reliability
The extent to which the measurement instrument gives the same results when the measurement is repeated.
當重復試驗的時候，依然會得到相同的結果, 即測量結果的可靠性。沒有random error.?
如何測量？how to assess scale reliability?
方法一：分兩半，分別計算score，再算二者的correlation。
好處：一次處理。解決了test-retest 方法中的問題。（什么問題？）
問題：如何分兩半
方法二：Cronbach's a?
a can be viewed as the expected correlation of two tests that measure the same construct.
a 可以看做是兩次對相同變量檢驗的相關程度。mean correlation over all possible split-half reliabilities.
它含蓄地假定了：the average correlation of a set of items is an accurate estimate of the average correlation of all items that pertain（屬于） to a certain construct
由公式推導出：a depends on 1)number of items ;2) mean correlation among items
a大于0.7 is acceptable
Validity
the extent to which the measurement instrument indeed measures the construct that we intend to measure.
確實是我們想測量的那個東西
content validity
content （items）是否覆蓋全面（the complex construct）
測量方法：定義construct- 產生items-專家看看是不是都fit- 做item analysis（factor analysis & cronbach）
assume uni-dimensional, confirmatory factor analysis.(上一章) 啥意思？
dimensionality of the scale matches construct?
criterion validity
和其他變量的關系，是否預測成功（scale predicts the behavior it is supposed to predict）
如何做？
--這個新的scale應該要predict （在此例中 tendency to purchase grooming products）檢驗什么（scale score）和什么(predicted behavior)的顯著相關性(significantly correlate with)
final scale scores = average?
SPSS-correlation （以前章節）
construct validity
是根據理論說的要測量的那玩意嗎
·convergent - 這個scale和其他measures of the same construct 正相關嗎
·discriminant - 這個scale和 measures of different constructs 不相關嗎
·nomological - 這個scale 和throry一致嗎
如何做？
建立一個network（是什么鬼）of caual relationships
檢驗hypothesis
if yes, validity is established.
if no, there are two possibilities?
? 1 scale is not valid (前兩個)?
? 2 theory is wrong?
