![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
blatherskiteA colleague over at LinkedIn asked me about whether there was any research to support the inclusion of lists of tables and figures (graphics) in the table of contents (TOC) of a large document such as a software user manual.
I'm not sure whether there's any research on this issue or not, but on basic principles, I'd be skeptical of taking any such research results as a given: the context of use changes so much between documents that findings for one context will often be irrelevant in another context.
I always approach these kinds of questions from a careful consideration of the user's perspective rather than from any particular theoretical orientation: Why is the user of the thing under discussion (here, the TOC) looking at it in the first place?
Questions in the original query:
1. Why have a list of tables and a list of images or any such list in the table of contents?
Because readers may know that they've seen a table or figure somewhere in the document, but be unable to tell from the chapter or section titles where that table might be located. Presenting all the tables and all the images in one convenient list eliminates the need to flip through the entire document looking for that one table or image they're seeking.
Of course, if you have time and money to pay a skilled indexer to index the document (rare these days), readers can find what they're looking for through the index. But even then, collecting the information together in a single place provides a benefit the index cannot provide: it reveals the structure and sequence of the document. In theory, you could infer the TOC from the index by rearranging all the page numbers into the correct sequence (sort of reverse-engineer or recompile the TOC), but that's an insane amount of work, and essentially impossible in practice. Easier to just produce the TOC in the first place.
2. What will be the arguments for and against this practice?
The arguments in favor, in addition to the aforementioned: Those who don't need this resource will ignore it; those who do need it will love you for providing it. Since you can generate the TOC and lists of tables and figures automatically, with no human input required other than the initial coding of the paragraph style definitions to include these elements in the TOC, labor requirements are nil. (In fact, some software offers predefined table and figure caption styles that are automatically included; you don't even have to manually set them to be included.)
The arguments against: Obviously, you increase the length of the document by a few pages, thereby increasing the cost. The increase is probably trivial for any document long enough to need these lists. Of course, if you're talking about an online document (increasingly the case these days), cost is irrelevant.
As always, the answer to such questions doesn't depend on any abstruse theory: you can get there yourself just by asking what the audience hopes to achieve, and spending a few moments thinking about how to help them achieve it.
I'm not sure whether there's any research on this issue or not, but on basic principles, I'd be skeptical of taking any such research results as a given: the context of use changes so much between documents that findings for one context will often be irrelevant in another context.
I always approach these kinds of questions from a careful consideration of the user's perspective rather than from any particular theoretical orientation: Why is the user of the thing under discussion (here, the TOC) looking at it in the first place?
Questions in the original query:
1. Why have a list of tables and a list of images or any such list in the table of contents?
Because readers may know that they've seen a table or figure somewhere in the document, but be unable to tell from the chapter or section titles where that table might be located. Presenting all the tables and all the images in one convenient list eliminates the need to flip through the entire document looking for that one table or image they're seeking.
Of course, if you have time and money to pay a skilled indexer to index the document (rare these days), readers can find what they're looking for through the index. But even then, collecting the information together in a single place provides a benefit the index cannot provide: it reveals the structure and sequence of the document. In theory, you could infer the TOC from the index by rearranging all the page numbers into the correct sequence (sort of reverse-engineer or recompile the TOC), but that's an insane amount of work, and essentially impossible in practice. Easier to just produce the TOC in the first place.
2. What will be the arguments for and against this practice?
The arguments in favor, in addition to the aforementioned: Those who don't need this resource will ignore it; those who do need it will love you for providing it. Since you can generate the TOC and lists of tables and figures automatically, with no human input required other than the initial coding of the paragraph style definitions to include these elements in the TOC, labor requirements are nil. (In fact, some software offers predefined table and figure caption styles that are automatically included; you don't even have to manually set them to be included.)
The arguments against: Obviously, you increase the length of the document by a few pages, thereby increasing the cost. The increase is probably trivial for any document long enough to need these lists. Of course, if you're talking about an online document (increasingly the case these days), cost is irrelevant.
As always, the answer to such questions doesn't depend on any abstruse theory: you can get there yourself just by asking what the audience hopes to achieve, and spending a few moments thinking about how to help them achieve it.
Figure and table captions
Date: 2010-09-09 02:31 pm (UTC)Jim
Re: Figure and table captions
Date: 2010-09-09 04:06 pm (UTC)In a TOC, the order follows the physical structure of the book and the search strategy is predominantly linear (from the first page to the last). The subject matter order should (ideally) have been determined based on some defensible set of criteria when the book's sequence of sections and chapters was first designed by the authors (i.e., at the outline stage). In that case, the list of tables and figures will follow a logical order to the same extent that the book itself follows a logical order. Additional navigation aids could be added to the list, such as chapter titles (or short versions of those titles that focus more narrowly on the content), if you're willing to do a bit more work on the TOC to increase its usability.
In an index, the organization is topic-based and nonlinear, with entries (topics) appearing in alphabetical order, ideally with one or more synonyms for readers who think of different keywords when they search for specific topics. In that case, no additional order beyond alphabetical is required. If you're "doing it right", you've either embedded an index by creating index entries as you write, or you've created an index after the main text is complete. The latter is a rare luxury these days. But in either case, topic-based access to the tables and figures will automatically be provided by the index.
For a book that emphasizes reference tables or reference figures as the key content rather than as support material (i.e., you will frequently refer to these aids rather than only using them when you read a procedure), being able to access the tables or figures quickly is an entirely separate task, and the document should be designed to support that task. Examples of such documents include lists of physical properties for chemical compounds (e.g., in the IRC handbook) or graphics of control layout and settings for a complicated device (e.g., the old devices that used DIP switches to set their behavior). At a minimum, it should be possible to easily find specific tables and figures using some kind of lookup aid—such as a separate TOC for tables and figures.
An example of this would be the "abstracting" journals that list all papers published by a category of researcher (e.g., biological) in a given year. These often include separate indexes for author names and keywords, since users will look for new papers based on the names of key researchers in their field and based on the keywords that appear in the journal articles.