Accuracy evaluation total test approach

To bypass these potential problems, we manually browsed a small subset of the literature for papers that actually use computational tools. Rather than read through all the papers in the selected journals, a random subset of papers was examined to see which used computational chemistry techniques. To test the accuracy of this approach, the results obtained from one-quarter of the total number of issues in a given year was compared with the results found by browsing all issues in that volume. The test case was the 1994 volume of Journal of Organic Chemistry. The random selection yielded 14%, whereas inspection of all 1376 papers also yielded 14%. Thus browsing a fraction of the issues of a journal should suffice for our purposes. At least one-fourth of the total number of issues of the journals listed in Table 1 were read, but one-third to one-half were evaluated for the smaller journals. 

---