How can we design for unsophisticated computer users?

Question

There's a thread on redit about computer users doing things in very unproductive ways. The first example:

My office mate is 62 and overdue to retire. If she has a website address, this is how she gets there:

Opens XP Internet Explorer, inhabited by so many toolbars that the actual display window resembles the eye slit on a hunting blind.

Uses Yahoo! toolbar to search for www.Yahoo.com.

Selects Yahoo! search engine from top of search results.

Types URL she has been given in full into Yahoo! search.

Selects random results on front page, and if they do not open desired website, closes IE and starts again from step 1.

Another example from the same thread:

I occasionally come back to my desk to see a post-it-note left by a team leader from elsewhere in the business.

The post-it-note says "See me please.", and is attached to a sheet

of paper. The sheet of paper is a print out of a Word document. The Word document contains a screenshot of Outlook. The screenshot of Outlook is showing an email I sent to the team leader. More precisely, it's a screenshot of her REPLY to my email that she has decided to screen-shot, paste, print, hand-deliver (from a different floor) and add an explanatory post-it-note to me instead of pressing Send... The reply reads "I am not sure what you mean here."

The phone rings before I can email the team leader, "Hi, did you see my note? I am not sure what you meant in your email..."

Are these just special cases? Does it fall outside the scope of UX and/or usability?

PS: First question here, so I realize it might not be appropriate.

Answer 1

Generational related differences in computer skills are well known and well documented.

It's also well known that users stick to suboptimal strategies when searching on the internet.

So those kind of behaviors aren't new things - just more extreme examples. Reading about seriously sub-optimal strategies can only be useful indirectly. It's not useful, profitable, and probably not possible to significantly improve things for these users. Our ROI is better where the UX changes benefit more users. However, the more extreme examples may give us insights that we wouldn't get another way.

How likely is that?

Take the quoted example as a test case. It's interesting to try and deconstruct how such a broken approach could come about. Here are some guesses that could be fed in to an interview plan relevant to a wider class of user.

• Why so many toolbars? Perhaps the Yahoo toolbar was installed by someone helpful, setting up yahoo mail for her? What factors make her like toolbars? Is the use of large buttons on the toolbar a factor - and is this related to eyesight?

• Follows a set strategy that 'works for her' without deviation. Some of the elements, like typing a website address into a search engine are quite common behaviors. Why do people do it? Usually because they are unsure of the exact URL, or their typing is not 100% accurate and their eyesight not good enough to catch errors. Type www.hootmail.com into google and it will give you hotmail sign-in as the first result. Doing this in the search bar of older browsers would not.

Is it worth following up this line of questioning? For the bulk of UX work, no. If we're designing a website specifically for seniors, then absolutely yes. We need to think more about users with poorer than normal vision. We need to consider that they might have a helper surf the site for them. We need to be aware they are more likely to be on an older version of the browser. We need to be aware that their strategies may have been formed for earlier versions of the internet and won't adapt. We're extremely unlikely to see the extreme behavior, but we are likely to see some contributing components.

These kind of things (eyesight issues, fixed strategies) can and have been found out over and over again in user tests. However, attempts to deconstruct the more extreme examples may guide us to ask questions sooner than we would from user tests. We won't find the answers until we ask the right questions, and we won't ask the right questions until we know to. Would the frequency of older people having some helper surf the site for them come out as easily in user tests? Probably not, as there would be a recruitment bias to people who can use the internet on their own.

Answer 2

Given that you rarely see people behaving in such a fashion, this type of behaviour is, with respect to the general populace, by definition, a special case.

That said, the behaviour of the general population is irrelevant. The behaviour of typical users of that which you are testing is all that is relevant.

Your main concern should be the consideration of whether such behaviour is significant to that which you are testing.

Despite careful test participant selection, you may still end up observing a user who appears incapable of performing the most rudimentary tasks in anything approaching a standard form. If shown to be statistically insignificant, data collected from such users would be invalid.

No form of behaviour is beyond the scope of UX or the study of usability. All forms of behaviour are relevant but only with respect to the target audience of that being tested.

We cannot learn anything valid from this one special case for two reasons: firstly, we can never learn anything from a single data point; secondly, we can learn nothing of such forms of behaviour outside the scope of testing users of a concrete product or service.

The most we can learn from this one special cases is that perhaps sometimes some people behave in ways that are far from standard.

Answer 3

Much of user experience is based on conventions, what you expect the user to understand without further instruction. If we lower the common denominator to the level of your examples, people who have next to no understanding of the technology they are interacting with, then designing an interface would be next to impossible without pages of help text and tutorials.

I think you need to ignore these cases as they vast extremes of the users you will typically need to deal with. Concentrate on the main stream user base instead.

Answer 4

The Redit thread demonstrates that many people have :

• little training on using computers

• that software comes with poor or non-existent instructions

• that even if software comes with instructions - people don't bother reading the instructions.

Answer 5

Old people stick to habits. People have habits. People resist learning new information if it contradicts what they know; the older they are the harder it is, some people are extremely stubborn.

Users give a f**k unless it does not works. This old dude find how to solve a problem, memorized it, uses it and it works. He does not care that there is some cooler way to do it. He just wants to solve his problem. And most likely his way.

Old dude found that way himself or somebody taught him. In first case, trying to show "the better way"(tm) might be understood as making him look dumb.

This old dude might be seen as creative because he found solution within set of limited tools. It is like painting Mona Lisa with MsPaint, You have to know Your tool perfectly. :)

Also note, that User Interface "experts" are obsessed with fallacy "newcomer does not understands it in a few minutes, therefore it is bad in general". It is just a heuristics, a guess, nothing more. Some good tools are hard to learn.

So Yes, it is about UX and about users.