A River Runs Through It:
Considerations and Issues when Evaluating StudentPortals

In October 2000, Stuart Brown and I talked at the NASPATechnology Institute about our shared desire for student affairsprofessionals to gain a better understanding of technology issues. Itwas from those discussions that this column evolved. The name of thiscolumn, A River Runs Through It, is borrowed from the NormanMaclean book of the same name. I must admit that I have not read thebook (or seen the 1992 Robert Redford film adaptation), but the titlehas always intrigued me. If a river runs through something, is itreally one thing, or two things separated by a river? It is in thatspirit that I would like to spend this time discussing studentaffairs and technology. Are these two things separated by a river, orone thing with a river running through it?

I tend to look at issues facing student affairs from more of atechnology perspective, and I'd like to share that perspective withyou all. I promise not to overwhelm anyone with technical jargon. Ithink you'll see as things progress that the gulf between technologyand student affairs isn't as wide as you may think. In fact, the moreI learn, the more I see that as student affairs professionals we havesome great theories and at experience at our disposal, we just needto understand how to apply all of that to our technologyneeds. This time around I'd like to spend some time talking about webportals.

By now, most of us probably have a pretty good idea what a website is, but portals are a fairly new concept for student affairs. Atits most basic, a portal is just a single place that a person (in ourcase students) can go to be presented with a wide variety ofinformation specifically pertinent to them. In many ways, a web pagewith a list of links is a very primitive portal, but it's not much ofone. A good portal builds on that by allowing the user to customizevarious parts of the site, from the "look and feel" of the site tothe actual information displayed. Commercial sites like myYahooare a good example of this kind of portal. Some higher educationinstitutions are getting into the game too, primarily with libraryportals (like North Carolina State University's MyLibrary)or services portals (like Virginia Tech's HokiePipeline). The truly excellent ones will even learn about theuser and begin presenting information that is similar to thingsthey've looked at in the past (this is called suggestive selling inretailing).

So how do you make sure your portal is truly excellent? Portaldesign (and, in fact, web design in general) is best evaluated onfour criteria: quality of content, quality of interface (ease ofuse), quality of infrastructure (speed of response), and degree ofcoupling (how the various pieces of the system interact). Write thesedown, tattoo them to your forehead, or engrave them on your desk -they are the four most important concepts of portal design. Most ofthe decisions you make during the design process will revolve aroundwhat trade-offs you are willing to make between these four conceptsto achieve your desired goal.

Quality of Content

It was not an accident that I listed content as the first area atwhich to look. With the proliferation of electronic media competingfor students' time, just because you build it doesn't mean they willcome, or more specifically come back. If the information orservices students want or need aren't where they are, they'll gosomewhere else, and if there isn't new or updated content or serviceson a pretty regular basis, they'll stop coming to the portal. Therearen't any real rules for how often, but my general goal is for astudent to visit the portal at least once a week. Exactly what kindof information and services should be available at your is going tobe up to you.

While we're talking about the quality of content, it's probablyworth noting quality's lesser known sibling: management, as inmanagement of content. An in depth discussion of strategies foreffectively managing the content of the portal is a bit beyond thescope of this installment, but here are some thoughts. Content shouldbe managed by (pause for dramatic effect) content experts.That's right. The people who know the most about a given topic shouldmanage the content for that topic. Of course that means your portalsystem has to make it easy to manage content because your contentexperts shouldn't have to be technology experts. In addition, a goodportal system will allow for distributed content management,as your content experts probably aren't all concentrated in onceplace.

Quality of Interface

If you have great content, but no one can figure out how to findit, your portal won't fare much better than if you had no content atall. That's where our second concept, quality of interface, comesinto play. How the content of the site is organized and displayedwill greatly affect the user experience. This isn't about how"pretty" or "slick" the portal looks, it's about how the designallows students to (or impedes them from) interacting with theportal. Dr. Richard Keeling of ReThinkInchas described today's students as "nonhierarchical, nonlinear,non-departmental" thinkers. But most of us (and our organizations)are hierarchical, linear, and departmental. The interface of theportal then becomes the bridge between their way of thinking andours. In the same way that we have content experts to manage thecontent of the site, so too should we have interface expertsto manage the interface of the site. There are people who get degreesin human interface design, so don't under estimate the importance ofgood interface experts, or the impact of bad ones. Just look at yourcomputer for an example. For you Windows users, what brilliant mindthought putting shutdown under the Start menu was a good idea? Forthe Mac folks, dragging a disk to the trash can to eject it iscertainly not the first, most obvious choice.

Quality of Infrastructure

So now we have great content and an excellent interface. But ifthe portal is slow to respond to requests or frequently unavailable,students will get frustrated and quit coming. That's where quality ofinfrastructure becomes important. When you're designing a portalsystem, there are a number of infrastructure issues to beconsidered.

• Bandwidth and throughput describe how much data the system can send and receive within a given time period and are often measured in megabytes per second (or if you're really lucky gigabytes per second). The design of your network, placement of servers, and type of hardware will all affect bandwidth and throughput. The good news is that you can almost always get the bandwidth and throughput you want, as long as you're willing to spend the money. The relationship between throughput and cost is roughly linear - the more you have the more it costs.
       
• Stability refers to how often the system is down, either unexpectedly or planned. The telecommunication industry rates stability by a scale of nines. Two nines represent 99% uptime (i.e. in a year the system will be down less than a total of four days), three nines represent 99.9% uptime (i.e. yearly downtime of less than eight hours), and four nines represent 99.99% uptime (i.e. downtime of only 52 minutes in a year). I doubt anyone in student affairs needs a four nine system, but a two nine system is worth shooting for.
       
• Scalability characterizes the ability of a system to grow to meet demand. We generally talk about it in terms of the number of simultaneous users that can be handled. A scalable system will allow you to build an installation that can handle 10 users but easily expand to handle 10,000 without having to redesign the whole system. The relationship between scalability and cost tends to be "stepped linear." As an example, the cost for up to 1000 users may be $10,000, while the cost for 1001 through 2000 users will jump to $20,000.
       
• Location of infrastructure describes where the system is physically, as well as logically, located. Many portal vendors offer to "host" the portal system for you at their location. While this can be a great solution for schools with limited infrastructure resources and expertise, there are some very real issues that need to be addressed if you choose to have parts of your infrastructure located off campus. As with content management, a full discussion of these issues will have to wait for another time. For now just realize that they exist.

Degree of Coupling

In physics and engineering, coupling defines the degree to whichthe function of one part of a system affects the other parts.Tightly coupled systems are ones in which a change of one partof the system will require changes to many of the other parts of thesystem to maintain functionality. Loosely coupled systemsallow for parts of the system to be changed or replaced withoutnecessitating wholesale changes to the entire system. For a portal,understanding the relationship between the content, interface, andinfrastructure will tell you a great deal about the degree ofcoupling in your system. Tightly coupled systems are bad. Youdon't want a system where a change in the interface requires majorchanges to the infrastructure (or vice versa), or, worse yet, majorrevisions to the content. Tightly coupled systems also tend torequire a single individual who is an expert in all three of theareas we've been discussing, and that's probably not realistic ordesirable. What you really want is a loosely coupled system whereexperts from the three areas can work somewhat independently on theirparticular issues (in the computer industry this is often calledseparating the business logic from the display logic).Obviously no system can be completely decoupled, or it wouldn'tfunction. The goal here is to have a system that is coupled in welldefined and reasonable ways and that allows everyone to do the workfor which they are best suited.

At this point I suspect you have almost as many questions as youdid at the start of this, but hopefully they're at leastdifferent ones. I'd like this column to reflect your concernsand issues, so let me know what you're thinking, what you'rewondering, and we can talk about it here together. As we all standwitnessing this river, most of you are probably looking from thestudent affairs side of the river across at the technology side. Goahead and wave, because I'm on the technology shore looking over atyou all. If you're willing to wade in and get wet, so am I. Maybe theanswers to all our questions are in the middle of this river - ifwe're just willing to look for them.