摘要
三个实验考察了不同心理距离(空间、社会、时间)和序列特征(趋势和发展速度)对判断预测中的趋势阻尼的影响。结果发现,当个体对人工时间序列做出预测时,会出现趋势阻尼;下降序列的阻尼大于上升序列,序列发展速度越快阻尼越大;不同心理距离对趋势阻尼的影响不一致,空间距离对趋势阻尼的影响不显著,为他人预测时的阻尼小于为自己预测,时间距离远时的阻尼大于时间距离近时。
Trend damping is one of the biases of judgmental forecasting, which means that when people forecast from time series with noises, they dampen data trends so that forecasts lie below upward trends but above downward trends. In other words, they underestimate the steepness of the series. Previous research focused mainly on the effect of the characteristics or presentation of trend damping, but rarely cared about situational factors such as psychological distance. The present research focuses on the effect of psychological distance and series characteristics （trend and slope） on trend damping in judgmental forecasting.
In terms of constmal level theory （CLT）, general trends convey a high-level, abstract rule about how the future will manifest itself, whereas deviation from trends represent a low-level, concrete exception to the rule. CLT-related research suggested that different psychological distances may show similar influences on construal and psychological consequence. We assumed that greater psychological distance may similarly lead to less trend damping, while less distance may lead to more trend damping.
Three experiments were designed to explore the effect of psychological distance and the series characteristics （trend and slope） on trend damping： Experiment 1 focused on spatial distance, Experiment 2 focused on social distance and Experiment 3 manipulated temporal distance. The materials were time series which were constructed using power-law functions of the general form： y=100＋300×（x/48）k, then added Gaussian noise with M=0, SD=10, where y represents the number of pixels above the x-axis, x being the dimension time, and k controlling the acceleration of the function. Downward trends were completely symmetric with upward ones, and the symmetry axis was y=250. The dependent variables were the D-value of predictive value and true value, true value minus predictive value for upward trend, predictive value minus true value for downward trend. When the dependent variable was 〉0, trend damping occurred. The larger the dependent variable, the larger the damping effect.
The results showed that： （1） Individuals tended to damp trends when forecasting. （2） The damping effect was greater when the trend was downward than upward, when the slope of time series was bigger. （3） The influences of different psychological distances on trend damping were different; though the damping effect was not different between different spatial distances, it was smaller when for others rather than for themselves and when for temporal near situations rather than temporal distal situations. The non-significant effect of spatial distance on trend damping may result from the fact that spatial distance does not affect the slope of prediction. As hypothesized, when forecasting for others rather than for themselves, individuals rely more on general trend, resulting in less damping effect. When forecasting for distant future, as the adaptation account of trend damping shows, individuals may rely more on the implicit negative acceleration mode, the representation of the developing trend of natural series, which leads to a larger damping effect. In other words, the implicit negative acceleration mode of trend may be more abstract than the global trend, and when forecasting for distant future, individuals rely more on it.
出处
《心理科学》
CSSCI
CSCD
北大核心
2016年第1期28-35,共8页
Journal of Psychological Science
基金
山东省自然科学基金项目(ZR2015CM026)的资助
关键词
判断预测
趋势阻尼
心理距离
适应理论
judgmental forecasting, trend damping, psychological distance, adaptation account
