Thursday, December 14, 2006

More Bad Interaction Design - University of Vermont

Part something in a continuing series about bad interaction design on university home pages.

OK, more nitpicking. Take a look at the University of Vermont home page. Mouse over the black and white photo. Kewl. It slides over to reveal a story. I don't really have a problem with that, though I wouldn't do it that way.

My problem is elsewhere. Look above the photo, just to the right of "University of Vermont". What does "✓FULL MEDIA" mean? Is it just a piece of system status information? Is it a link?

Ah! When you click on it, it changes to "✓ LOW MEDIA" and the Flash animation is replaced by a static image. "Full Media" really meant "High Bandwidth" and "Low Media" meant "Low Bandwidth". Why didn't they say so?

But what's this? If I click on it again it changes to "✓ TEXT ONLY" and just text is displayed instead of the image.

Cool!!! Or is it?

Not really. Why? Because I really had no idea that it was a link at first. And clicking on it didn't really function as we would expect a checkbox to operate. There is a well-defined and consistent checkbox metaphor on most Operating Systems. Checkboxes are supposed to be checked or unchecked but the options aren't supposed to change. This checkbox violates that model by leaving the box always checked but changing the option. That's wrong. You shouldn't mess with metaphors that people understand.

In addition, I had no idea what the options were. Once I clicked on it, I figured that it was just a toggle between high and low bandwidth. I had no idea that there was a further option. That's poor communication on their part.

How to Fix It
They have a couple of options.
  • Ditch the Flash animation. Stick with the static photos
  • Don't use a checkbox metaphor unless people can see the options and the box is going to function according to the standard metaphor
  • Show all the options. Perhaps moving it to the lower right-hand corner of the photo.
  • Change the words to match familiar terms: "Full Media" to "High Bandwidth", etc.
Moral of the Story
  • Don't use Flash animation unless it really adds something to the content.
  • Don't mess with metaphors people understand. It confuses and frustrates them
  • Don't hide options or people won't know what choices they have
  • Use terminology that people are familiar with

Here's one relatively unobtrusive idea, but certainly not the only - or even best - one. Notice that it uses radio buttons instead of checkboxes to be consistent with standard OS interface widgets.

Where's the Branding?

A good example of bad site branding.

OK, you have five seconds to tell me what school this is. Go!

Come on, folks! You really need to hit people over the head with your branding. Prospective students are going to look at dozens, maybe hundreds of college and university sites. Yours needs to be clearly identified and memorable.

What school is this? You have five seconds to tell me! didn't need even half a second, because the home page screams the school's name. Or at least, it screams the name of the state the school is in.

MECA is a tiny little school with no where near the name recognition that the University of Wisconsin has. They need to make a greater impression with their name and brand that UWisc does - by far! So why hide their name? Dare I suggest that an artist and not a Web designer designed the site? Perhaps that's unfair, but it wouldn't surprise me.

Cool Home Page Stuff - Student Blogs

A brief discussion of using student blogs on your home page.

On this last day before Christmas break (oh man, do I need it), I thought I'd start a recurring post on cool stuff universities are doing on their home page. This is just my opinion (and as we all know, I don't know jack), so feel free to ignore or deride it. I'm stealing most of this stuff from anyway.

Student Blogs

OK, these are university-sponsored student blogs, and so have some strings attached. Typically, students are compensated and there are often rules about not bashing professors by name or writing in gory details about the underage kegger/orgy that took place Saturday night.

A lot of schools have sponsored student blogs, but not that many feature them on their home page.

The value of sponsored student blogs are several:

  1. Students are blogging anyway, so you might as get involved and have at least a little control over the message.
  2. Prospective students are very tuned into social networking on the web, and getting the direct perspective of other students is highly valuable.
  3. Blogs personalize your Web presence, and can make your site appear more personal, friendly and honest.

The downsides include:

  • Picking the right students with the commitment to post regularly
  • Maintaining control over the message (at least to the point of preventing explicit or inappropriate posts)
  • Cost. Students are usually compensated, often with digital cameras and such that they use to maintain their blogs

Some sites, like Capital University, allow people to subsribe to the blogs, so that they can easily follow what's happening on campus over time.

Overall, most campuses that have used student blogs as a recruiting device have been very happy with the result.


Wednesday, December 13, 2006

Home Page Branding Photos/Spotlight Stories

An examination of how the main photo area on the home page is used.

In looking at the home pages of the Top 20 university sites, I kept looking at the main photo area of the pages. Though some of these pages may have many photos on them, there is virtually always a "main" photo prominently placed on the home page (see graph, below).

And, of course, I was supposed to look at that main photo. That's what it's there for: to catch and hold the viewer's attention. And as you can see, most universities give the most important real estate on the page to these photos.

So, guessing that it might be important, I decided to take a look at exactly what universities were doing with these photos.

First of all, I decided to add 10 universities to my list of the top 20. I guess I'm just getting bored with the same old 20 home pages. These were randonly picked out of the Google search results, somewhere on page 6 or 7. The additional 10 are as follows:
  1. USC
  2. Oregon State University
  3. Vanderbilt
  4. University of Connecticut
  5. Rutgers
  6. University of Wyoming
  7. Washington State University
  8. Michigan State University
  9. SMU
  10. University of Montana

All of the 30 sites had some sort of main photo area, even if it was relatively small. Of the 30 home pages, 7 (23%) were using a Flash-based animation for the main photo area. Twenty (67%) of the sites had photos that randomly changed, either as part of the Flash animation or upon refreshing the page. Six sites (20%) had photos that only changed periodically (once a day, once a week, etc.).

Twenty-three (77%) sites included at least some text with the photo, ranging from a short caption to a story headline and/or teaser. Half of the sites tied the photos to some sort of feature article or spotlight story, and even more put a link with the main photo area (67%). Those sites that were not using the main photos for a feature or spotlight linked to a variety of things, from a slide show of the campus to a mission statement to a eCard builder.

Photos of What?

The subject of the photos also varied, with some campuses focusing heavily on faculty and research and others more evenly spread out between faculty, students, etc.

Totals in the above graph don't equal 100% because most sites had changing photos that featured multiple subjects.

Of the 15 sites that used the main photo area for links to features/spotlights, 12 (80%) had photos of research or faculty. Only seven (47%) had photos/stories of students.

Of the 15 sites that did not use the main photo area for links to features/spotlights, 11 (73%) had photos of campus scenes (environment, buildings, etc.) and only 3 (20%) had photos of research or faculty.

People or Scenics?

On our campus, there is some discussion of what the ratio of people photos vs. campus/nature photos should be. The "consensus" (i.e., two people threw out that number) is that 3 to 1 is proper.

In order to see what other universities were doing, I tried to look at every photo on the sites (19 of the 30) that had randomly rotating photos (as opposed to photos that only changed daily). I broke down the photos into categories where the photos were primarily of people, primarily of the campus (though people may be in the photo), scenics (areas not on campus), sports and other (most commonly art or research oriented).

The results were as follows:

As you can see, people-oriented photos are only slightly more common than photos of the campus. If you add the sports-oriented photos to the people photos, the gap widens a bit. But overall, there is a fairly even balance between campus/scenic photos and people/sports photos.

To me, this makes sense. I never had any desire to visit the University of Wisconsin or the University of Montana until I saw just how beautiful their campuses and environs were - as revealed by the photos on their home pages. CSU, Chico has both a beautiful campus and a beautiful setting; those should be exploited in attracting prospective students to visit the campus.

Tuesday, December 12, 2006

Home Page Flash I Don't Like - UCLA

A look at a flash implementation at UCLA that isn't bad, but doesn't give the user enough control.

I'm not feeling so good today, and that makes me hate every Web site I see, so today I'm going to complain about Flash on university home pages. I know, yesterday I was all peaches and cream about the FSU home page, and I still like what FSU has done, but today I feel like whining.

Let's start with UCLA's home page. Overall, I really like the look of the UCLA home page. It's very clean. And their use of Flash isn't totally gratuitous. But there are a few things. First of all, what's with the "'Tis the season to be mellow"?? What a waste of real estate! Sure, stress management is important... but not on the home page of a major university!

But my main issue is with the Flash-based "Happenings" on the right side of the page. First of all, there really isn't anything to indicate that these are changable. Yes, there are up and down arrows, but they don't really scream "interaction" to me.

And then, if you do try to click them, they don't act quite as you'd expect an arrow button to work. In fact, you can't click on them. Mousing over the button causes the happenings to scroll up or down at a constant rate.

Though this is OK, it takes control away from me the user. I'm impatient. I don't want to sit there and watch the thing scroll at it's snail's pace. I want to control the speed. How about making it so mousing over the arrows causes it to scroll, but clicking on the arrows causes it to skip ahead to the next screen of happenings? That gives control back to me. I control my experience, not the Flash developer.

Monday, December 11, 2006

Flash on Your Home Page - FSU Does it Right

Incoherent babblings about using Flash on your home page, including one good example.

Everybody loves Flash animations. Everybody except usability designers, that is. Usability designers are into functionality, simplicity, accessibility - all things that Flash excels at doing poorly. Well, to be fair, it isn't just that Flash does these things poorly; it's more that it encourages Flash designers to do these things poorly.

Ah, but things have changed, you say. Flash is being used more appropriately and provides real functionality. In some cases, there is actually a tiny grain of truth to this. Let's go so far as to say that flash isn't being used quite so gratuitously these days.

It's not that I hate Flash - I used to be a Flash developer. But no matter the thousands of considerations you have as you build your university's Web site, remember that content is king. Forever and always. People go to any Web site for one of three reasons (sometimes all): to find some information, to conduct some business, or to have some social interaction. That's all about content. Sure, maybe that information is porn, maybe that business is online gambling, and maybe that social interaction is a singles chat room. But finding information, doing business and interacting socially are the three main classes of Internet activities. And they are all ruled by content.

What does Flash contribute to these activities? Not much usually. Sure, there's Flash video, but it ain't the only game in town for video. Yeah, there are Flash-based shopping sites, but how common are they?

Nope, Flash doesn't really contribute much to the content of a Web site. And whatever content your Flash doodad might have is inaccessible. Ditto for kewl animated Flash navigation. Oops! That's a no-no in the CSU.

So, death to Flash, right? Well, not so fast. Flash can have a role to play, and some schools and businesses are using it reasonably well in that role. Flash's strengths are in visual presentation, animation and interaction. Obviously, all of those can be good things on your home page if not over done.

From my perspective, Flash can be used well as a marketing tool on the home page, as an attention-getter - providing visually interesting views of the campus and campus life, teasers for spotlight stories, or other visually heavy but content light material. Certainly, it should not be used for core home page content or navigation (because of accessibility issues).

Take a look at Florida State's home page. They have a Flash-based rotating spotlight teaser, mostly about faculty research and achievements. What's nice about this?
  1. It's visual. Each piece has a photo related to the teaser.
  2. It's personal. Most of the pieces have a photo of the faculty member looking at (visually engaging) the viewer.
  3. It's prominent. It takes up the prime real estate on the page and the text is large enough to read easily.
  4. It's short. Though I think the text could be a bit punchier, the headline and teaser are short and to the point. A "More" button lets users see the rest of the story.
  5. It gives the user control. With built-in "previous", "next" and "pause" buttons, the user can skip to the next story, go back to the previous story or stop the animation altogether.
  6. It isn't annoying. The animations are simple and relatively unobtrusive. And there is no audio.

If you're going to do Flash on your home page, this is a great way to do it.

Personally, I would mix it up more. Fewer faculty and more students. I get the impression that FSU is trying to appeal to potential doctoral and post-doc students, and maybe potential research faculty. Maybe that's the kind of institution it is. I don't actually know.

But for Chico State...well, need I really say more? We're obviously going to be appealing mostly to potential undergraduate students, so highlighting student life (the beauty of the campus, local activities, etc.) and student achievements might be a better focus. A lot of it depends upon who the primary audience for our home page is.

Wednesday, December 6, 2006

Analysis of Navigational Links on Top 20 Home Pages - Part 2

An analysis of the labeling and terminology used in links on the top 20 home pages.

This Analysis is based partly on this post.

In Part 1, I looked at the most common navigational links on university home pages. In this post, I'll be taking a look at labeling and terminology.

Labeling is vital because the words and phrases that we use can have very different meanings to different people and audience groups. Ambiguous labels can confuse and frustrate users. And the use of non-standard labels can leave users lost at sea.

The value of understanding labeling is in being able to use terms and labels that your users both understand and expect. Some labels become so common as to be de facto standards; witness the "home page" and the "shopping cart". Everyone knows what these are, but there is no defined standard requiring that people call them that. Those standards arose solely through common usage.

In the same way that "home page" has become an ubiquitous label, other terms have risen to the point of de facto standards on university Web sites.

Academics and Research
"Academics" and "Research" are the two most common home page link concepts. As such, they need to communicate clearly, and you would expect - or at least hope for - some uniformity in terminology and labeling.

Term Number
Academics & Research 2
Academic Programs
Research & Innovation 1

Term Number
Research 15
Academics & Research 2
Research & Libraries 1
Research & Innovation 1

Both "Academics" and "Research" are overwhelmingly used in one permutation or another on virtually every site. Personally, I think that "Academics" seems potentially vague and ambiguous to me, but it is clear when you follow the link that "Academics" has a well-defined and accepted meaning. What do you find when you click on "Academics"? The following are the most common links:

  • Schools/Colleges
  • Departments/Academic Units
  • Undergraduate Education
  • Graduate Education
  • Advising
  • Centers & Institutes
  • Catalogs
  • Academic Support Services

It seems obvious that - unless we intend the term to mean something very different - we should follow the de facto standard that has clearly been created in labeling links to "academics".

“By creating a label we are building the site's information architecture. The labels define the categories of information, and the categories define the site's structure.”

Looking further into this you begin to realize that it isn't really just about labels and terminology; it's really about the structure of the site itself. By creating a label called "Academics" and placing the items in the above list there, we are building the site's structure. The labels define the categories of information, and the categories define the site's structure.

Audience Groups

Some audience group names can only be expressed in certain ways. "Alumni" are always referred to as alumni; "faculty" always as faculty. But other groups are a bit more open to interpretation. For example, there used to be some consensus that prospective students would be referred using that label, but more and more sites are now using "Future Students" to describe that audience group, though "Prospective Students" is still more common. This change may be occurring for several reasons. Universities may see "future students" as more hopeful, encouraging, and welcoming than "prospective students", or it may simply be a matter of trying to squeeze the phrase into a limited space.

Term Number
Prospective Students
Future Students

Once you get past "prospective students" to parents, friends, visitors, etc., it becomes something of a free for all.

Term Number
Visitors 7
Friends 5
Parents 3
Community 3

Other Terms


To me, "Athletics" and "Sports" carry two slightly different meanings, where "athletics" are the official university extramural sports and "sports" is that plus the recreational and intramural campus sports. But that's just my personal take.

Term Number
Athletics 10
Athletics & Recreation 2
Sports 2
Sports & Recreation 1

No real consensus here, though the word "employment" appears in eight of the labels.

Term Number
Employment 5
Jobs 3
Staff & Employment 2
Working at… 2
Job Opportunities 2
Employment Opportunities 1

Monday, December 4, 2006

Analysis of Navigational Links on Top 20 Home Pages - Part 1

An analysis of the most common links found on the top 20 university home pages.

The purpose of this analysis is to examine the navigational systems used on university home pages, in order to shed light on:
  1. Which links and types of links are most common
  2. Which types of organizational schemes are most common
  3. Which types of labeling and terminology are the most common

This analysis focuses only on navigational systems (e.g, "menus") and not on all of the links on each page. Links to specific news or feature stories, events, specific schools or colleges, video clips, etc. were not included in this analysis.


For this analysis I counted the number of times a particular link appeared on the 20 home pages. In some cases different terms were used to describe functionally equivalent things (e.g., "Athletics" and "Sports"). In those cases I group the items under the same heading. In addition, I only counted those items plainly visible on the home page; items occurring in pop-up menus were not included in the count (45% of sites used pop-up menus to one degree or another).

I did not always count every single link on each page. There were some links - that I either felt weren't relevant to my goals (e.g., links to specific news and feature stories, privacy policies, "About this Site" links) or that were unique to the specific site - that I did not count. In total I counted 78 different links, and a total of 493 links on the 20 home pages (an average of 25 links per page).

Loosely following Rosenfeld and Moreville's (2002) list of organizational systems, I divided links into one of four types:

  1. Topical/informational - pertaining to specific topics
  2. Organizational - pertaining to the organization of the university
  3. Audience - addressed toward specific audience groups
  4. Task - pertaining to actions and things that can be done

I chose these four types as opposed to Rosenfeld and Moreville's "topic", "task", "audience" and "metaphor" both because I felt that they ignored the fact that many Web sites organize their sites to reflect their internal organization, and because metaphoric organizational systems are pretty much a thing of the past (for good reason). These four types follow those used in my post on the Navigational Structure of the CSU, Chico Home Page.

Defining the types of some links was difficult. Is "Libraries" and organizational link or a topical link? How about "Directory"? I made my best guess and you may not agree


The links that I counted and their relative frequency appear in the table below.

LinkPercent of Sites
Contact Us75%
Campus Map55%
Prospective Students55%
Colleges & Schools50%
Medical Center40%
Continuing/Distance Ed35%
Financial Aid30%
Student Life25%
Academic Calendar20%
Campus Life20%
Centers & Institutes20%
Graduate Studies20%
Health & Safety20%
U & Town20%
Business & Industry15%
Find People15%
Site map15%
Class Schedule5%
Human Resources5%
Undergrad Research5%
Foundations & Corps5%


Far and away the most common type of link on university home pages are topical/informational in nature.

Type Total Percent
Topical/Informational 54 69%
Audience 12 15%
Organizational 8 10%
Task 4 5%

However, when you look at only the most popular links (occurring on 50% or more of home pages), a somewhat different picture emerges. Audience links double in frequency, accounting for 30% of the most popular links, and indicating that audience-oriented links are among the most popular. In fact, 18 of the 20 sites had a specific navigational structures for audience groups.

Type Total Percent
Topical/Informational 10 48%
Audience 6 29%
Organizational 3 14%
Task 2 9%

Other than that, there's no discernable pattern as to the nature of the most common links. Of course Academics and Research get a lot of play, but so do Jobs/Employment which is on a clearly different hierarchical level than the first two.

On the other hand, there is a long tail of links that occur on three or fewer sites, suggesting that there are many different potential approaches in building home page navigation and that different universities have different needs and priorities for their home page navigation.

Discovering and addressing those needs is perhaps the biggest challenge in redesign the CSU, Chico home page.

Rosenfeld, L. & Morville, P. 2002. Information Architecture for the World Wide Web. O'Reilly. 2nd Edition.

Thursday, November 30, 2006

So What's It All Mean? Part 2 - Most Viewed Pages

A discussion of the most commonly viewed pages on the CSU, Chico Web site.

This post is based on this analysis.

In some ways there is less to learn from the list of most viewed pages than there is from the most popular searches.

For one thing, many of these pages are buried deep within the hierarchy and are part of specific courses. Certainly, "Skull Model" (#14) and "Indian Caste System" (#17) are parts of very popular courses. The Skull Model has been around for years and has always been a highly viewed page.

One thing that is new is the rise of social networking pages, like "Progressive Students Union WIKI" (#2) and "Taiwanese Students Association" (#8, #19). Pages like these aren't part of the "official campus web presence" that is somewhat centrally maintained. So, though they are an important part of what is going on on campus, they aren't about to appear on the home page any time soon.

What's left is pretty much the usual suspects; library, class schedule, campus directory, catalog, etc. Student Computing is prominent, but that's because all of their labs computers use that as a home page.

There are a number of "deep link" pages in the top 30 most viewed pages. Some of these, like "Job Opportunities" (#21) are important enough to consider placing on the home page. However, most of the deep links can be reached via from the home page via a main page link (e.g., "Athletics - Baseball" can be reached from the "Athletics" link on the home page). The only real anomoly is the "Career Planning - Cover Letter Examples" page (#6). This page might have been used in a course or other activity, so without hard evidence I'd hesitate to consider placing a link to this on the home page.

Overall, based on the analysis, most of the bases are covered with respect to having the most viewed pages linked from the home page.

So What's It All Mean? Part 1 - Search Terms

A discussion of what to do about the most commonly searched for terms on the CSU, Chico Web site.

I'm a big advocate of the school of thought that says that your most popular search term is your hardest to find page. In other words, if people prefer to browse first and search after a browse fails (see this post), then a large portion of searches are for things that they couldn't find via browsing.

So, what does that mean? Should we put big flashing links for campus map, transcripts, and financial aid on the home page?

Maybe it does.

These are obviously high demand pages and should have high visibility. Of the top 30 search terms, only eight had visible links on the home page (campus map, class schedule, athletics, bookstore, portal, Orion, library, and catalog).

When you look at the placement of these eight links, you can see that all of them are either in the lower part of the page, or use very small text (or both in the case of "campus map").

Five of the links occur in the "photo/link" strip toward the bottom of the page, suggesting that perhaps this bar either is not highly visible or does not communicate effectively the fact that there are links there.

In any case, none of these links are prominently placed on the page, and this probably contributes to their popularity as search terms.

Another seven popular search terms appeared in the blue pop-up menus on the home page (financial aid, housing, jobs, academic calendar, records and registration, human resources, and associated students), indicating that the pop-up menus perhaps don't get as much use as we would like.

So overall 15 of the top 30 searches appear some place on the home page, but none of them have a high visibility.

I don't actually recommend big flashing links on the home page, but I do recommend that we take a closer look at who is doing the searching and whom the home page is serving.

A close look at who is likely doing the searching reveals the following about the top 57 search terms (based on this post):

Likely Searchers Terms Percent
Current Student 54 95%
Prospective Student 20 35%
Community 6 11%
Staff 5 9%
Alumni 3 5%
Faculty 3 5%

If, in fact, current students are likely to be searching for 54 of the top 57 terms, then I think it becomes clear that they are the major audience for our home page as it exists. It doesn't have to stay that way, of course. Humboldt State redesigned their home page to be a marketing tool for recruiting new students. That's great, but right now I think it's clear that current students are doing searches more than all other groups combined and would most benefit from a redeveloped home page.

IF the goal of the home page remains to serve current students, then a reorganization of the link structure to increase the visibility of the most popular search terms is an obvious idea.

How to do it?
Some campuses put a list of "Quick Links" on their home pages (University of Michigan, Princeton, for example); a list of supposedly popular links. But I'm not a fan of these for the same reason I'm not a fan of FAQ pages: it's an unorganized grab bag of stuff and there is no apparent reason for me to expect that what I'm looking for will be there. The "information scent" is too faint for me to want to go down that trail. In the case of the U Mich site, they at make them all visible (though they're in no discernible order).

Personally, I'm more of a fan of the approach that the University of Florida takes, which is to have main headings with a list of popular links below it, reminiscent of the old Yahoo! home page. Western Washington University does something similar. In fact, I suggested something similar during the last redesign in 2003:

What's the advantage?
The advantages of this approach IMO are:

  1. Links are part of a clear organizational structure (e.g., "campus map" could go under "About Chico State") instead of a haphazard list-o'-links.
  2. Links are plainly visible on the page without having to click on a "Quick Links" list or a pop-up menu.

The challenge is coming up with an organizational structure that includes as many of the high profile search terms as possible.

However, page structures like this have fallen out of favor in recent years, probably because people think that pop-up menus are more effective. This is certainly true from a space usage perspective, but I'm not actually convinced that this is the case from a findability perspective. My feeling is that pop-up menus are OK if the user knows they are there and if the menus are short and clear enough to be quickly scannable, but otherwise may actually reduce findability by making links on the home page less visible.

This is just one idea, and I'm sure that there are other ways to do this that present the links as part of an organized scheme and doesn't hide them in a drop-down list or pop-up menu somewhere.

Another Approach for Campus Map
The University of Wisconsin has put a link to their interactive campus map in a very prominent place on the home page, using an icon to draw attention. For something is as high demand as the campus map (and something that lends itself to a visual representation), this is not a bad approach. This wouldn't work for everything, of course, because the page would be littered with icons, but for the campus map (which is a perennial favorite), the idea has some merit and would make it clearly stand out.

Search vs. Browse

Insane babblings about why people use search instead of browsing, and visa versa.

Why do people use search instead of browsing or visa versa? We'll I'm sure that there are studies out detailing why, but I prefer to wing it based solely on opinion with no facts to support me. So here are my reasons why people prefer searching:

  1. They think it's quicker (but will try browsing if searching fails)
  2. They know exactly what they're looking for
  3. They don't trust (or have had a bad experience with) the site's browsing architecture
  4. They want to get instant results without a lot of steps (this is a combination of 1, 2, and 3)

Ultimately (particularly for a younger audience who are a) more web savvy, and b) want instant results without intermediary steps) searching is about getting to the desired information as quickly as possible without having to make decisions and without thinking about choices. This approach is referred as teleporting. Whether or not this is a good thing I'll leave to others to debate.

However, "The Perfect Search Engine Is Not Enough: A Study of Orienteering Behavior in Directed Search" (now I'm starting to do my research) suggests just the opposite, that people tend to browse first and search only when that fails, and that people tend to return to browsing (or "orienteering") even after a successful search.

One interesting paper (Effects of Scent and Breadth on Use of Site-Specific Search on E-Commerce Web Sites) found that the decision of a user to use search or browse approaches to finding information depended on several things, including:

  • How clearly organized and labeled a site's menu system was ("information scent")
  • How prominent search and browse areas were
  • The user's inclination to search or browse

Here's an article that summarizes the two papers mentioned above - that I found only after I found the other two.

The Upshot
What's the upshot of all this for administrators and others trying to plan a new website?

My take on it is this: If it's true that people prefer to browse (I plan to test this hypothesis), then we need to make absolutely certain that our site structure is extremely clear, well-designed, intuitive and easy to use. As Katz and Byrne say, "providing site search should not be used to compensate for poor menu design, and provide[s] further evidence regarding the design of effective menu structures."

Let's face it, a perfect search engine would always give you exactly what you wanted the first time every time. If that happened, no site would even have a menu; they'd all look like the Google home page.

But search engines aren't perfect for several reasons (ambiguity of terms, lack of context, limited algorithms, and poor labeling being just a few). And if people prefer to use menu systems over search, it is a condemnation of the ability search engines to return desired results, because we all know how poorly organized most menu systems are.

People are choosing the lesser of two evils in this case. And though there are a few things we can do to improve search results, we have much less control over that than we do over the information architecture and menu systems of our sites.

Katz, M. A. and Byrne, M. D. (2003). Effects of Scent and Breadth on Use of Site-specific Search on E-Commerce Web Sites. ACM Transactions on Computer-Human Interaction, 10(3) pp 198-220. (Link)

Teevan, J., Alvarado, C., Ackerman, M. and Karger, D. (2004). The perfect Search Engine is not Enough: A Study of Orienteering Behavior in Directed Search. Proceedings of ACM CHI 2004, pp. 415-4422. (Link)

Wednesday, November 29, 2006

Home Page Structure of the Top 20 Sites

Analysis of the layout, structure and space usage on the top 20 university home pages.

I was curious as to what were the most common page elements on the top 20 sites. How many sites included news or events or welcome messages, as well as how many used left hand navigation vs. top navigation, etc. The following analysis breaks down my findings.

Most Common Page Elements

ElementPercent of Sites
Search Box100%
Top Banner90%
Copyright Notice75%
Left Navigation70%
Quick Links45%
Top Navigation35%
Contact Info30%
Right Navigation15%
Inside "University"15%

Surprisingly, a search box and photos were the only things that all 20 sites had on their home page. But maybe that's not so surprising given the difficulty in organizing something as complex as a university website.

Of course, all sites also had identity/branding that identified the university on the home page.

Content Analysis

In terms of actual content (as opposed to structural or navigational elements), there was less agreement among the 20 sites. Only two items - News and Feature/Spotlight occurred on more that 50% of the 20 home pages, and the majority of content types occurred on 10% or less of the sites.

ElementPercent of Sites
Inside "University"15%

Page Element Placement

Another thing I was curious about was the placement of page elements on the page. I know that the issue of where to put the search box will inevitable arise, and I wanted to have hard data on what the "big boys" are doing.

Surprisingly, there were no real surprises. If you've read Neilsen, you realize that everybody is pretty much following the banner-on-the-top/navigation-on-the-left standard.

The following "heat maps" (OK, they're not heat maps in the traditional meaning) show where each of the most common page elements were placed on the page. To make these maps, I took a screen shot of each home page and then drew transparent rectangles over each page element. When I was done I removed the screenshots, leaving a stack of rectangles behind. Note that not every page was the same width, and I made the screenshots the size of the biggest page.

Top Banner
As expected, most sites use a top banner of varying heights, and most of them are placed - unbelievably enough - at the top of the page.

Here I'm only referring to the campus name and a logo or seal, if any. Some campuses put their logo in various places on the page, but almost all campuses put their name in the upper left part of the banner.

Search Box
Search boxes tend to be scattered around a bit more than the identity due to page layout considerations, but it's still clear that the most popular location is in the upper right-hand corner of the page.

Instead of trying to break down the navigation into different types, I just highlighted any area that was clearly navigational in nature. Many home pages had multiple navigation areas. Again, no surprise as the left side of pages and the area directly below the banner were the most popular locations for navigation.

Photos are used to create interest and connection on a home page, humanizing the content by bringing people/and or the environment onto the page. Photos also add visual richness, whatever their content. Many sites use photos as part of features or spotlights on the home page, so there is considerable overlap in the placement of photos and the placement of features.

Again, not surprising, photos - with their attention-grabbing abilities - are most commonly prominently placed in the upper center of home pages. Many sites had photos in multiple areas of the home page, serving a variety of purposes.

News and announcements (mostly labeled "News") were the most common content element on home pages. Though there was no overwhelming consensus on the placement of News, the most common location was near the center of the page, or on the center left, just to the right of the navigation.

Features or spotlights are either short articles or article teasers placed on the home page, mostly highlighting some interesting or newsworthy aspect of the campus. For example, Cornell spotlighted a talk given by Shimon Peres; UW spotlighted sports stories and student research in ANWR.

Placement of Features varied widely, but was most concentrated in the center of the page, and closely follows the placement of photos, since virtually all Features were accompanied by photos.

Events and/or calendars only appeared on about a third of home pages, but tended to all be placed lower center or lower right portion of the page.

Quick Links

Less than half of sites contained "Quick Links", typically as a drop-down list of links, but in a couple of cases as an actual list of links. The placement of Quick Links on the home page shows no discernible pattern.

The "Average" Home Page

On the "average" home page, 21% of the page area is taken up by photos, and 13% is taken up by navigation. Over a quarter of the page is taken up with white space or other content.

The following graphic shows the relative average size of each content area, in proportion to each other and the page.

The "Perfect" Home Page

Assuming that you wanted to build your home page based on the most common practices (which might or might not be a good idea), you would probably lay it out something like this:

Of course, this doesn't take into account the purpose of your home page, the primary intended audience(s) for your home page, your campus's particular branding, etc., etc.

Monday, November 27, 2006

Review of the Top 20 University Home Pages

Introduction to the analysis of the top 20 university home pages.

As part of the preparation for the redesign of the CSU, Chico home page, I decided to take a look at what other universities are doing on their home pages (following my personal mottos of "reinventing the wheel is for idiots" and "steal from those who actually know what they are doing").


I conducted a simple Google search for "university" on November 27th, 2006, and took the first 20 sites that came up. I choose this approach because I didn't want to bias the results by selecting site that I personally liked, and because I felt that the top sites listed in Google would have a level of prestige and legitimacy that a truly random selection might not.

The top 20 sites were, in this order:
  1. Harvard
  2. Stanford
  3. University of Florida
  4. University of Michigan
  5. Yale
  6. Cornell
  7. University of Delaware
  8. University of Washington
  9. Duke
  10. University of Texas at Austin
  11. University of Virginia
  12. Georgetown University
  13. Columbia University
  14. UC Berkeley
  15. Princeton
  16. Indiana University
  17. University of Pennsylvania
  18. University of Illinois at Urbana-Champaign
  19. University of Wisconsin - Madison
  20. New York University

I eliminated all sites with table-based layouts as being too old and non-standards compliant for consideration. This eliminated eight universities (highlighted in red, above). I then proceeded down the list of results in the Google search until I found the next eight CSS-based university home pages. I ended up looking at 45 university sites before finding 20 that were CSS-based, meaning that less than half of the top listed universities in the country have CSS-based home page layouts. The final list appears as follows:

  1. Stanford University
  2. University of Florida
  3. University of Michigan
  4. Yale University
  5. Cornell University
  6. University of Washington
  7. Duke University
  8. University of Texas at Austin
  9. Princeton University
  10. University of Illinois at Urbana-Champaign
  11. University of Wisconsin - Madison
  12. New York University
  13. Ohio State University
  14. Boston University
  15. University of Colorado at Boulder
  16. Florida State University
  17. Brown University
  18. University of Arizona
  19. Utah State University
  20. Syracuse University
Basic Analysis

I took a look at the DocType for each site and ran the W3C HTML validator and the Cynthia Says accessibility validator against each page. I also wanted to see the minimum width of each page (the point where content would be hidden and scrollbars would appear). Here's what I found:

University DocType Validates Sec 508 Min Width
Stanford University XHTML 1.0 Strict Yes Passed 789
University of Florida XHTML 1.0 Transitional Yes Passed 770
University of Michigan XHTML 1.0 Strict Yes Passed 820
Yale University XHTML 1.0 Transitional No Failed 783
Cornell University XHTML 1.0 Transitional No Passed 809
University of Washington XHTML 1.0 Strict Yes Failed 942
Duke University XHTML 1.0 Transitional No Failed 718
University of Texas at Austin XHTML 1.0 Strict Yes Passed 775
Princeton University XHTML 1.0 Transitional Yes Failed 796
University of Illinois at Urbana-Champaign HTML 4.01 Transitional No Passed 770
University of Wisconsin - Madison XHTML 1.0 Strict Yes Passed 762
New York University XHTML 1.0 Strict No Failed 815
Ohio State University HTML 4.01 Strict No Passed 767
Boston University XHTML 1.0 Strict Yes Failed 795
University of Colorado at Boulder HTML 4.01 Transitional No Failed 765
Florida State University XHTML 1.1 Yes Passed 797
Brown University XHTML 1.0 Strict Yes Passed 765
University of Arizona XHTML 1.0 Transitional No Passed 858
Utah State University XHTML 1.0 Transitional Yes Passed 788
Syracuse University XHTML 1.0 Strict No Failed 721

The most common DocType was XHTML 1.0 Strict, with nearly half of all sites using this DocType. Next most common was XHTML 1.0 Transitional, with 35% of sites.

Of the 20 home pages tested, only 11 (55%) actually validated, regardless of the DocType. Seven of the nine XHTML 1.0 Strict site validated, indicating that the people who maintain these sites are paying attention.

Section 508 Compliance
Only 12 (60%) of sites passed Cynthia Says automated accessibility validation. The most common problem was lack of labels on input elements.

Minimum Page Width
Minimum page width measures the smallest window width a page can viewed at before content begins to be obscured and horizontal scrollbars appear.

The narrowest page was 718 pixels wide (Duke); the widest was University of Washington at 942 pixels. The average minimum page width was 790 pixels, and the majority were between 760 and 800 pixels.

Use of Flash

Six of the 20 sites (30%) used Macromedia Flash technology on their home page. The most common uses were to provide rotating photos with stories and links and to provide pop-up menus. However, Boston University uses Flash to create an interactive recruiting environment with stories, photos, interactive maps, and multimedia pieces including video.

Wednesday, November 22, 2006

Most Popular Searches on the CSU, Chico Website

An analysis of the most commonly searched for terms on the CSU, Chico Web site, 2003 and 2006.

Part 4 in a series of analyses of the CSU, Chico home page and website that I conducted in 2003 and then repeated in 2006.


Grouped and Ranked Search Terms

The following list groups similar search terms from the raw data. Terms such as "map", "campus map", and "map of campus" are grouped together to give a clearer picture of the true ranking of search terms.

Rank Term Total Percent
1 enter search term 3132 4.81%
2 campus map 1695 2.60%
3 portal 1534 2.36%
4 financial aid 1441 2.21%
5 as bookstore 1047 1.61%
6 jobs 1030 1.58%
7 transcripts 818 1.26%
8 webct 722 1.11%
9 athletics 610 0.94%
10 housing 574 0.88%
11 tracs 469 0.72%
12 student employment 447 0.69%
13 class schedule 424 0.65%
14 tuition 421 0.65%
15 email 373 0.57%
16 parking 356 0.55%
17 orion 354 0.54%
18 textlink 287 0.44%
19 human resources 239 0.37%
20 cave 235 0.36%
21 grades 209 0.32%
22 health center 181 0.28%
23 calendar 176 0.27%
24 wildcat card 169 0.26%
25 study abroad 167 0.26%
26 library 161 0.25%
27 nursing 156 0.24%
28 sororities 151 0.23%
29 cns 147 0.23%
30 career planning and placement center 138 0.21%
31 academic calendar 134 0.21%
32 celt 133 0.20%
33 records and registration 132 0.20%
34 associated students 118 0.18%
35 user services 116 0.18%
36 seo 113 0.17%
37 writing center 107 0.16%
38 sports 101 0.16%
39 as presents 94 0.14%
40 getting connected 89 0.14%
41 work study 87 0.13%
42 open classes 82 0.13%
43 geography 77 0.12%
44 address 76 0.12%
45 application 75 0.12%
46 open university 73 0.11%
47 scholarships 72 0.11%
48 regional and continuing education 70 0.11%
49 directions 68 0.10%
50 adventure outings 65 0.10%
51 student computing 65 0.10%
52 fraternities 62 0.10%
53 catalog 60 0.09%
54 syllabus 59 0.09%
55 clubs 57 0.09%
56 fall 2003 54 0.08%
57 whitney hall 54 0.08%

These statistics are very revealing.

First of all, they tell us that the link to the Portal is not seen by students, even though it is right next to the search box.

Further, they tell us that the lack of links to "financial aid" and "campus map" on the home page are a problem for many students. These would be prime additions to the home page in a redesign.

"Employment", "jobs", "human resources", etc., are also popular, indicating that links to jobs and employment might not be amiss on the home page. Many university sites do this, and a few even do it well.

Also, the fact that "enter search term" every other term says that consideration should be given to adding a single line of JavaScript code that would remove that text as soon as the user clicked in the search field.

I repeated the analysis in 2006, the results of which are below.

Rank Search Term Total Percent
1 campus map 6530 2.24%
2 transcripts 5484 1.89%
3 financial aid 4701 1.62%
4 housing 4673 1.61%
5 jobs 4384 1.51%
6 academic calendar 3545 1.22%
7 webct 3246 1.12%
8 tuition 3144 1.08%
9 health center 3039 1.04%
10 final exam schedule 2881 0.99%
11 class schedule 2870 0.99%
12 career planning and placement center 2679 0.92%
13 calendar 2661 0.91%
14 athletics 2429 0.84%
15 as bookstore 2235 0.77%
16 study abroad 1994 0.69%
17 nursing 1992 0.68%
18 summer orientation 1956 0.67%
19 portal 1929 0.66%
20 orion 1802 0.62%
21 library 1681 0.58%
22 graduation 1668 0.57%
23 cave 1572 0.54%
24 records and registration 1529 0.53%
25 human resources 1508 0.52%
26 student employment 1486 0.51%
27 orientation 1398 0.48%
28 parking 1381 0.47%
29 as 1294 0.44%
30 catalog 1224 0.42%
31 general education 1109 0.38%
32 add/drop 1058 0.36%
33 eop 1029 0.35%
34 badm 101 1018 0.35%
35 admissions 970 0.33%
36 gpa calculator 952 0.33%
37 theme 877 0.30%
38 upward bound 839 0.29%
39 photos 755 0.26%
40 rec sports 743 0.26%
41 geography 722 0.25%
42 internship office 700 0.24%
43 logo 695 0.24%
44 peoplesoft 686 0.24%
45 bmu 645 0.22%
46 textlink 595 0.20%
47 intramural sports 591 0.20%
48 box office 547 0.19%
49 facilities management services 546 0.19%
50 psychology 531 0.18%
51 commencement 513 0.18%
52 research foundation 512 0.18%
53 scholarships 504 0.17%
54 university village 504 0.17%
55 sap 501 0.17%
56 regional and continuing education 497 0.17%
57 baseball 489 0.17%


Obviously, there were several changes in relative search popularity from 2003 to 2006, but no clear pattern is apparent.

Despite adding a link to Campus Map on the home page, "campus map" (or similar terms) is still the most popular search term. I attribute this to the fact that the link on the home page was essentially added as an afterthought in the footer of the page where it is not very visible.

The drop in "email" from 15th to 95th can probably be attributed to the change in the prominance of the email links on the home page. Similarly, the drop of "grades" from 21st to 87th is probably due to the fact that grades are easily accessible via the portal now, and most students are aware of that fact.

The biggest success story of the 2003 redesign was in adding a single line of JavaScript to the search box so that when a user clicked in the box, the words "enter search term" disappeared, resulting in removing "enter search term" from the top 100 search terms altogether.

The second biggest success story was changing the name of the link to the portal from "Chico State Connection" to "Portal", which is what everyone was calling it anyway. That moved searches for the portal from 2nd most popular to 19th.

Some unexplainable changes include the drop of "student employment" from 12th to 26th, the rise of "career planning and placement center" from 30th to 12th, the rise of academic calendar from 31st to 6th, the rise of "catalog" from 53rd to 30th, and drops in "user services", "sports", "getting connected", "work study", "open university", and several others.

Rank 2003 Rank 2006 Term
enter search term
2 1 campus map
3 19 portal
4 3 financial aid
5 15 as bookstore
6 5 jobs
7 2 transcripts
8 7 webct
9 14 athletics
10 4 housing
12 26 student employment
13 11 class schedule
14 8 tuition
15 95 email
16 28 parking
17 20 orion
18 46 textlink
19 25 human resources
20 23 cave
21 87 grades
22 9 health center
24 68 wildcat card
25 16 study abroad
26 21 library
27 17 nursing
28 63 sororities
30 12 career planning and placement center
31 6 academic calendar
33 24 records and registration
34 29 associated students
35 65 user services
36 59 seo
writing center
38 58 sports
39 60 as presents
40 91 getting connected
41 80 work study
open classes
43 41 geography
44 84 address
46 93 open university
47 53 scholarships
48 56 regional and continuing education
50 62 adventure outings
student computing
53 30 catalog
55 72 clubs
fall 2003
whitney hall


Overall, most terms that were popular in 2003 remained popular in 2006. Of the top ten terms in 2003, six were still in the top ten three years later.

Of the most popular search terms, virtually all of the top search terms appear to come from the current student audience group. Some terms (e.g., jobs, parking, financial aid, etc.) might also be used by other audience groups (community, staff, potential students, etc.), but it is impossible to separate those out accurately.

The table below gives a rough idea of the audience groups that would be most likely to use each search term. This is only my guesstimate, so you might have a different opinion.

RankSearch Term
Cur. Students
Pros. Students
1campus map


3financial aid



6academic calendar



9health center

10final exam schedule

11class schedule

12career planning and placement center



15as bookstore

16study abroad


18summer orientation






24records and registration

25human resources

26student employment





31general education



34badm 101


36gpa calculator


38upward bound


40rec sports


42internship office





47intramural sports

48box office

49facilities management services



52research foundation


54university village


56regional and continuing education


stu = current students
ps = potential students
alm = alumni
fac = faculty
stf = staff
com = off-campus community