Reading-to-learn and cognitive loads

While I was in Vancouver, you may remember, I read a great article about cohesion and coherence from the ISTC magazine. I mentioned that there was an STC Journal article about a year ago that related but I couldn’t remember how.

I found the article.

It’s The Effect of Heading Frequency on Comprehension of Print Versus Online Information by Alexandra L. Bartell, Laura D. Schultz, and Jan H. Spyridakis. Technical Communication, Vol 52, no. 4, November 2006. pages 416 to 426.

So, what’s the article say, Sharon?

Oh, many interesting things. Get another cup of coffee. And cookies. Everyone needs some cookies.

Except the woman who stopped in front of me this morning on the freeway and then slowly rolled 30 feet backwards into me, despite blowing my horn, yelling out the window, waving my arms, and backing up my car as much as possible. She doesn’t get any cookies.

She was going less than a mile an hour so there was no damage to my car or hers. Which is a good thing, considering how mad I was. But she still doesn’t get any cookies. (My car won’t go, she said, when asked why she hit me. Well, I said, it seems to have no problems going backwards. Maybe you should put it in Park until I’m gone.)

Back to the article

OK – now that you have coffee and cookies, let’s settle in.

On the first page, they talk about:

One area of empirical research with considerable information concerning print documents is that of “signaling.” Text signals consist of preview statements, overview sentences, headings, and other cues about the content and structure of a text. Empirical studies tell us that when readers who are “reading-to-learn” encounter signals in print text, they are better able to comprehend new or difficult information. These signals help readers create mental roadmaps or schemata of a text’s structure and content that in turn help them absorb new information. As readers take in new information, they instantiate existing schemata, adapt them as needed to fit the new information, and at times, even reject the new information. [p 416]

These seem to be the cohesive devices mentioned in Ward’s article. And they seem to be saying that cohesive devices create the coherence in the users’ minds. I’m good so far. We’re consistent with Ward’s article.

What are headings good for anyway?

They go on to wonder how the frequency of headings in print or online might impact comprehension (comprehension being cohesion, as it is in the mind of the reader and is testable). Headings are important as signals because they help the readers understand the structure of the text and assign meanings (create schemas). “The importance of text signals is evident when one realizes how much readers use cues in a text to help them discern a text’s meaning.” [p. 417]

However, difficulty of the text or reader motivation seems to matter in using signals for coherence.

[…] if readers lack domain knowledge or find a text difficult to comprehend, signals that provide text coherence may help them better understand the material. In contrast, text that readers perceive to be easy does not benefit from signals—in such cases, readers have little need for the extra help that signals could provide. Similarly, if readers find the information enjoyable to read, they may be less likely to need headings or other signals to help them understand the content. [p. 417 with references removed]

However, headings may be problematic in online text:

Because readers tend to scan online text, many Web usability specialists have called for the use of more headings in online text. Although such advice may be valuable for readers who are “reading-to-do” and who are thereby searching online texts for specific information, it may be less appropriate for readers who are “reading-to-learn.” [p. 417 with references removed]

Reading-to-learn vs reading-to-do

Reading-to-learn are people who are reading for learning the material for some reason. For example, they may be interested in consumer goods and how to reduce their consumption, as my son and his family are interested in. So my son sends me lots of reading-to-learn information links to help me know more. One of the goals to reading-to-learn is recall.

Reading-to-do is reading information with the intent to act on that information. Product documentation is a good example of reading-to-do material. You expect that your reader will read some of the material and then turn to the product and do something. Reading-to-do is short term. Recall is not usually a goal.

Cognitive loads can be so heavy

They also give good definitions of cognitive load, which you should be familiar with if you’re not. It’s an important concept for us, because our users don’t come to our materials with nothing else going on.

[…] three types of cognitive load that could help explain comprehension differences in print versus online environments:

  • Intrinsic cognitive load, which is inherent in the material itself
  • Extraneous cognitive load, which is imposed on the user by the format and way that the material is presented
  • Germane cognitive load, which involves the learner’s attempts to process and understand the information [page 418]

They suggest that online docs users may have an increased overall cognitive load because of the nature of online docs. Navigation, hyperlinks, and other common online devices may add to the cognitive load while the reader is trying to understand the material. So getting to that germane cognitive load part of the brain could be a little harder for online readers than book readers.

The point to their research

They found that reading-to-learn readers reading online material did better overall with headings that appeared at a medium frequency (1 heading for about every 200 words).

[…] readers of print-based text are able to comprehend more of the text than readers of online text regardless of heading frequency. And this may be particularly true of our reader population, readers with little prior knowledge of the topic or personal motivation for reading the information—even though our readers were extremely experienced in Web environments. [p 422]

Another of the interesting findings (and they had several more I won’t get into here)

[…] readers of the print text were much more resilient to heading frequency extremes than the online readers. In comparison to the online readers, readers of the print texts had relatively similar comprehension scores regardless of heading frequency. In contrast, readers of the online texts had higher scores with the medium-frequency heading condition and considerably lower scores with the high-frequency and no-heading conditions. The significant interaction of display medium and heading frequency reveals that the comprehension of online readers is much more susceptible to weak structural cues (such as too many or too few headings) than is the comprehension of print-based readers. [p. 422]

So what does all this mean?

Oh, sure ask me that. I was hoping you could put this together yourself. Fine, I’ll do the heavy lifting. You just sit there and eat that cookie.

What this all means is that the signals you use to create cohesion (headings) must account for the delivery method for coherence to happen.

For example, people who are reading-to-learn online need heading signals about every 200 words. So, if your sentences are less than 25 words each and your paragraphs are less than 5 sentences, you need a heading about every 2 paragraphs if you are delivering the info online to your readers.

If you are delivering the information in a printed form, you have more wiggle room with your heading signals. You could go as far as a heading every 300 words and still get coherence.

Future research directions

In good Graduate School form, I end this article with a call for further research.

But in truth, we really need someone to do this work with reading-to-do material. We have guidelines for teaching people the concepts in our materials now. But what about the people who just need to program the damn remote control again? What signals are best for them so they get coherence? Do they need coherence? Do they even need cohesion?

If you know about any research that speaks to this, let me know. I’m very interested in that.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: