The reliability of randomly selected final year pharmacy students in drug product assessment

Abstract

Employing ANOVA, factorial experimental analysis, and the theory of error, reliability studies were conducted on the assessment of the drug product chloroquine phosphate tablets. The G–Study employed equal numbers of the factors for uniform control, and involved three analysts (randomly selected final year Pharmacy students), conducting three tests each (drug content, friability, and hardness), on each of three identical samples taken from a single batch. The results were converted to percentages of “true scores” for uniformity and their collective statistical treatment. The highest source of variance, is the analysts\' factor which gave a component variance of 7,221 compared to 333 and 330 for the samples and tests factors respectively. The D-studies showed how to improve assessment reliability (R_A) of the students, and the developed equation below relating the numbers of samples (n_S), tests (n_T) and analysts (n_A), was used to calculate the number of students that would give the acceptable reliability level of 80%, for any given pair of numbers of samples and tests, and tabulated for instant reading off: 3,049.366 = 333.424/n_S + 330.234/n_T + 7,221.525/n_A + 272.84/n_Sn_T + 1,414.4/n_Sn_A + 2,295.13/n_Tn_A + 329.91/n_Sn_Tn_A. For one sample and one test, the needed six students for achieving the acceptable reliability, may be reduced more efficiently by increasing the number of tests rather than samples. Automation of assessment also improved the reliability of the G-Study from 0.798 to 0.979, and that of the D-Study starting point from 0.5 to 0.906.

Key Words: Reliability; Pharmacy Students; Drug product assessment.

Journal of Pharmacy and Bioresources Vol.1(1) 2004:17-23