Figure B 
Sampling Design of the longitudinal study 
Project IDSS - Sampling Design 
Social Class 
middle class 
lower class 
Class 4 
Class 5 Class 6 
Class 3 
Class 1 
Class 2 
 
 
 
 
 
 
 
 
 
 
TM 
 
JOw 
C 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
N=5 
N =5 
general 
 
 
 
— 
 
 
 
 
 
 
 
 
ability 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4 
 
 
 
 
high 
 
 
1 
 
 
 
 
 
 
urban (N = 120) 
 
— 
— 
 
 
 
 
3 non-urban 
 
communities 
N= 19 
N=21 
 
(complete cohorts) 
 
 
 
 
 
 
N =65 
L 
— 
 
— 
 
South 
North 
West 
farming 
fishing 
service 
community 
community 
community 
The sampling design of the longitudinal study was introduced with the aim of maximizing 
interindividual variance. Individual differences are taken to derive from competence level at the 
onset of school, to children's sex, to socio-economic status of parents and to the social and 
cognitive ecologies of different lifeworlds. 
The three dimensions according to which the urban sample was stratified were treated as factors 
in a quasi-experimental design. Although general ability level constituted a systematic 
stratifying dimension in the urban sample only, it could be derived retrodictively also for the 
rural children. Because the rural samples formed entire birth cohorts, the socio-economic status 
of parents is not equally distributed. 
Competence level 
During the first two weeks after school entrance teachers in all first grades of the city of 
Reykjavik were asked to nominate three children in the upper third, three children in the middle 
third and three children in the lower third of the general ability distribution in their particular 
classes. Subsequently, the middle third was discarded from the study. In the absence of 
3