When word frequency meets word order: factors determining multiply-constrained creative association

Abstract

AbstractCreative association is inherent and essential to creativity and insight. Here we utilised a Chinese compound Remote Associates Task (cRAT) to identify the potential impact of word order (i.e., solution position hereinafter) and word frequency on creative association across two behavioural experiments. Experiment 1 identified the effects of (a) word order and word frequency on cRAT-induced association without considering the specific strategies used during solving such problems and (b) their interaction not only on performance in solving the cRAT, including solution time and accuracy, but also on difficulty rating. Following a coarse-to-fine approach, Experiment 2 examined the effect of both factors on creative association determined according to each participant’s trial-by-trial reports regarding insight solution strategies. Main effects of word order and word frequency, and a two-way interaction, were found on the accuracy and self-rated difficulty. These findings shed light on creative association, multiply-constrained problem solving and constructing compounds.Keywords: Creative associationword frequencysolution positionword orderproblem solving Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1 Word frequency was mainly determined according to a well-known and widely used word frequency dictionary, listing more than 1.8 million Chinese words or characters, published by the Beijing Language College Institute of Language Studies (Citation1986; the English version is available from the University of Virginia). If the frequency of a word was fewer than 10 times, with a percentage of cumulative frequency less than 5% (the taken threshold of 5% is largely based on a general threshold view drawn from the widely used statistical significance of p < 0.05 in psychological studies wherein an event with a probability/frequency of occurrence of less than 5% is considered a small probability or uncommon/infrequent event. Actually, the word-frequency dictionary authors also considered a word with fewer than 10 appearances, namely, word frequency, as a low-frequency word, which is also manifested in their sampling process), it was considered a low-frequency word; otherwise, it was viewed as a high-frequency word (not fewer than 10 times), occupying more than 95% in terms of percentage of cumulative frequency. It should be noted that the selected items in this study were from an original set of 192 developed items, such that some confounding variables (e.g., chronological effect) were excluded.2 In these two experiments, participants were instructed to try their best to solve each cRAT problem and simultaneously process them. To reduce the complexity and difficulty in design and statistical analysis, the three Chinese characters of a cRAT problem was controlled (not fixed but randomly), and presented simultaneously (not sequentially).3 Here we provided a brief description on the simple effect of two-way interaction regarding accuracy, and more details can be available by contacting the first author. The results are: HHHFFF > HLLIII, HHHIFF > HHHIII, HHHIFF > HLLIII, HHHIFF > LLLIFF, HHHIFF > LLLIIF, HHHIIF > HHHIII, HHHIIF > HLLIII, HHHIIF > LLLIFF, HHHIIF > LLLIIF, HHLFFF > HLLIII, HHLIFF > HLLIII, HHLIIF > HLLIII, HHLIIF > LLLIFF, HHLIIF > LLLIIF, HHLIII > HLLIII, HLLIFF > HLLIII, HLLFFF > HLLIII, HLLIIF > HLLIII > LLLFFF, and HLLIII > LLLIII, wherein “>” only denoting the difference between two types is greater for the former than the latter at the level of p<.05. Building on these effects, we believe that word frequency may work in a U-shaped approach rather than a linear way, which is still an open question that needs more careful evaluation in the future.Additional informationFundingThis work was supported by the National Social Science Foundation of China (20BXW118). The first author was also supported by the sixth high-level personnel training project in Jiangsu province (333) and the Research fund of Jiangsu Provincial Key Constructive Laboratory for Big Data of Psychology and Cognitive Science (No. 72592162002G), and Dr Yuan was funded by the Jiangsu Qinglan Project of Jiangsu Higher Education Institutions and Natural science fund for colleges and universities in Jiangsu Province (No. 23KJB180016).