Abstract: On the basis of the theory of prototype elicitation in insight, this study design two experiments to examine whether different positions will be influenced by task difficulty and presentation time of prototype elicitation logogriph in four-word logogriph problem solving. 16 target logogriphs and 16 prototype elicitation ones are selected from logogriph database established by Cong and Zhang (2013) to test. Each of the two types is composed of 8 slack block logogriphs (SBL) and 8 compact block logogriphs (CBL). Every participant is required to finish 16 random logogriphs from the two types of logogriphs. Experiments are programmed by E-prime 2.0, where the logogriph type is tested as a within-subject variable including the SBL and CBL, the presentation time of prototype is a between group variable consisting of prototype existence (PE) and prototype disappearance (PD). The first experiment is 2（logogriph type）×2（presentation time of prototype）mixed design, the prototype is in front of the target logogriph(BTL) and 77 undergraduates are randomly assigned into the PE and PD groups. Repeatedly measured variance analysis shows that the accuracy of the SBL is significantly higher than that of the CBL on whichever conditions of PE or PD (p=.001), and the accuracy of the PD is remarkably higher than that of the PE (p=.0036),we call this time offset effect. Additionally there is a strong interaction between logogriph type and presentation time of prototype (p=.018). Further analysis on the simple effect displays that the scores of CBL on PD conditions are significantly higher than those on the PE ones (p<.001). In order to check the time offset effect and further examine whether the different positions of prototype can influence logogriph problem solving, we design experiment 2. It is 2（logogriph type）×2（the presentation time of prototype）×2 (position) mixed design. The position includes two cases , the prototype is placed behind (BTL) or in front of the target logogriph(FTL) . A repeated-measures ANOVA is conducted with logogriph type (SBL vs. CBL) as a within-subject variable, presentation time of prototype (PE vs. PD) and position (BTL vs. FTL) as between subject variables. The results not only demonstrate experiment 1 , but also show a strong interaction between logogriph type, the presentation time and position of prototype(p=.00). Further analysis on the simple effect display that ,for BTL, the scores of the SBL on the PD conditions are significantly higher than those on the PE ones (p=.003), While for FTL, the scores of the CBL on the PD conditions are significantly higher than those on the PE ones. These results suggest there are time offset effect and prototype position effect in logogriph problem solving. Specifically, the SBL compare to the CBL and the PD compared to the PE are easier to solve in logogriph problem, for BTL ,the SBL is easier to be influenced by the presentation time of the prototype, whereas for FTL,the CBL is easier to be influenced by the presentation time of the prototype. In order to check the time offset effect and further examine whether the different positions of prototype can influence logogriph problem solving, we design experiment 2. It is 2（logogriph type）×2（the presentation time of prototype）×2 (position) mixed design. The position includes two cases , the prototype is placed behind (BTL) or in front of the target logogriph(FTL) . A repeated-measures ANOVA is conducted with logogriph type (SBL vs. CBL) as a within-subject variable, presentation time of prototype (PE vs. PD) and position (BTL vs. FTL) as between subject variables. The results not only demonstrate experiment 1 , but also show a strong interaction between logogriph type, the presentation time and position of prototype(p=.00). Further analysis on the simple effect display that ,for BTL, the scores of the SBL on the PD conditions are significantly higher than those on the PE ones (p=.003), While for FTL, the scores of the CBL on the PD conditions are significantly higher than those on the PE ones. These results suggest there are time offect effect and prototype position effect in logogriph problem solving. Specifically, the SBL compare to the CBL and the PD compared to the PE are easier to solve in logogriph problem, for BTL ,the SBL is easier to be influenced by the presentation time of the prototype, whereas for FTL,the CBL is easier to be influenced by the presentation time of the prototype. Experiments are programmed by E-prime 2.0, where the logogriph type is tested as a within-subject variable including the SBL and CBL, the presentation time of prototype isa between group variable consisting of prototype existence (PE) and prototype disappearance (PD). The first experiment is 2（logogriph type）×2（presentation time of prototype）mixed design, the prototype is in front of the target logogriph(BTL) and 77 undergraduates are randomly assigned into the PE and PD groups. Repeatedly measured variance analysis shows that the accuracy of the SBL is significantly higher than that of the CBL on whichever conditions of PE or PD (p=.001), and the accuracy of the PD is remarkably higher than that of the PE (p=.0036),we call this time offset effect. Additionally there is a strong interaction between logogriph type and presentation time of prototype(p=.018). Further analysis on the simple effect displays that the scores of CBL on PD conditions are significantly higher than those on the PE ones (p<.001). In order to check the time offset effect and further examine whether the different positions of prototype can influence logogriph problem solving, we design experiment 2. It is 2（logogriph type）×2（the presentation time of prototype）×2 (position) mixed design.The position includes two cases , the prototype is placed behind(BTL) or in front of the target logogriph(FTL) . A repeated-measures ANOVA is conducted with logogriph type (SBL vs. CBL) as a within-subject variable, presentation time of prototype (PE vs. PD) and position (BTL vs. FTL) as between subject variables. The results not only demonstrate experiment 1 , but also show a strong interaction between logogriph type, the presentation time and position of prototype(p=.00). Further analysis on the simple effect display that ,for BTL, the scores of the SBL on the PD conditions are significantly higher than those on the PE ones (p=.003), While for FTL, the scores of the CBL on the PD conditions are significantly higher than those on the PE ones. These results suggest there are time offset effect and prototype position effect in logogriph problem solving. Specifically, the SBL compare to the CBL and the PD compared to the PE are easier to solve in logogriph problem, for BTL,the SBL is easier to be influenced by the presentation time of the prototype, whereas for FTL,the CBL is easier to be influenced by the presentation time of the prototype. Two experiments are programmed by E-prime 2.0, where the logogriph type is tested internal variable including the SBL and CBL and the presentation time of prototype is between group variable consisting of prototype existence (PE) and prototype disappearance (PD). The first experiment is 2（logogriph type）×2（the presentation time of prototype）mixed design, the prototype is in front of the target logogriph(BTL) and randomly-selected 77 undergraduates are randomly assigned into the PE and PD groups. Repeatedly measured variance analysis on the experimental results show that the accuracy of the SBL is significantly higher than that of the CBL on whichever conditions of the PE or the PD (p=.001), and that the accuracy of the PD is remarkably higher than that of the PE (p=.0036), thus suggesting strong interaction between the logogriph type and time offset (p=.018). Further analysis on the simple effect display that the scores of the CBL on the PD conditions are significantly higher than those on the PE ones (p<.001). In order to check the results from the first experiment and further examine whether the prototype’s different positions can influence logogriph problem solving, we design experiment 2. It is 2（logogriph type）×2（the presentation time of prototype）×2 (position) mixed design. The position includes two cases ,there are the prototype is placed behind the target logogriph(FTL) or the prototype is in front of the target logogriph(BTL) . A 2 × 2 × 2 repeated-measures ANOVA is conducted with logogriph type (SBL vs. CBL) as a within-subject variable, time offset (PE vs. PD) and position (BTL vs. FTL) as between subject variables. The results not only demonstrate experiment 1 , but also show a strong interaction between logogriph type，time offset and position (p=.00). Further analysis on the simple effect display that for BTL, the scores of the SBL on the PD conditions are significantly higher than those on the PE ones (p=.003), While for FTL, the scores of the CBL on the PD conditions are significantly higher than those on the PE ones. These results suggest that the logogriph type as well as the position and presentation time of the prototype can indeed affect the logogriph problem solving. Specifically, the SBL compare to the CBL and the PD compared to the PE are easier to solve the logogriph problem, and as the prototype is placed behind the target logogriph it is easier to solve the SBL problem whereas the SBL relative to the CBL is easier to be influenced by the presentation time of the prototype, and vice versa. This study discusses the influence of task difficulty and time offset of the prototype elicitation on logogriph problem solving from a new viewpoint, which will lay the theory’s foundation on better solving problems from study and life. In order to check the results from the first experiment and further examine whether the prototype’s different positions can influence logogriph problem solving, the second experiment was designed that all the conditions were same to the first experiment except that the prototype was placed behind the target logogriph(FTL). The results demonstrate experiment 1 , and A 2 × 2 × 2 repeated-measures ANOVA was conducted with logogriph type (SBL vs. CBL) as a within-subject variable, time offset (PE vs. PD) and position (BTL vs. FTL) as between subject variables. The results show a strong interaction between logogriph type，time offset and position (p=.00). Further analysis on the simple effect displays that for BTL, the scores of the SBL on the PD conditions are significantly higher than those on the PE ones (p=.003), While for FTL, the scores of the CBL on the PD conditions are significantly higher than those on the PE ones. These results suggest that the logogriph type as well as the position and presentation time of the prototype can indeed affect the logogriph problem solving. Specifically, the SBL compared to the CBL and the PD compared to the PE are easier to solve the logogriph problem, and as the prototype is placed behind the target logogriph it is easier to solve the SBL problem whereas the SBL relative to the CBL is easier to be influenced by the presentation time offset of the prototype, and vice versa. This study discussed the influence of task difficulty and time offset of the prototype elicitation on logogriph problem solving from a new viewpoint, which will lay the theory’s foundation on better solving problems from study and life.

Key words: insight, prototype elicitation, logogriph problem solving, time offset, prototype position effect

摘要： 根据顿悟的原型启发理论，通过两个实验考察原型启发的时间抵消和位置效应。实验1采用2（字谜类型）×2（原型呈现时间）混合设计，结果发现：松组块字谜的正确率显著高于紧组块字谜；原型谜面消失的正确率显著大于谜面存在的正确率。实验2采用2（原型位置）×2（字谜类型）×2（原型呈现时间）混合设计，结果不仅证明了实验1，且原型位置、字谜类型、原型呈现时间的交互作用显著，原型在靶字谜之前时，紧组块字谜上谜面消失的正确率显著大于谜面存在的正确率；原型在靶字谜之后时，松组块字谜上谜面消失的正确率显著大于谜面存在的正确率。表明字谜顿悟问题解决中存在原型的时间抵消和位置效应，二者连同任务难度共同影响问题解决。

关键词: 顿悟, 原型启发, 字谜问题解决, 时间抵消效应, 原型位置效应