Others, Miscellaneous, and the Like...

By Gerd Waloszek

Sometimes, I receive tough questions in my inbox, as the following example shows:

I read several of your articles on the Web and, although you do not discuss what the selections in a dropdown should be, I thought you might be able to assist me.

I am charged with finding research to support the contention that the selection choice "other" should not be included in a dropdown menu. I support this contention, because it has been my experience that users will select that choice because they lack the time to make a more careful selection.

If you could direct me to anyone or any institution that has published anything in this regard, I would be very appreciative.

That was a tough question – at least, I had never read anything about that problem in the literature. To make sure that the enquirer had really searched the Web and not just sent me an e-mail, I started a Web search on my own. But I could not find anything useful and gave up. As I did not want to leave the questioner out in the rain, I sent him my personal opinion about the matter:

I am always glad to find that category in a selection list because:

It does not force me to select a category that does not fit or that I do not want to select

It allows me to "hide" my opinion/state/etc. In many cases, I do not want to expose myself to the company that wants me to fill a form (e.g. when registering software). "Other" is a nice and humane way out...

I seldom select this category because I do not want to read the others – but other people may well do so...

For me, the second point is the most important one. Often, companies require me to answer questions that I do not want to answer (and that are not at all necessary for the respective purpose). "Other" allows me to "save my face"...

This reply did, however, not satisfy the needs of the enquirer, and he made his case more clear to me:

While I certainly agree with you that there are times the selection "Other" is a valid choice, I am concerned with indexing and retrieval of information. We are establishing an interface for users to index content for a central repository. They will be using a controlled vocabulary. (Those searching will be using the same vocabulary.) For example, when an indexer is assigning a "document type" to a piece of content, if he/she selects "Other," that piece of metadata information is useless. I have seen this before in our archives. Half of the documents are assigned the "Other" choice in the document type field. This is not real useful when performing a search for a particular type of document. And since I am working with an identifiable group of people and identifiable materials, if there is a document that does not fit in the available choices, there is always the option of adding a new choice.

Although I would never use "Other" as a selection in indexing, I find myself in the position of having to "prove" my position.

So, the ball had been rolled back to me but I still did not have an answer – only an opinion, namely that there is a difference between the users' interests and the indexers'... My last hope and resort was a textbook, which offers a wealth on empirical results, Jens Wandmacher's book "Software-Ergonomie" from 1993. As the date indicates, it's definitely and regrettably a little bit dated... While the book has nothing to say about dropdown lists and their options, it contains an extended section on menus. I scanned that section eagerly and finally found a short paragraph on menu categories. There, Wandmacher cites a report by Dumais and Landauer from 1983 and makes a short remark on the "Miscellaneous" category. My even shorter translation for the enquirer was: According to Dumais & Landauer (1983), removing a category "Miscellaneous" can increase the hit rate for other categories. In other words: it's no good. Then I searched the Web for the original article, found it in the ACM portal, downloaded from there, and sent back my answer to the enquirer. After taking a look at the article, he answered: "This article is as close to what I need as anything I have seen. Thank you very much." So, I had at least come close to what he needed. However, as I am not a native English speaker, I had to ask a few colleagues, what the first sentence really meant...

So far, so good – but an answer to the original question has still to be found...

Afterthought

I thought about what would happen if, for example, a software company would recommend an "Other" category in its UI guidelines. Would that help designers, or might they run into problems even with well thought out rules? Perhaps, they might – as the following examples with "mimicked" dropdown lists indicate:

This one should be OK:

Here we might already start to brood – it looks like a "philosophical" question:

Last, but not least we might run into the equivalent of an "impossible figure":

Your Favorite Color:

Red
Green
Blue
Yellow
Other

Gender

Female
Male
Other

Your Yearly Income:

Less than $100,000
$100,000-$500,000
$500,000-$1,000,000
More than $1,000,000
Other

By the way, I would select "Other" because I am paid in Euros...

After all, these examples tell us that purely "syntactical" rules would not be sufficient...

References

Dumais, Susan, Landauer, Thomas K. (1983). Using Examples to Describe Categories. In: Smith, Raoul N., Pew, Richard W., Janda, Ann (ed.): Proceedings of the ACM CHI 83 Human Factors in Computing Systems Conference. December 12-15, 1983, Boston, Massachusetts, United States. p.112-115.
Wandmacher, J (1993). Software-Ergonomie. Berlin: De Gruyter, ISBN 3-11-012971-X

Originally Published: 07/19/2006 - Last Revision: 02/01/2009