Quantifying the time course of tantrum anger

描述(申请人提供)：虽然愤怒是一种重要的情绪，存在于公开攻击之下，具有巨大的社会代价，但人们对愤怒与其相关行为之间的关系知之甚少。蹒跚学步的孩子们几乎无处不在的发脾气为人们提供了一扇了解愤怒的窗口，愤怒可能是如此强烈，以至于无法进行系统的观察。对330个以前收集的发脾气行为的分析表明，不同愤怒行为的起伏反映了愤怒的瞬间水平。我们建议构建一个量化模型，将观察到的每种愤怒行为的时间变化与驱动它们的愤怒强度的变化联系起来。该模型将与目前正在收集的更详细的Tantrum数据集进行测试。 因为所有容易观察到的愤怒行为都已经被纳入到模型中，所以需要新的发脾气愤怒衡量标准来进行实证检验。为了满足这一需要，同时开辟一个有前途的新研究领域，我们将从声学上分析Tantrum尚未探索的多样性 声乐。通过缝在孩子衬衫上的麦克风记录下的尖叫、喊叫、哀鸣和哭泣，将被数字化，并由DFT进行频谱分析。发声概率、持续时间、F0特征和发音困难的时间变化将与根据同时录制的行为计算出的愤怒轨迹进行比较。我们还将使用FACS编码的面部表情和父母和儿科专家的全球愤怒评级来测试该模型。 成功建模的主要好处之一是能够根据愤怒强度来衡量行为，估计总体或最大发脾气愤怒，以及比较儿童内部和儿童之间的愤怒强度(例如，与发脾气触发因素、父母干预和儿童精神病理学有关)。该项目还将为此类时间进程数据的分析提供一个范例。这项研究的创新性，以及它对发脾气的声学分析的新颖应用，使这一应用程序符合R21计划的要求。

DESCRIPTION (provided by applicant): Although anger is an important emotion underlying overt aggression of great social cost, little is known about the relations between anger and the behaviors with which it is associated. The almost ubiquitous tantrums of toddlers provide a window onto anger which can be so intense as to be otherwise inaccessible to systematic observation. Analyses of 330 previously collected tantrums suggest that the rise and fall of different angry behaviors reflects the momentary level of anger. We propose to construct a quantitative model which relates the observed temporal variation in each of the angry behaviors to the changing intensity of anger which drives them all. The model will be tested against a more detailed tantrum data set now being collected. Because all the easily observed angry behaviors are already incorporated into the model, new measures of tantrum anger are needed for empirical testing. To meet this need while opening up a promising new area of study, we will acoustically analyze the unexplored diversity of tantrum vocalics. Screams, shouts, whines, and cries, recorded by microphones sewn into the child's shirt, will be digitized and spectrally analyzed by DFT. Temporal changes in vocalic probability, duration, F0 characteristics, and dysphonation will be compared to the anger trajectory calculated from the simultaneously videotaped behaviors. We will also test the model with FACS coded facial expressions and global anger ratings by parents and pediatric experts. Among the major benefits of successful modeling would be the ability to scale behaviors by anger intensity, to estimate total or maximum tantrum anger, and to compare anger intensity within and between children (e.g., related to tantrum triggers, parental intervention, and childhood psychopathology). This project will also provide a paradigm for the analysis of such time course data. The innovative nature of this research, and its novel application of acoustic analysis to tantrum vocalics, qualifies this application for the R21 program.