重新檢視閱讀困難與視覺功能的關係－－從一個大量行為樣本的角度

Revisiting the Relationship between Visual Functions and Reading Difficulty: A Large Sample Perspective

在閱讀障礙的解釋假說裡，有一主張特定型態（主要是視覺型）的閱讀困難者，其大腦裡的特定皮質下區域─側膝核的巨細胞─缺損假說，目前仍是頗具爭議。其中的一些早期支持證據，也就是利用特定作業表現的閱讀困難vs.正常組的差異，來當做支持「對比敏感度作業可用來鑑別閱讀困難程度」的證據─被認為是忽略了閱讀困難的發生率與特定作業表現的其他交集可能─如篩選作業差，但卻閱讀正常─的比例，且此種的學童的比例，可能人數比「篩選作業差／閱讀困難」的學童還多，消弱其對巨細胞缺損假說的證據力。以上的說法，迄今尚未看到驗證。因此本研究以目前在台灣學校教室經常使用的「IRS教學回饋系統」，進行一個小學二、三年級大型樣本的檢驗。一共成功蒐集到381名（點作業協調性作業），與165名（對比敏感度測驗）學童的有效樣本，對照現行標準閱讀能力測驗的分數作相關，並與學生該學期之國語及數學成績作對照。結果非但修正了早先對於閱讀正常／困難的臆測之外，也發現到儘管點協調性與對比敏比感度測量皆與閱讀能力有顯著相關，點協調與對比敏感兩者的相關卻未達顯著，顯示兩者與閱讀能力的關聯顯非源自於同一來源。在列舉了一些其他解釋力較低的可能─如視角或亮度，或兒童視力差異等的可能影響─後，本研究結論雖支持視覺功能與閱讀能力的關聯，但同樣不支持巨細胞的缺損假說。最後，本研究除可當作一國小學童閱讀困難團體檢測驗的高可行性方案，亦可供其他領域的研究者作參考。

Among the various etiologies of developmental dyslexia, the hypothesis of the defected specific subcortical structure, known as the magnocellular (M) pathway, is still contentious. Its supporting evidence is limited in that some task comparisons, which normally show group differences between dyslexic and normal readers, fail to consider the prevalence rate of dyslexia. Such negligence may have missed some poor visual task performers who have normal reading, and the normal vs. dyslexic-at-risk poor visual performers may be equal ratio or larger in the former, making magnocellular hypothesis much less persuasive. So far, this assertion remains untested. In the current study, large sample were collected from the second and third Taiwan elementary school graders, on reading comprehension (as well as their academic performance on mathematics and language for corroboration) and the two often cited M-related tasks: dot coherence (DC, N = 381) and contrast sensitivity (CS, N = 165). The results disagreed with the previously suggested 1 ~ 2:1 poor-DC/CS with normal or with poor reading ratio, suggesting that the real ratios could be larger (4 ~ 8:1). In addition, significant DC-reading and CS-reading, but insignificant DC-CS, correlations were found, indicative of shared yet distinct reading-related mechanisms that are less likely from a single origin. Among other possible but unfavorable explanations (such as test position variability, light conditions during testing, and the students' visual acuities), these results concur the relations between visual functions and reading, but unsupportive of the magnocellular hypothesis of dyslexia. Lastly, the current study exemplifies an efficient group testing/screening protocol applicable to developmental dyslexia and other research domains.