FamilySearch recently reached a significant milestone: one billion images of historical documents are now viewable on FamilySearch.org. That’s one billion pictures of documents. Of those images, how many would you say are indexed and searchable by name? All of them? Half of them? Would you believe less than 22 percent?
Clearly more indexers and arbitrators are needed, and many are responding to the call, but as important as their participation is, making the best use of volunteers’ gifts of time and talent is equally important. Our current process (two volunteers index and another arbitrates), which will continue to be used as needed, produces high-quality indexes but is labor intensive and can be inefficient for some types of records.
FamilySearch engineers are constantly researching more ways to improve and simplify the indexing process. The goal is to improve the indexing program and indexing processes to make the work of every indexer and arbitrator go farther and count more, while meeting or even exceeding the high standard of quality patrons have come to expect from FamilySearch.
As you read the “progress report” below detailing these advancements (most coming in the new indexing program later this year), we think you will agree that some fabulous, even inspired, ideas have come forward. Most of these new processes will be seamless and largely invisible to volunteers.
Single-key indexing of selected projects. Simple projects (especially printed or typewritten records) that are indexed accurately by the majority of volunteers may be transcribed only once. As with all projects, the quality will be monitored closely. Audit processes catch major indexing mistakes. If the quality of a project drops, a peer review or arbitration process will be added. Research shows that using this process for easy projects produces an index of comparable quality to the traditional double-key-plus-arbitration process with less than half the work. More importantly, this approach frees experienced volunteers to focus their advanced skills on more difficult projects.
Single-key indexing plus peer review. In this process, which will be introduced as part of the new indexing program, a single volunteer will index and submit a batch, and a second indexer will review the completed work. For some projects, the person reviewing can add corrections to the values entered by the first indexer, and both values can be included in the searchable index. Volunteers will need to index a minimum number of records (yet to be determined) before they will be allowed to review others’ work. To maintain the focus on quality, the new indexing program will also assign batches submitted by newer volunteers to more experienced indexers for review.
Single-key indexing by field. Some fields are very straightforward, such as the Age or Event Year fields. These fields are generally transcribed accurately by most indexers, without any kind of review or arbitration needed. In this new process, the system would bypass additional steps for these fields, allowing volunteers to continue to review or arbitrate other fields where errors are more likely to be found.
Single-key indexing by qualified volunteers. As volunteers gain significant experience and expertise, they sometimes reach a point where review or arbitration of their work actually increases the likelihood of introducing errors. The new indexing program can recognize volunteers who meet this level of quality and route their work directly to publication, bypassing review processes altogether. As with all submitted work, periodic audits would help ensure that the quality of their work is consistently high.
Stay tuned for more information about how these and other ideas may help magnify your efforts to increase the pace of quality record publication on FamilySearch.org.