中介變數 （粵拼 ：zung1 gaai3 bin3 sou3 英文 ：mediator variable 統計學 上嘅一個概念：如果有三個變數，
X
{\displaystyle X}
Y
{\displaystyle Y}
M
e
{\displaystyle Me}
X
{\displaystyle X}
M
e
{\displaystyle Me}
M
e
{\displaystyle Me}
Y
{\displaystyle Y}
M
e
{\displaystyle Me}
X
{\displaystyle X}
Y
{\displaystyle Y}
迴歸分析 分析中介效應 （mediation effect）最簡單嘅有三個步驟[ 1] [ 2]
用迴歸分析睇吓
X
{\displaystyle X}
Y
{\displaystyle Y}
Y
=
β
10
+
β
11
X
+
ε
1
{\displaystyle Y=\beta _{10}+\beta _{11}X+\varepsilon _{1}}
β
11
{\displaystyle \beta _{11}}
顯著 。
用迴歸分析睇吓
X
{\displaystyle X}
M
e
{\displaystyle Me}
M
e
=
β
20
+
β
21
X
+
ε
2
{\displaystyle Me=\beta _{20}+\beta _{21}X+\varepsilon _{2}}
β
21
{\displaystyle \beta _{21}}
顯著 。
用迴歸分析睇吓
X
{\displaystyle X}
M
e
{\displaystyle Me}
Y
{\displaystyle Y}
Y
=
β
30
+
β
31
X
+
β
32
M
e
+
ε
3
{\displaystyle Y=\beta _{30}+\beta _{31}X+\beta _{32}Me+\varepsilon _{3}}
β
32
{\displaystyle \beta _{32}}
顯著 ，以及
β
31
{\displaystyle \beta _{31}}
絕對值 係咪細過
β
11
{\displaystyle \beta _{11}}
中介效應嘅圖解 一個諗中介嘅統計模型可以涉及多過一個層次嘅變數，即係話當中有啲變數係量度緊某啲個體（例：員工），而高層次嗰啲個體（例：公司 ）係由低層次嗰啲個體組成嘅。不過一般嚟講，高層嗰啲變數會係預測低層變數數值嘅自變數 [ 3]
↑ MacKinnon, D. P. (2008). Introduction to Statistical Mediation Analysis . New York: Erlbaum.
↑ Baron, R. M. and Kenny, D. A. (1986) "The Moderator-Mediator Variable Distinction in Social Psychological Research – Conceptual, Strategic, and Statistical Considerations", Journal of Personality and Social Psychology , Vol. 51(6), pp. 1173–1182.
↑ Tofighi, D., & Thoemmes, F. (2014). Single-level and multilevel mediation analysis (PDF). The Journal of Early Adolescence , 34(1), 93-119.
