The Accuracy of Self-Reported Achievement Scores among High School Students in Addis Ababa, Ethiopia

Abstract

Researchers in various disciplines, including psychology and education, usually use self-report measures
to collect data from research participants. They prefer to use self-report data because they are easily
accessible and more efficient. This study examined the accuracy of high school students’ self-reported
mathematics, English, and chemistry achievement scores from correlational and discrepancy
perspectives. Data were collected from 123 twelfth-grade students drawn from two (one public and one
private) high schools in Addis Ababa. Self-reported achievement scores were collected from the students
using a questionnaire whereas the corresponding scores were collected from official school records. The
results indicated that the correlations between self-reported scores and the corresponding actual scores
range from .59 to .81. On the other hand, the discrepancy perspective yielded statistically significant
mean differences for all pairs of self-reported and actual scores. The self-reported mean scores were
significantly larger than the corresponding actual mean scores. A significantly larger proportion of the
students over-reported their scores than those who underreported or accurately reported their scores.
Over-reporting of scores was more common among low achievers than among average or high achievers.
The number of students who accurately reported their mid-semester scores was larger than the number of
students who accurately reported their first or second semester scores. The overall pattern was similar
across the three school subjects. In conclusion, whereas self-reported scores were more or less valid from
the correlational perspective, their accuracy from the discrepancy perspective is concerning.