Designing studies to support or refute design is actually quite simple. The hard part are the pragmatics; finding a research scientist who is willing to collaborate, writing grant proposals for funding, access to technology such as eye-tracking equipment, finding someone to help you build the materials be they signs or simulations, finding someone to help you with the data analysis, finding an editor to help you write the manuscript and my personal favorite, getting ethical approval to conduct the study approval as you will be using human subjects.

If you can jump through all these hoops, then your work has a chance of getting published in a peer-reviewed journal. This is an enormous undertaking and would probably take ~three years before publication depending on how thorough your study is and who publishes it. I have no idea how much it would cost but I’d hazard >one million at least.

Experimental design is actually quite fun, and for any designer who favours a structured approach to problem solving, you will probably find yourself naturally drawn to scientific method. In this case an observation has already been made. Next we need to ask some directed research questions such as “Is this typeface better than Clearview Highway?”

Following this, we need to define what we mean by “better.” Better = Can someone discriminate words faster in low light?
Better = Can someone spot a word set in this typeface as opposed to that one?
Better = Can comprehend language in this typeface with greater accuracy compared to others?

Next, define the independent(IV) and dependent(DV) variables (the ones you change and the ones you measure).

Potential IV’s:
Typeface
Size
Lighting
Colour
Speed of vehicle

Potential DV’s:
Crashes (obviously in a simulation)
Eye movements
Detection
Comprehension
Reaction time
Search time

Next, make a testable hypothesis. A simple example would be “IF I change the typeface from, THEN I expect reaction time to decrease (measured by subjects ability to read key-words while speeding down the highway).”

Now you’re ready to design your experiment. For example, put someone in a car, place them on the track, ask them to spot a target word, change the font, repeat.

With an experiment designed you’ll be in a position to build your materials be they billboards or video games.

Following this, you’ll have to go through the process of statistical data analysis (I still need help with my math homework).

The comes the writing. And editing. And writing. And editing… This will take several rounds before it’s ready for submission, and once accepted, there will be several more rounds of editorial depending on the journal.

This type of study would make a wonderful contribution to the typographic community as it would give practicing typographers an awareness of a new (to them) method of problem solving and more important;y, provide data to support or refute the decisions and conclusions we traditionally make by intuition.

I would be more than willing to participate in a study like this.

——

The video does not work on my browser either: Firefox 3.5.8, OSX 10.6.2

I cannot find an example of the Legibility Test Tool

How do I upload an avatar?

Regards,
Vinicius

Did you look at the research of Jock Kinnear and Margaret Calvert: http://designmuseum.org/design/jock-kinneir-margaret-calvert

Nice article! I’m a student yet,so I didn’t know how many tests,and things you need to do for creating a typefont,especially if it’s for wayfinding. I just thought that was like ‘bigger,better’! hehe

You said in the ending of the article to contact you to receive a trial version of the font,if you can send it to me, I would enjoy!;~)

wonderful article.
Reading along, something occurred to me:
Dutch license plates have this weird ‘missing parts’ in the characters: http://sunpig.com/martin/archives/2003/09/20/dutch-car-license-plates-and-traffic-control.html
Can you find out with your Legibility Tool whether this actually helps to improve legibility from far away?

Onno