目的：本研究的目的在驗證魏氏成人智力測驗第三版(WAIS-III) 的Blyler 和其它四個分測驗組成的簡式版本在住院慢性精神分裂症病患的適用性。方法：本研究的實驗樣本來自三所醫院的住院慢性精神分裂症病患共103 人，控制樣本則來自社區內正常自願者共 66 人。每位受試者皆接受WAIS-III 所有分測驗的施測，並記錄所花費的時間。結果：本研究顯示精神分裂症病患組在WAIS-III 四個指標智商的全距較正常組大；在檢驗所有由四個指標智商組合而成的簡式版本中，共有S-BD-A-DSS；I-BD-DS-DSS；I-MR-DS-DSS；和I-MR-DS-SS 等四種組合，相對於Blyler 的簡式版本，對於精神分裂症病患組和正常組的FSIQ 解釋變異量較佳，所需要的施測時間較短，而且FSIQ 預測值與FSIQ 真實值的平均數差異的絕對值落於5 分內的百分比較高。結論：本研究發現有四種組合的簡式版本，在使用於國內的年輕住院慢性精神分裂症病患時，優於Blyler 等人的簡式版本(I-A-BDDSS 和I-BD-S-C)，唯使用時應注意其應用上的限制。

Objectives: To assess the appropriateness of Blyler's and four other subtests of the short form of the Wechsler Adult Intelligence Scale-III (WAIS-III) for Taiwanese inpatients with chronic schizophrenia. Methods: The subjects enrolled in this study were inpatients recruited from three different hospitals and the normal controls were volunteers recruited from communities. A total of 103 patients with schizophrenia and 66 normal controls were enrolled. All participants were administered the WAIS-III (Chinese Version) with 14 subtests in standard procedures described in the test manual. In addition, the administration time for each subtest was also recorded. Results: Results showed that the schizophrenia group had a larger full range than the normal control group among four index IQs. After analyzing all combinations from the four subtests, which belonged to the four index IQs, there were four combinations that were better than Blyler's short forms for accounted variance in full scale IQ (FSIQ), administration time and percentage of estimated FSIQ scores within 5 points of actual FSIQ. These forms were: Similarity - Block Design - Arithmetic - Digit Symbol Substitution (S-BD-A-DSS); Information - Block Design - Digit Span - Digit Symbol Substitution (I-BD-DS-DSS; Information - Matrix Reasoning - Digit Span - Digit Symbol Substitution (I-MR-DS-DSS); and, Information - Matrix Reasoning - Digit Span - Symbol Search (I-MR-DS-SS). Conclusion: Four compositions (S-BD-A-DSS, I-BD-DS-DSS, I-MR-DS-DSS and I-MR-DS-SS) were superior to Blyler's short forms for obtaining gross estimation of FSIQ in young inpatients with chronic schizophrenia, although interpretation of estimated IQ scores should be undertaken with caution.