3. 1. Response Style?
Number use (Chen, Lee, Stevenson, 1995)
• 1-7 point scale
Slightly agree (5), Agree (6), Strongly agree (7)
• 5, 6, 7 → 3
• 4 → 2
• 1, 2, 3 → 1
7-point scale 3-point scale
• U.S-Japan d = .81 d = .79
• Chile-Japan d = 1.24 d = 1.25
4. 2. Measurement Equivalence?
• Do well-being items function in the same way in two
cultures?
• Differential Item Function (DIF) Analysis
5. The DIF Analysis of Life Satisfaction
(Oishi, 2006)
0
0.2
0.4
0.6
0.8
1.0
-3 -2 -1 0 1 2 3
1
2
3
4
5
6
7
Ability
Probability
Item Characteristic Curv e: 1
Graded Response Model
0
0.2
0.4
0.6
0.8
1.0
-3 -2 -1 0 1 2 3
1
2
3
4
5
6
7
Ability
Probability
Item Characteristic Curv e: 6
Graded Response Model
Item 1: In most ways my life is close to ideal
China
US
Latent LS score
“disagree” “agree”
Latent LS score
6. 0
0.2
0.4
0.6
0.8
1.0
-3 -2 -1 0 1 2 3
1
2
3 4 5 6
7
Ability
Probability
Item Characteristic Curv e: 4
Graded Response Model
0
0.2
0.4
0.6
0.8
1.0
-3 -2 -1 0 1 2 3
1
2
3 4
5
6
7
Ability
Probability
Item Characteristic Curv e: 9
Graded Response Model
China
US
Item 4: So far I have gotten the important things I want in life
“disagree”
“agree”
Satisfied Chinese (+1SD)
Satisfied American (+1SD)
7. Does mean difference disappear, if we use only
the equivalent items?
Answer: Not really
Original Mean Difference:
d = 1.18
Weighted Latent Mean Difference:
d = .83
9. 5. Meaning & Desirability of Happiness
Happiness more desirable in some cultures than others
→ Above or below-expected level of happiness?
Columbia
Puerto Rico
Peru
Thailand China
Nepal
HK
Japan
Taiwan
R = .37*
10. Measurement Issues Across Cultures
1. Response style?
Probably not
2. Lack of Measurement Equivalence?
Slightly
3. Self-presentation?
Slightly
4. Judgmental and memory bias?
Moderately
5. Desirability of happiness?
Moderately