I read the explanation of reasons for finding low ICC values in chapter 26 of the book, Foundations of Clinical Research application to practice.
One of the reason is variance. The authors said "the variability among subjects' socres must be large to demonstrate reliability. A lack of variability can occur when samples are homogeneous, when raters are all very lenient or strict in their scoring, or when the rating system falls within a restricted range........If subjects' socres are homogeneous, this source of variance will not be significant....Therefore, the researchers have to try to obtain heterogeneous samples whenever possible."
My question is that I don't really understand the meanings of homogeneous and heterogeneous mentioned here. Did the authors suggest that we have to recruit the clients with different diagnoses in order to obtain heterogeneous samples??
那您掌握了那些作者的看法?
回覆刪除different diagnoses = heterogeneous samples??
stroke 病人無法組成 a heterogeneous sample??
我重覆看了許多次，我尚未掌握作者的看法，仍覺得困惑，到底異質性和同質性的樣本是指什麼?
回覆刪除我在想，例如：要看血壓和中風之間的關係，研究者得到的樣本是來自同一個家族，此樣本的同質性很高，同一個家族成員會有可能因為基因遺傳因素，導致之間的變異性較小。
所以要確定血壓是否為中風的預測因子，就要找不同家族和或樣本之間的相似性要小，才能看出變異性。
我的想法對嗎??
您還沒回覆: 掌握了那些看法(看懂了那些)?
回覆刪除I will respond it as soon as possible.
回覆刪除今日事今日畢
回覆刪除拖拖拉拉，習慣不良。
我看書上variance(變異)對reliability(信度)的解釋，信度是true score(真實值)佔所有觀察值的比例是多少。真實值的變異度越大，所獲得的信度也越高。
回覆刪除增加樣本的變異，就可以增加真實值的變異，就可增加信度。
以我對書上的例子理解，增加不同類型的樣本即可增加樣本的變異。
我看了另一本書，Research in Occupational Therapy: Methods of Inquiry for Enhancing Practice, written by Gary Kielhofner.
書上提出ICC is not like other reliability estimates. ICC is not sample dependt.(page 165)