The aim of this study was to define the level of quality and types of analytical errors for measurement instruments used [i.e., interview forms, achievement tests and scales] in doctoral dissertations produced in educational sciences in Turkey. The study was designed to determine the levels of factors concerning quality in research methods and the case study model was used. Theoretical universe for the study was 324 doctoral dissertations in educational sciences in Turkey from 2003 to 2007. Sampling group was consisted of 211 doctoral dissertations accessed through online in the National Thesis Center. In order to collect the data, an evaluation form was developed by the researcher and the data analysis method was epistemological document analysis. In the analysis process, frequencies, descriptive statistics, and typology analysis techniques were used. The findings indicate that the properties of measurement tools used in dissertations in educational sciences were absent and that the most common analytical mistake was the absence of validity.