Akademik Veri Yönetim Sistemi
8th International Researchers, Statisticians and Young Statisticians Congress (IRSYSC 2024), Adana, Türkiye, 28 - 30 Kasım 2024, ss.1
A
pilot study is a small-scale feasibility study planned to test various aspects
of the methods used before moving on to more comprehensive, rigorous, or
confirmatory main studies. These studies are used as an important tool for
early detection of potential errors, especially in studies that may impose
significant burdens in terms of time and cost. Researchers resort to pilot
studies to evaluate the applicability of the methods and procedures they plan
and to test the adequacy of data collection tools regarding suitability and
quality. Pilot studies contribute to correcting the deficiencies encountered in
this process and prepare the basis for a more successful main study. After the
necessary corrections are made in line with the findings obtained as a result
of the pilot study, the real study is started. This research explores whether
the quantitative data obtained from pilot studies can be integrated into the
main study and how this process might be executed. Three different cases are
considered: 1. The pilot study data are accepted as a part of the main study
and included in the main sample size; 2. The pilot study data are added as an
additional component to the main study and thus increase the main sample size;
3. The pilot study data are excluded and only data from the real study are
used. In each of these cases, the efficiencies of simple mean estimators and ratio
estimators are compared under simple random sampling. As a result of the
simulation studies carried out with synthetic data sets generated from both
symmetric and skewed distributions, it is seen that the case of adding the
pilot study externally to the main study (case 2) increased the efficiency of
the estimators in all conditions. Also, incorporating the pilot study data
directly into the main study as part of the overall sample (case 1) has been
observed to have a negative impact on the estimator efficiencies.