The Effect of Sample Size Differences on Calculating the Conditional Standard Error of Achievement Tests

The aim of the research was to study the effect of sample size differences (200, 400, 600) on calculating the conditional standard error using the Lord and Keats methods in achievement tests according to the classical measurement theory. The results showed that the difficulty coefficients ranged between 0.42 and 0.73 with an average of 0.60, and the discrimination coefficients ranged between 0.17 and 0.47 with an average of 0.36, with the exception of two items that did not meet the conditions. All the wrong alternatives were effective with negative values, and the Cronbach’s alpha reliability coefficient was 0.879, The Lord and Keats methods were compared using arithmetic means and standard deviations, and the results showed statistically significant differences in favor of the Keats method due to the lower average standard errors compared to the Lord method. The effect of sample sizes was also studied using one-way analysis of variance, which showed statistically significant differences between the three sample sizes (200, 400, 600). The results showed that the conditional standard error decreased with increasing the sample size, and the larger sample size (600) was the better, the researcher presented a number of conclusions, recommendations and suggestions.