Flynn (1994, 1999, 2007) suggested that sustained improvements in IQ over the 20th century were due to an increasing proportion of the population being provided with a new way to process and organize information, especially abstract concepts. To adequately test this assertion we collected data in a non-Western forager horticulturalist society, the Tsimane of Bolivia, where rapid acculturation and formal education has altered the landscape of traditional living and localized learning. Drawing from a four year repeated measures design, as well as 10 year cross-sectional data, among Tsimane children, this study investigates changes in IQ scores within schooled and unschooled villages.
Our previous findings among the Tsimane demonstrated that (at the lower end of the educational continuum) better schooling and resources led to higher cognitive task performance. Following up on these findings, two predictions are tested: 1) Using a longitudinal design, children’s IQ scores in schooled villages will improve over time, while children in unschooled villages will show no increase; and 2) Using a cross-sectional design, a second cohort of children in schooled villages will show a greater average IQ than the preceding cohort.
Methods: Using Raven’s Coloured Progressive Matrices (RCPM) we tested 336 children at time 1 for a baseline assessment of cognitive performance across seven villages (four schooled & three unschooled). Three years later we began a longitudinal study and followed 104 children in four villages (three schooled & one unschooled) over a four year period. Finally, in the fall of 2017 we tested an entirely new cohort of 140 children in three villages (two schooled & one unschooled) used in the baseline sample 10 years earlier.
Results: To examine changes over time with a repeated measures design, we tested RCPM performance on 104 children over a four year period across the entire sample. We found no effect on RCPM across the sample by age (
β= -0.66, SE = 0.46, p = 0.151) or sex (β= 1.13, SE = 0.57, p = 0.138). (F(1, 75.52) = 7.28 , p = 0.009) when we looked at both schooled and unschooled villages. However, quality of school was among the greatest predictors of RCPM performance over time, with children attending higher quality schools scoring 4.6 points higher on the test than children attending moderate or lower quality schools (β= 4.60, SE=.57, p = 0.036).
Preliminary results of the cross-sectional design tested 120 children within three villages (two with a school and one without) with 10 years between the samples. We find an average 8 point increase in IQ between the first and second cohorts within schooled villages, which suggests a Flynn Effect. Surprisingly, however, the cohort's increase in IQ was much higher than observed increases in Western samples. We found no significant increase in IQ among unschooled children over that same time period.
Summary: Our initial findings suggest that among the Tsimane, pronounced performance gaps within their own communities across a four year period were dependent upon the presence and quality of schooling in their villages. Furthermore, dramatic improvement on cognitive scores over a ten year period were also observed. Overall, results suggest that access and quality of schooling is one of the greatest predictors of cognitive performance at the lower end of the educational continuum, and that school quality should be considered when comparing cognitive performance scores cross-culturally.
