comparison of four search engines and their efficacy with emphasis on literature research in addiction (prevention and treatment)

conclusions as literature research is one of the most important stages of research, it&apos;s better for researchers, especially substance-related disorders scholars to use different search engines with the best recall, preciseness and importance in that subject field to reach desirable results while searching and they don’t depend on just one search engine. results results have shown that the search engines had different operations with regard to the evaluated criteria. since p value was 0.004 < 0.05 for preciseness and was 0.002 < 0.05 for importance, it shows significant difference among search engines. pubmed, science direct and google scholar were the best in recall, preciseness and importance respectively. background surveying valuable and most recent information from internet, has become vital for researchers and scholars, because every day, thousands and perhaps millions of scientific works are brought out as digital resources which represented by internet and researchers can’t ignore this great resource to find related documents for their literature search, which may not be found in any library. with regard to variety of documents presented on the internet, search engines are one of the most effective search tools for finding information. objectives the aim of this study is to evaluate the three criteria, recall, preciseness and importance of the four search engines which are pubmed, science direct, google scholar and federated search of iranian national medical digital library in addiction (prevention and treatment) to select the most effective search engine for offering the best literature research. materials and methods this research was a cross-sectional study by which four popular search engines in medical sciences were evaluated. to select keywords, medical subject heading (mesh) was used. we entered given keywords in the search engines and after searching, 10 first entries were evaluated. direct observation was used as a mean for data collection and they were analyzed by descriptive statistics (number, percent number and mean) and inferential statistics, one way analysis of variance (anova) and post hoc tukey in spss. 15 statistical software. p value < 0.05 was considered statistically significant.