CAJ | 학술논문

在数字自动加工研究中,已有研究的实验对象一般局限在小数量数字上,大数量数字加工中是否也存在自动激活现象尚未获得实验证据。本研究以大数量为研究对象,以Stroop效应和SNARC效应为自动加工的指标,实验1发现,无论是数字语义比较还是数字个数比较,都出现了典型的数字Stroop效应;实验2的结果表明,在大数量个数的比较任务中存在SNARC效应和顺序效应。由实验结果初步推断,在大数量加工中也存在无关维度数量信息的自动激活。
재수자자동가공연구중,이유연구적실험대상일반국한재소수량수자상,대수량수자가공중시부야존재자동격활현상상미획득실험증거。본연구이대수량위연구대상,이Stroop효응화SNARC효응위자동가공적지표,실험1발현,무론시수자어의비교환시수자개수비교,도출현료전형적수자Stroop효응;실험2적결과표명,재대수량개수적비교임무중존재SNARC효응화순서효응。유실험결과초보추단,재대수량가공중야존재무관유도수량신식적자동격활。
In recent years, in number cognitive domain numerical Stroop paradigm was used to study the relationship of numerical semantic and numerical physical size in number process. In addition to Stroop effect, SNARC effect was also an important index of number automatic processing. Based on existing fi ndings, this study conducted two experiments to explore whether large numbers could be activated automatically. By using Stroop effect as index, the fi rst experiment explored whether the number of numeral and number semantic could produce interference and facilitation; the second experiment explored whether the number（i.e., quantity of asterisks） represented spacially by using SNARC effect as index. Experiment 1 was divided into two parts. In fi rst part subjects were asked to compare the numerical semantic of standard stimulus with that of comparing stimuli, but to ignore the numbers of numeral. The second part took the reverse procedure. The results indicated that the main effect of task type was not signifi cant, F（1, 19） =.030, p〉.01, the main effect of Stroop effect（ between accord and disaccord） was signifi cant, F（1, 19） = 83.838, p 〈.001. The results suggested that the response speed of numeral semantic comparison was infl uenced by the number of numeral. In detail, under accord condition（i.e. smaller numeral matched with less numerals, bigger numeral matched with more numerals）, response was quicker, otherwise, response was slower. Stroop effect of numeral semantic comparison was about 15 ms（461ms and 476 ms respectively）. In the task of numeral number comparison, the response was infl uenced by numeral semantic, under accord condition, response was quicker, under disaccord condition, response was slower. Stroop effect of numeral number comparison was about 24 ms（459ms and 483 ms respectively）. Average error rate was analyzed by repetitive measure ANOVA. Main effect of task type was signifi cant, F（1, 19） = 1521.000, p=.000, the error rate of numeral semantic comparison was less than that of numeral number comparison（3.7% and 4.9% respectively）. Main effects of Stroop effect was signifi cant, F（1, 19） = 12705.575, p=.000, error rate of accord condition was less than that of disaccord condition（2.6% and 6.0% respectively）. All the stimuli of Experiment 2 were composed of starlet pictures, all of which covered equal area but the numbers of starlet within them were different. Experiment 2 included two parts, all of wich completed the same task： comparing the starlet numbers of comparison stimuli with that of standard stimulus, but the key-pressing rules were reverse. Data analyzing indicated that the main effect of number presentation order was very signifi cant, F（1, 20） = 31.421, p 〈.001,the average response time of from less to more was 478 ms,Std=14.9,the average response time of from more to less was 509 ms, Std=17.1. SNARC effect was signifi cant, F（1, 20） =9.514, p〈 .01. Under accord condition, the average response time was 469 ms, Std=14.2; under disaccord condition, the average response time was 519 ms, Std=20.7. SNARC effect was about 50 ms. Average error rate was analyzed by repetitive measure ANOVA. The main effect of number present order was not signifi cant, F（1, 20） = 2.913, p〉.05. SNARC effect was signifi cant, F（1, 20）=5.969, p〈.05. Error rate of accord condition was 4.3%, Std=.008, error rate of disaccord condition was 6.1%, Std=.009. To sum up, experiment 1 found that Stroop effect exists in both numeral number comparison task and number semantic comparison task. Experiment 2 found that there exists SNARC effect in number comparison task. An presentation order effect was also found, that is to say, when the numerals＇ semantic of standard was smaller than that of comparison simuli, Rts were faster, otherwise Rts were slower.