Skip all navigation and go to page content
NN/LM Home About PSR | Contact PSR | Feedback |Site Map | Help

Getting the most from PubMed indexing: Announcing the availability of NLM’s online indexing training module

PubMed is designed so that the novice user can simply type search terms in a box with automatic mapping to MeSH terms. Therefore, some ask: “Do I really need to know how articles are indexed?”

“Yes, you do!” experienced searchers generally agree. They know how often they are challenged to find the five most useful references from a database of over 17 million citations. They also intuitively recognize that a feature such as PubMed’s Related Articles is shaped not only by the occurrence of words in article titles and abstracts but by the use of MeSH headings and subheadings.

Therefore, we were delighted to learn that the National Library of Medicine has now made available, freely over the internet, the online indexing training modules that are used to train indexers at NLM. In addition to providing information on the principal concepts of indexing, these training modules include online exercises that may be beneficial to both individuals learning about indexing as well as online searching. These modules are very useful to the online searcher including those who have had some exposure to MEDLINE indexing or who have search experience.

You will find the training modules by going to: http://www.nlm.nih.gov/bsd/indexing/index.html. Although the site looks deceptively simple, it is chock full of useful information and exercises to reinforce indexing principles and practice. The first lessons deal with the indexing process, how to scan an article for indexing, and how indexing is different for the main points of an article versus the minor points. Then you proceed to an in-depth presentation about check tags – age groups, human versus animal, gender, historical chronology tags, etc. and indexing principles relating to these tags. You then learn about the hierarchical nature of MeSH, the rule of three, the application of subheadings to MeSH terms, and how to read a MeSH record. Then you will be taken to all 15 topical categories of MeSH where you will be able to study indexing principles for the category; there are exercises for each category. Finally you will review publication types and gene indexing. Whew! This takes time!

In the last analysis, I was reminded of why PubMed/MEDLINE does so well – its rigor, depth of MeSH terms, use of check tags, major and minor points of an article, consistent indexing of human/gender/age, subheadings (qualifiers) — in summary the quality and depth provided through its indexing policies, process, and application. I appreciate even more that PubMed is a very powerful system.

My suggestion: take a day or two to study the online indexing training tools. If you can’t devote an entire day, plan to spend a few good thinking hours spread out over several days. Go through the exercises. Take the quizzes. Use the MeSH Browser to select terms. Read the definitions and indexing notes. Think about what you are doing. Reflect on what you’ve learned.

Then, if you want a challenge, test your own indexing skills. Pick an article from your collection – one that’s been out for at least a month so that it’s likely to have been indexed already. Without looking at indexing in PubMed, scan this article using the training guidelines and try indexing them yourself. Use the MeSH Browser to find appropriate index terms. When you have finished, use PubMed’s Citation Format to see what index terms were applied. Undoubtedly, there will be differences between what you did and what the indexer did. What is different? Did you capture the main points of the article? Did you use correct heading/subheading combinations? What surprises did you find? Why do you think the indexer chose the terms he/she did? Read the indexing annotations for the terms the indexer used. If you are really brave, try two more articles.

Even professional indexers will vary in how they index an article. You should not be surprised if your indexing is different than theirs. Perhaps the greatest challenge you have as a searcher is thinking about how to account for these differences and to take them into account in your searching. You might even want to put together a study group – get a group of colleagues who want to be better searchers, take the online indexing course, and then compare notes on how each of you would index a series of articles. Is anyone up to this challenge? If so, let us know!

Here are some additional resources: