http://dx.doi.org/10.6129/CJP.2009.5102.03
Chinese Journal of Psychology, 51(2), 2009

陳韻如（國立台灣大學心理學系暨研究所）；雷庚玲（國立台灣大學心理學系暨研究所）；鄭欣宜（林口長庚醫院兒童心智科）；王麗娟（林口長庚醫院兒童心智科）；柯乃綺（林口長庚醫院兒童心智科）；黃奕偉（臺北市立關渡醫院身心科）；謝家智（臺北市立關渡醫院身心科）；詹和悅（國立臺灣大學附設醫院精神醫學部）；楊智凉（國立臺灣大學附設醫院精神醫學部）； 李素貞（國立臺灣大學附設醫院精神醫學部）；吳佑佑（林口長庚醫院兒童心智科）；張學岭（林口長庚醫院兒童心智科）；黃玉書（林口長庚醫院兒童心智科）；劉弘仁（臺北市立關渡醫院身心科）；蔡文哲（國立臺灣大學附設醫院精神醫學部）

摘要

自閉症兒童不論在能力、症狀、情緒及行為上皆呈現個別差異。本研究欲探究氣質是否可做為獨立於自閉症狀之外且用以描述自閉症兒童個別差異之構念。考量國內已累積的實徵資料及臨床使用度，本研究擇定以Carey與McDevitt（1978）根據Thomas和Chess（1977）的理論所編製的氣質風格量表（Behavior Style Questionnaire, BSQ）做為初探自閉症兒童氣質的工具。共有45位生理年齡三歲一個月至七歲七個月之自閉症男童之母親填寫BSQ。分析結果發現：自閉症組在BSQ之九個向度上皆具有可被接受的內部一致性與再測信度，在六個向度中與一般男童的平均值有顯著差異、另在五個向度中的變異量較一般男童大，並有四個向度與症狀嚴重度有顯著相關。階層迴歸分析顯示，在排除了發展能力與症狀嚴重度的貢獻後，活動量、情緒本質、反應閾及反應強度等四個向度之迴歸係數仍可預測親職壓力，可視為獨立於症狀之外的自閉症兒童個別差異之構念。本研究並建議BSQ中的注意力分散度與反應閾兩分量表應拆解成不同面向以便能更精確地描述自閉症兒童的個別差異。

關鍵詞：自閉症、個別差異、氣質、親職壓力

The Individual Differences of Temperament and Parenting Stress in Preschoolers with Autism

Yunn-Ru Chen（Department of Psychology, National Taiwan University）；Keng-Ling Lay（Department of Psychology, National Taiwan University）；Shin-Yi Cheng（Department of Child Psychiatry, Linkou Chang-Gung Children’s Hospital）；Li-Chuan Wang（Department of Child Psychiatry, Linkou Chang-Gung Children’s Hospital）；Nai-Chi Ko（Department of Child Psychiatry, Linkou Chang-Gung Children’s Hospital）；Yi-Wei Huang（Department of Psychiatry, Taipei Municipal Gan-Dau Hospital）；Chia-Chih Hsieh（Department of Psychiatry, Taipei Municipal Gan-Dau Hospital）；Hoe-Yuek Chan（Department of Psychiatry, National Taiwan University Hospital）；Chih-Liang Yang（Department of Psychiatry, National Taiwan University Hospital）；Su-Chen Li （Department of Psychiatry, National Taiwan University Hospital）；Yu-Yu Wu（Department of Child Psychiatry, Linkou Chang-Gung Children’s Hospital）；Hsueh-Ling Chang（Department of Child Psychiatry, Linkou Chang-Gung Children’s Hospital）；Yu-Shu Huang（Department of Child Psychiatry, Linkou Chang-Gung Children’s Hospital）；Hung-Jen Liu（Department of Psychiatry, Taipei Municipal Gan-Dau Hospital）；Wen-Che Tsai （Department of Psychiatry, National Taiwan University Hospital）

Abstract

Children with autistic spectrum disorders vary significantly in developmental ability, symptomatology, as well as emotional and behavioral patterns. The large variance in behavioral and emotional patterns poses a major challenge to parents and clinical workers while interacting with children with autism. This research tested whether temperament, as an important area in understanding the individual differences of normal children, poses a unique contribution to the stress of parenting in children with autism. Forty-five Chinese mothers of male preschoolers diagnosed as autism (CA = 3 yr. 1 mo. to 7 yr. 7 mo.; MA = 1 yr. 5 mo. to 6 yr. 3 mo.) residing in Taiwan completed the Behavioral Style Questionnaire (BSQ-Chinese version; Carey & McDevitt, 1978), the Chinese Children Developmental Inventory (CCDI; Hsu & Hsu, 1976), the Clancy Checklist (Chinese version; Hsieh, Soong, & Hsu, 1983), and the Parenting Stress Index–Short Form (PSI/SF, Abidin, 1995) for assessing temperament, developmental age, autistic symptomatology, and perceived parenting stress, respectively. Results indicated acceptable internal consistency and six-month test-retest reliability of BSQ subscales for assessing autistic population. Compared with the BSQ scores of 243 normal male preschoolers between age five and seven residing in Taipei (Wang, 2003), preschoolers with autism received significantly higher scores in activity level (M = 4.11, SD = 1.01) and intensity of reaction (M = 4.07, SD = 1.09), and poorer scores in rhythmicity (M = 4.23, SD = .79), adaptability (M = 4.62, SD = 1.02), and attention span and persistence (M = 3.44, SD = .97). They were also less likely to be distracted (M = 3.94, SD = .95). In addition, larger group variances were found in the autistic sample on the dimensions of activity level, adaptability, quality of mood (M = 4.70, SD = .96), threshold of responsiveness (M = 3.83, SD = 1.08), and intensity of reaction. None of the dimensions yielded smaller variance for the autistic group than their normal counterparts. The severity of autistic symptomatology was significantly correlated with four BSQ dimensions, namely, activity level, adaptability, distractability, and approach/withdrawal (M = 3.88, SD = 1.02), rs = .30, -.30, -.55, -.41, ps < .05. The fact that there was a fair range of within-group temperamental differences among children with autism and that there were five BSQ dimensions not significantly related to autistic symptomatology yields that some of the BSQ dimensions may be regarded as independent constructs from the autistic symptomatology. In addition, hierarchical regression analyses yielded that after controlling for developmental ability and symptomatology, the dimensions of activity level, quality of mood, threshold of responsiveness, and intensity of reaction were still significant predicators for parenting stress. These four temperamental dispositions thus may be especially effective in predicting stressful parenting. This study also suggested that the dimensions of distractibility and threshold of responsiveness in BSQ may need to further break up into several distinctive aspects to better grasp the individual differences among children with autism.

Keywords: autism, individual differences, parenting stress, temperament