Correlational findings

Study Feather & O'Brien (1986): study AU 1980

School-leavers, Australia, followed 2 years,1980-1982
N = 5446
Non Response:
T2: 25%, T3: 34%
Questionnaire: paper
Mailed orr administered in the schools.


Authors's Label
Our Classification
Level of happiness at T1(1980), T2(1981) and T3(1982). T1 and T2 happiness by T2 employment(sample A): T2-empl. T2 unempl. difference
Related specification variables
0  Employed at T2/T3.
1  Unemployed at T2/T3

Observed Relation with Happiness

Happiness Measure Statistics Elaboration / Remarks A-AOL-g-mq-n-24-a AoV = + ns happiness at T1  Mt'= 6.53 Mt'= 6.46  -.07
happiness at T2  Mt'= 6.61 Mt'= 6.16  -.45
Interaction employment x time (nonsignificant):
happ.  T1 to T2       +.08      -.30  -.38

T1 and T3 happiness by T3 employment(sample A):
                 T3-empl.  T3 unempl. difference
A-AOL-g-mq-n-24-a AoV = - p < .01 happiness at T1  Mt'= 6.60 Mt'= 6.36  -.24
happiness at T3  Mt'= 6.50 Mt'= 5.80  -.70
Interaction employment x time ( p<01 ):
happ.  T1 to T3       -.10      -.56  -.46

T2 and T3 happiness by T3 employment(sample A+B):
                 T3-empl.  T3 unempl. difference
A-AOL-g-mq-n-24-a AoV = - p < .01 happiness at T2  Mt'= 6.27 Mt'= 5.93  -.34
happiness at T3  Mt'= 6.43 Mt'= 5.60  -.83
Interaction employment x time ( p< 01):
happ.  T2 to T3       +.16      -.33  -.49

Significance assessed by variance-analysis, also involving employment.

The original depression score is ranging from
4 to 28. Mt' is standardized into a happpiness score (0-10 and reversed)

Possible testing effects checked by comparing three sub-samples:
- sample A: selected at T1, tested at T1, T2, T3
- sample B: selected at T1, tested at T2 and T3
- sample C: selected at T1, tested at T3 only
At T3 happiness appeared lower in sample A than in sample B, but there is little evidence for systematic testing effects.