SACMEQ III Study Results

Achievement and test scales
Scale creation
  • An international reading performance scale and an international mathematics performance scale were constructed using item response theory (Rasch partial credit model).
  • For both scales, the international average is 500 points and the standard deviation is 100 points, with all countries being given equal weighting.
  • There are eight competency levels for both reading and mathematics tests.
  • Minimum and desired knowledge levels were developed for the HAK test.
List of achievement scales
  • Reading, mathematics, and HAKT achievement scores – Grade 6 pupils and their teachers
  • Reading competency scale – Grade 6 pupils and their teachers
  • Mathematics competency scale – Grade 6 pupils and their teachers
  • HAKT levels – Grade 6 pupils and their teachers
Questionnaire and background scales
  • Indexes were compiled to summarize responses to the background survey questions put to pupils, teachers, and school heads.
  • The computation of these indexes followed the same development process as the test scores (item response theory, the Rasch partial credit model)
  • To facilitate interpretation of the indexes, results were adjusted to an international scale for which the average is 50 and the standard deviation is 10.
  • School infrastructure index
  • Classroom equipment index
  • Personal possessions index
  • Academic qualifications index
  • Home condition index
  • Socio-economic status index
Overview of key study results


  • The results showed that existing HIV/AIDS prevention education programs in SACMEQ school systems were not effective because a majority of the Grade 6 (age 13) learners did not have the minimum level of knowledge about HIV and AIDS that is required to preserve and protect health.


Gender differences

  • In the sixth grade, in 2007, reading achievement was mostly higher among girls; mathematics achievement was mostly higher among boys.
  • Trend (2000 - 2007):
    • While gender equality in Grade 6 participation has improved in many SACMEQ countries, the size and the direction of gender differences in learning achievement in the SACMEQ countries remained stable from 2000 to 2007, no matter who (boys or girls) performed better.
    • In rural schools and lower SES groups across SACMEQ countries, gender differences in both reading and mathematics were slightly larger in 2007 than in 2000, while there was a slight improvement in urban schools and a reasonable improvement in higher SES groups.


School infrastructure and resources

  • There were considerable variations among school systems in terms of the condition of school buildings, provision of teachers, provision of toilets, and pupils’ and teachers’ behavioral problems.
  • School systems with high levels of teacher problems tended to have higher levels of pupil problems, and vice versa.


School heads and teachers

  • There were large variations in school heads among the school systems in terms of their personal characteristics (age and gender), academic education, pre-service training, and special training on school management.
  • There were variations in the characteristics of Grade 6 teachers across SACMEQ school systems in terms of age, gender, general education, professional training, and teaching experience.
  • Trend: There was an increase in the average achievement scores of teachers between 2000 and 2007; however, large variations in teacher achievement scores across SACMEQ school systems were noted.