Data for the neural mechanisms of the insight memory effect induced by metaphorical soluitons

The neural mechanisms underlying the successful encoding of insight events in the field of psychotherapy are largely unknown. This study investigated the insight memory effect and its related neural mechanisms in a psychotherapeutic scenario using metaphorical micro-counselling dialogues. 31 participants read a mental distress problem followed by a metaphorical or literal solution during functional magnetic resonance imaging scanning in the encoding phase, a memory retrieval test was performed outside the scanner one week later. The results revealed that metaphorical solutions were associated with stronger insightfulness in encoding and better memory performance in testing than literal solutions, and this effect was associated with greater activation in the bilateral hippocampus, amygdala, thalamus, caudate, left superior frontal gyrus, and middle temporal gyrus, suggesting that successfully learning from insight may rely on strong episodic memory encoding in a psychotherapeutic scenario that transmits the information of positive emotion, useful knowledge, and intrinsic reward.There were two data files named "Insight scores and brain activity for the interaction of insight and memory" and "Insight scores and brain activity for metaphorical and literal solutions regardless of memory", respectively. No missing data was included in two data files.Behavioural data, such as insight scores, accuracy, reaction time and confidence were collected for each trial via E-prime software (Version 2, Psychology Software Tools). fMRI data was acquired via a Siemens Trio 3.0 Tesla MR scanner at the Beijing MRI Center and image data analyses were conducted using the software of statistical parametric mapping (SPM8, http://www.fil.ion.ucl.ac.uk/spm/).There were 22 column variables in the Data 1, and the names of column variables and specific meanings of each variable were included in related labels when you open the "variable view". There were 31 subjects and each of them has two rows (conditions of metaphorical and literal solutions), which was suitable for hierarchical linear modelling (HLM) analyses.(1) For each subject，36 trial scores were collected for metaphorical and literal solutions by E-Prime for the variable of "Insight", "Confidence" and "ATTW" , then the average of 36 scores of each variable was calculated for each condition, respectively. (2)Variable "Hit": means hit rate for the old/new recognition in the testing phase. Hit = Numbers of correctly recognized old solutions/36(3) Variable "FA": means false alarm rate for the old/new recognition in test testing phase. FA= 1-Numbers of correctly recognized new solutions/36(4) Variable "Pr": means Pr for the old/new recognition in the testing phase, Pr = Hit - FA(5)Variable "LMTG", "RMTG", "LSTG", "RSTG", "LSFG", "LAMYG", "RAMYG", "LHIPPO", "RHIPPO", "LTHALA", "RTHALA", "LCAUDATE", "RCAUDATE ": Brain activity values in regions identified in the contrast of insight (metaphorical) > non-insight (literal) during the presentation of solutions. These regions were defined as spheres with a radius of 0.5 mm centred on the peak coordinates defined by the paired-samples t contrast of Insight (metaphorical) > Non-insight (literal).There were 31 subjects and each of them list in one row, and there were 37 column variables in the Data 2. The names of column variables and specific meanings of each variable were included in related labels when you open the "variable view". Based on behavioural memory performance during the test phase, each metaphorical and literal solution learned in the scanner during the encoding phase was further divided into either later-remembered (REM, solutions answered correctly both in the old/new recognition test for solutions and the target theme word recognition test for problems out of four choices) or later-forgotten (FOR) items. Therefore, four conditions were created: metaphorical solutions later remembered (Insight-REM, "IR" in short), metaphorical solutions later forgotten (Insight-FOR, "IF" in short), literal solutions later remembered (Non-insight-REM, "NIR" in short), and literal solutions later forgotten (Non-insight-FOR, "NIF" in short).(1) Bilateral hippocampus, amygdala, thalamus, and caudate were brain regions identified in the contrast of ([Insight-REM]-[ Insight-FOR]) > ([Non-insight-REM]-[Non-insight-FOR]), and, and these regions were defined as spheres with a radius of 0.5 mm centred on the peak coordinates defined in this contrast. Then brain activity values in regions were extracted via SPM8 for each condition.(2) For each subject, learning trials of metaphorical and literal solutions were divided into either later-remembered or later-forgotten items, and then related insight scores were calculated for each condition.

心理治疗领域中，顿悟事件成功编码的相关神经机制目前仍未被充分阐明。本研究采用隐喻性短咨询对话，在心理治疗情境下探究了顿悟记忆效应及其相关神经机制。31名被试在编码阶段的功能磁共振成像（functional magnetic resonance imaging, fMRI）扫描过程中，先阅读一则心理困扰问题，随后阅读隐喻性或直白性解决方案；一周后，在扫描仪外完成记忆提取测试。结果显示，相较于直白性解决方案，隐喻性解决方案在编码阶段能引发更强的顿悟体验，且在后续测试中表现出更优的记忆成绩；该效应与双侧海马、杏仁核、丘脑、尾状核、左侧额上回以及颞中回的激活增强相关，这表明在心理治疗情境中，从顿悟中成功学习或许依赖于强有力的情景记忆编码，此类情境可传递积极情绪、实用知识以及内在奖赏的相关信息。

本研究包含两个数据文件，分别命名为《顿悟与记忆交互作用的顿悟得分及脑活动数据》与《不考虑记忆效应的隐喻与直白解决方案的顿悟得分及脑活动数据》。两份数据文件均无缺失值。

本研究通过E-prime软件（版本2，心理学软件工具公司）收集每一试次的行为学数据，包括顿悟得分、准确率、反应时以及信心评分。功能磁共振成像数据由北京磁共振成像中心的西门子（Siemens）Trio 3.0特斯拉磁共振扫描仪采集，图像数据分析采用统计参数映射（statistical parametric mapping, SPM8，http://www.fil.ion.ucl.ac.uk/spm/）软件完成。

数据集1包含22个列变量，打开“变量视图”即可查看各列变量的名称与具体含义。本研究共有31名被试，每名被试对应两行数据（分别对应隐喻性与直白性解决方案条件），适用于分层线性模型（hierarchical linear modelling, HLM）分析。
(1) 针对每名被试，E-Prime收集了其在隐喻性与直白性解决方案条件下共36个试次的“顿悟得分”“信心评分”与“ATTW”变量数据，随后分别计算每种条件下各变量36个得分的平均值。
(2) 变量“击中率（Hit）”：指测试阶段新旧再认任务的击中率，计算公式为Hit=正确识别的旧解决方案数量/36。
(3) 变量“虚报率（FA）”：指测试阶段新旧再认任务的虚报率，计算公式为FA=1-正确识别的新解决方案数量/36。
(4) 变量“辨别力指数（Pr）”：指测试阶段新旧再认任务的辨别力指数，计算公式为Pr=Hit-FA。
(5) 变量“LMTG”“RMTG”“LSTG”“RSTG”“LSFG”“LAMYG”“RAMYG”“LHIPPO”“RHIPPO”“LTHALA”“RTHALA”“LCAUDATE”“RCAUDATE”：代表解决方案呈现阶段，顿悟（隐喻性）>非顿悟（直白性）对比范式所识别脑区的脑活动值。上述脑区以“顿悟（隐喻性）>非顿悟（直白性）”配对样本t对比得到的峰值坐标为中心，定义为半径0.5mm的球形区域。

数据集2包含31名被试的单一行数据，共37个列变量；打开“变量视图”即可查看各列变量的名称与具体含义。基于测试阶段的行为学记忆表现，编码阶段在扫描仪内学习的每一则隐喻性与直白性解决方案可进一步划分为后续记忆成功组（REM，指在解决方案新旧再认测试与问题主题词四选一再认测试中均答对的项目）与后续记忆失败组（FOR）。由此形成四种实验条件：后续记忆成功的隐喻性解决方案（顿悟-REM，简称IR）、后续记忆失败的隐喻性解决方案（顿悟-FOR，简称IF）、后续记忆成功的直白性解决方案（非顿悟-REM，简称NIR）以及后续记忆失败的直白性解决方案（非顿悟-FOR，简称NIF）。
(1) 双侧海马、杏仁核、丘脑与尾状核为([顿悟-REM]-[顿悟-FOR]) > ([非顿悟-REM]-[非顿悟-FOR])对比范式所识别的脑区；上述脑区以该对比范式得到的峰值坐标为中心，定义为半径0.5mm的球形区域，并通过SPM8提取各条件下的脑活动值。
(2) 针对每名被试，将隐喻性与直白性解决方案的学习试次划分为后续记忆成功与失败项目，随后分别计算每种条件下对应的顿悟得分。