因果关系的判断

Research on causality judgments

目的:考察强化概率、强化比例和代价任务对因果关系判断的影响。方法:于2005-05选择西南大学四年级学生和研究生114人作为被试。按照Reed的实验范型,利用自编程序在电脑上进行,数据在电脑上自动收集。实行个别施测法。随机抽签法将被试分为6个组:①VR4任务无代价组19人,采用VR4(1～8)程序,只要求判断因果关系。②VR8任务无代价组20人,采用VR8(1～16)程序,只要求判断因果关系。③VR8任务低代价无分数呈现组18人,按1次键扣1分,三角形亮1次得10分,在进行因果关系判断后呈现得分。④VR8任务低代价有分数呈现组20人,按1次键扣1分,三角形亮1次得10分,在因果关系判断前呈现得分。⑤VR8任务高代价无分数呈现组19人,按1次键扣2分,三角形亮1次得10分,在进行因果关系判断之后呈现得分。⑥VR8任务高代价有分数呈现组18人,按1次键扣2分,三角形亮1次得10分,在进行因果关系判断前呈现得分。在实验过程中,要求被试按照电脑屏幕上给出指令进行按键操作和因果关系判断。当被试完成所有任务后,口头询问被试是否有发现三角形亮起来的规律以及作出判断的依据,作为考察其对按键与三角形亮起来之间因果关系的认知依据。组间比较采用t检验,F检验。结果:各组被试均完成实验任务,全部进入结果分析。①被试完成VR8强化程序作出的判断分数显著低于完成VR4强化程序作出的判断分,差异有显著性意义[(24.4500±20.2211),(55.8421±22.5887)分,t=4.578,P<0.05]。②VR4和VR8程序下做出高反应次数的7名被试的因果关系判断分数高于做出低反应次数的7名被试的判断分数,但差异无显著性意义[(44.7143±29.0787),(25.2857±24.1641)分,t=1.36,P>0.05]。③VR8程序下不同代价任务的5组被试在3min任务中作出反应的次数和任务结束时的因果关系判断的分数差异均无显著性差异。事后检验表明,无代价组因果关系判断的分数显著高于低代价有分数呈现组,差异有显著性意义(P<0.05)。结论:强化程序中的强化概率和强化比例会影响人们的因果关系判断。代价任务会影响因果关系判断,在无代价任务情况下人们的因果判断高于有代价任务情况下的因果判断。

AIM： To explore the effects of reinforcement probability, reinforcement rates and the cost of a response on human＇s causality judgments. METHODS： A total of 114 seniors and postgraduates from Southwest University were selected during May 2005 with individual measure method. A self-programmed computer procedure was used, and data was collected automatically, according to the experimental pattern of Reed. Participants were randomly distributed into 6 groups： ①No-cost VR4 （n=19）： VR4 （1-8） procedure was adopted to judge causality;②No-cost VR8 （n=20）： VR8 （1-16） procedure was adopted to judge causality;③Low-cost without mark VR8 （n=18） and low-cost with mark VR8 （n=20）： One-point deduction for one press on button, 10-point gain for one triangle blink, and the scores were presented after or before causality judgment;③High-cost without mark VR8 （n=19） and high-cost with mark VR8 （n=18）： Two-point deduction for one press on button, 10-point gain for one triangle blink, and the scores were presented after or before causality judgment. During experiment, all the subjects were required to conduct keying operation and causality judgment. And after completing all the tasks, they were inquired orally whether the triangle blinked in formula so as to investigate the causality between button pressing and triangle blink as the cognitive evidence. And t-test and F-test were used to compare between the groups. RESULTS： All the participants completed the tests and entered the result analysis. ①The scores on causality judgment were significantly lower under VR8 schedule than that under VR4 schedule [（24.450 0±20.221 1）, （55.842 1±22.588 7） points, t=4.578, P 〈 0.05].②The scores on causality judgment were higher in 7 participants with high response＇s number than those with low response＇s number, but the difference was not significant [（44.714 3±29.078 7）, （25.285 7±24.164 1） points, t=1.36, P 〉 0.05].③ There were insignificant differences in the response rates within 3 minutes task and the scores on causality judgment of the different costs under VR8 schedule. But after LSD check, the scores on causality judgment were significantly higher in no-cost groups than those in low-cost with mark groups （P 〈 0.05）. CONCLUSION： Reinforcement probability and reinforcement rates in schedules of reinforcement may affect human＇s causality judgments. The cost of a response may also influence human＇s causality judgments, and causal effectiveness judgment under condition requiring no cost of a response are higher than that under condition requiring a cost of a response.