臺灣正常人班頓視覺保留測驗之常模研究

Abstract

背景：班頓視覺保留測驗(BVRT)是最廣泛被用來評估視覺能力、視覺記憶能力及視覺空間建構能力的神經心理測驗工具之一，具有施測時間短、使用方便及平行複本等優點。然而現有之BVRT常模樣本代表性不足，且缺乏心理計量特性等相關的資料，不利於研究及臨床應用。目的：本研究主要目的為建立臺灣正常人BVRT之常模，主要目標為：一、探討人口學變項對測驗表現的影響；二、評估BVRT正確分數及錯誤分數之心理計量特性；三、本常模於臺灣健康成年人之適用性。方法：本研究共收集臺灣308位16至90歲、教育程度0至18年之健康成年人於BVRT之常模資料，同時收集魏氏記憶量表之圖繪測驗及視覺形狀辨認測驗等資料以檢驗BVRT之效度。此外比較臨床病人與健康成年人在本測驗之表現以檢驗效度。部分樣本同時收集複本資料及再測資料以檢驗BVRT之信度。 結果：年齡及教育程度為主要影響測驗表現之人口學變項，性別則無影響；另外提供以年齡及教育程度做校正之常模資料。研究結果顯示本測驗具有良好之再測信度、效標關聯效度及建構效度，及中等程度之複本信度。本研究提供之常模資料具有良好之代表性、新近性及適切性。結論：本研究確認人口學變項對測驗表現的影響，並使用統計方法以排除人口學變項的效果，再以校正分數建立BVRT之正確分數及錯誤分數之百分等級對照表，可供臨床工作者偵測早期偵測、鑑別診斷及預後評估等應用。儘管臺灣樣本之複本信度不如預期，研究結果仍顯示BVRT具有良好之信效度。未來應重複驗證臺灣樣本於BVRT之複本信度，並建立其他複本之常模資料，以利於臨床上評估認知功能的變化或治療之成效。

Background: The Benton Visual Retention Test (BVRT) is increasingly utilized within research and clinical settings to assess visual perception, visual memory, and visuoconstructive ability due to its convenience to use, short administration time, and the availability of multiple parallel forms. However, the current available norms with questionable representativeness of samples, lack of examination of relations between the demographic variables and the test, and lack of examination of psychometric properties limits the utility. Objective: The present study was aimed to establish a norm for the BVRT. The specific objectives were to investigate: (1) the influences of demographic variables on the test, (2) the issue of psychometric properties, and (3) the issue of norm appropriacy. Methods: 308 participants were recruited through stratified sampling by current age (ranging from 16 to 90 years old), education (ranging from 0 to 18 years), and area of residence. Some participants were additionally administered the Visual Reproduction subtest of WMS-III and the Visual Form Discrimination Test to investigate the validity. Clinical Patient group was also recruited for verifying the validity. The test-retest and the alternate-forms reliability were also derived from a subgroup of participants. Result: The effects of age and education rather than gender were significant on the BVRT performance. The results also showed sound test-retest reliability, criterion-related validity, and the construct validity of the BVRT, with the exception of moderate alternate-forms reliability. The normative data have good representativeness, recency, and relevance. Conclusion: The present study identified the influences of demographic effects on the BVRT performances, established an appropriate normative data for evaluating the visual memory, and provided the referential table for percentile ranks. Despite the fact that the alternate-forms reliability was not as high as expected from Taiwanese samples, this study also verified the adequate reliability and the validity of the BVRT. It is useful for early detection, differential diagnosis, treatment evaluation, and prognosis. Further investigation on the issue of the alternate-forms reliability and the normative data of the Forms D and E are necessary.