下,下降序列的阻尼会显著大于上升序列,但在损失框架下,上升序列的阻尼显著大于下降序列。为了验证哪种解释更为合理,实验2选取不可控事件(股票波动)探讨框架和序列趋势对趋势阻尼的影响,若结果与实验1一致,则乐观偏见更具解释力,若结果不一致,则扭转期望的解释更为合理。两个实验的预测均采用点预测,即让被试根据序列中前48个点,预测后6个点的走向。两个实验中的因变量均为预测值与真值之间的偏差。研究结果发现:(1)序列发展速度的主效应显著,序列发展速度越快,趋势阻尼越大;(2)可控事件中收益—损失框架与序列趋势的交互作用显著,收益框架下,下降序列的阻尼显著大于上升序列;而在损失框架下,上升序列和下降序列的阻尼没有显著差异;(3)不可控事件中收益—损失框架与序列趋势的交互作用显著,收益框架下,下降序列的阻尼显著大于上升序列;而在损失框架下,二者之间的差异不显著。可见,乐观偏见能够更好地解释框架和序列趋势对阻尼影响。
[Abstract]:There are two kinds of forecasting methods: statistical forecasting and judging forecasting. Compared with statistical prediction, judgment prediction is often subjective, so it is easy to deviate, among which trend damping is the most common. Trend damping means that when people predict noisy sequences, they tend to underestimate the trend of descending sequences, and overestimate the tendency of ascending sequences, that is, to underestimate the steepness of sequences. The influence of sequence trend on trend damping has been studied, and it is found that the damping of descending sequence is greater than that of ascending sequence. However, most of the experimental materials used in the previous studies are the better data. People may perceive the ascending sequence as the gain and the descending sequence as the loss. That is to say, previous studies may confuse the effect of benefit-loss framework on trend damping. The research shows that the frame has an effect on the trend damping. Based on these two experiments, the influence of frame and sequence trend on trend damping and its causes are discussed systematically. Experiment 1 selects controllable events (corporate profit and loss) to explore the influence of frame and sequence trend on trend damping. There are two possible explanations for this effect: optimistic bias and reversing expectations. According to these two explanations, this study assumes that the damping of the descending sequence is significantly greater than that of the ascending sequence under the income framework, but the damping of the ascending sequence is significantly greater than that of the descending sequence under the loss framework. In order to verify which explanation is more reasonable, experiment 2 selects uncontrollable events (stock volatility) to study the influence of frame and sequence trend on trend damping. If the result is consistent with experiment 1, the optimistic bias is more explanatory, and if the result is not consistent, The explanation for reversing expectations is more reasonable. The two experiments used point prediction, that is, the subjects were asked to predict the trend of the last 6 points according to the first 48 points in the sequence. The dependent variables in both experiments are deviations between the predicted and true values. The results show that the main effect of the speed of sequence development is significant, and the faster the speed of sequence development, the greater the trend damping. (2) the interaction between the pay-loss framework and the trend of the sequence in controllable events is significant, under the benefit framework, The damping of descending sequence is significantly larger than that of ascending sequence, but there is no significant difference between the damping of ascending sequence and descending sequence in the frame of loss. The damping of descending sequence is significantly higher than that of ascending sequence, but the difference between them is not significant under the frame of loss. Therefore, optimistic bias can better explain the influence of frame and sequence trend on damping.
【学位授予单位】：山东师范大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：B848

【参考文献】