11. 形成型與反映型指標的差異
“Whereas reflective indicators are essentially interchangeable (and
therefore the removal of an item does not change the essential nature of
the underlying construct), with formative indicators ‘omitting an indicator
is omitting a part of the construct’.”
(Diamantopoulos
Reflective
Construct
/ Winklhofer, 2001, p. 271)
形成型指標重點
在互補的指標之
間重疊的最小化
Formative
Construct
反映型指標重點在指標
之間重疊的最大化
20. SEM兩大家族
• Covariance-based SEM (CBSEM)
– Goal:樣本矩陣與模型期望共變異數最接近
例如:
樣本矩陣為…
模型期望共變異數為…
– Data source: raw data, covariance matrix
or correlation matrix with standard
deviation
– Software: Amos, LISREL, EQS, Mplus…
21. SEM兩大家族
• Variance-based SEM (PLS)
– Goal:內生變數的被解釋能力最大
– Ex:死亡
– Data source: raw data
(format: .csv 或.txt)
– Software: SmartPLS, PLS-Graph,
VisualPLS (中文), PLS-GUI, SPAD PLS…
25. 結構模式與測量模式
Inner model and Outer model
Outer model
e1
x1
e2
x2
e3
x3
Inner model
Outer model
D
Lx1
Lx2
Ly1
F1滿意度
b
Lx3
測量(CFA)模型
F2忠誠度
Ly2
Ly3
結構模型
y1
e4
y2
e5
y3
e6
測量(CFA)模型
25
53. PLS估計方法選擇
• Factor Weighting Scheme
– the estimation of LV will be estimated
using correlations (inner estimation)
• Path Weighting Scheme
– the estimation of LV will be estimated
with regression and correlations
(depends on the arrows)
• Centroid Weighting Scheme
– used when the other does not run
(it is an older 三星統計服務有限公司
procedure)
75. EXAMPLES from MISQ
An Empirical Examination of Individual
Traits as Antecedents to Computer
Anxiety and Computer Self-Efficacy
Managing Client Dialogues during
Information Systems Design to
Facilitate Client Learning
三星統計服務有限公司
85. 結構係數t值
Path Coefficients (Mean, STDEV, T-Values)
• Report Html Report
• t >1.96 at p < 0.05, t > 2.58, at p < 0.01,
t > 3.29 at p < 0.001 for two-tailed tests
Original
Sample (O)
Sample
Mean (M)
Standard
Deviation
(STDEV)
Standard
T Statistics
Error
(|O/STERR|)
(STERR)
CI -> ATT
0.326
0.326
0.062
0.062
CI -> SAT
0.197
0.191
0.072
0.072
SAT -> ATT
0.304
0.304
0.057
0.057
SC -> ATT
0.204
0.206
0.063
0.063
SC -> SAT
0.088
0.090
0.053
0.053
SQ -> SAT
0.473
0.480
0.062
0.062
The bootstrap procedure is just used to compute standard error and
t-values of outer loadings, outer weights, path coefficients.
For these reasons: t = original / std error
5.273
2.745
5.349
3.260
1.654
7.616
87. 總效果Total Effect
Original
Sample
(O)
Sample
Mean
(M)
Standard Standard
T Statistics
Error
Deviation
(|O/STERR|)
(STDEV) (STERR)
CI -> ATT
0.385
0.385
0.066
0.066
5.867
CI -> SAT
0.197
0.193
0.074
0.074
2.658
SAT -> ATT
0.304
0.301
0.058
0.058
5.265
SC -> ATT
0.231
0.233
0.065
0.065
3.566
SC -> SAT
0.088
0.092
0.056
0.056
1.576
SQ -> ATT
0.144
0.144
0.033
0.033
4.396
SQ -> SAT
0.473
0.478
0.059
0.059
8.023
顯著性檢定