ELearning/Course building/Fidelity and learning

Fidelity is the degree to which an image, representation, simulation, description, or explanation resembles reality.

A continuum


1. Three views of the brain, higher to lower fidelity. Consider their different uses.

On a continuum, visuals with the highest fidelity are photographs in full and natural color, and photorealistic 3-D renderings. The high-fidelity visual contains detail, depth, shadow, texture, and nuance of color we see in our environment. High fidelity simulations and descriptions add realism by including details, variables, and complications to match reality as much as possible.

On the other end of the continuum are visuals with low fidelity, such as line drawings, silhouettes, and iconic images. The low-fidelity image uses fewer visual elements and qualities that resemble an object. Reducing realism reduces the fidelity of the image. Likewise, low fidelity simulations and descriptions reduce the number of details and other variables to a minimum in order to focus on the intended message. Low fidelity is effective when the goal is to draw attention to the big picture, larger story, or specific details, to induce a quick response, or strengthen the impact of a message or action (Malamed, 2011).

Designing from a minimalist approach has many advantages when it comes to comprehension. Minimalism makes every phase of the human information-processing system more efficient as we perceive a graphic, situation, or concept, hold the information in working memory, internally represent the information, and interpret it. A graphic or simulation composed of primitive features or an explanation void of details will take less time to probe and assimilate compared to one that is more complicated, such as a natural scene, photograph, or real-life scenario.

However, we know that simplicity can only take us so far along the learning curve. Detail, finesse, emotion, and scope must all increase to the point of being transferable to real life. Research in this area (Alessi, 1998; Noble, 2002; Clark & Lyons, 2004; Malamed, 2011) supports the common sense approach of limiting details for novice learners and increasing them as learners gain experience and expertise. In the language of cognitive load theory, we increase germane load as the learner masters lower levels of skill. It may even be appropriate to add extraneous load for expert learners.

Graphical techniques that reduce or increase the perception of fidelity (Malamed, 2011):

Low Fidelity		High Fidelity
Hard shadows	< >	Soft shadows
Smooth surfaces	< >	Rough surfaces
Sharp color	< >	Soft color transitions
Sharp countours	< >	Soft countours
No blurring of distant objects	< >	Appropriately blurring distant objects
Severely limited number of elements, details	< >	Realistic detail
Focuses on specific intent of the message	< >	Total realism
Pristine (featureless) background	< >	Realistic visual noise
Silhouttes, icons, line art	< >	Photorealism

Diminishing returns

As Alessi (1998) points out in Figure 1 below, however, we reach a point of diminishing returns even for expert learners. At some point, the additional investment made toward fidelity brings little return in increased learning. This is as true for simulating flying an airplane as it is for simulating a counseling session. The better approach is to maximize fidelity for aspects that really matter.


1. Fidelity and learning for different learners (Alessi, 1998)

Talking about instructional design generally, Jimenez (2013) says instructional designers … “always face the temptation of overloading our narratives with detailed information and data. In our eLearning development work, it might be best to follow the unwritten law in flirting: Don’t show too much, leave something for the imagination. Surprisingly, even people from science think that the best way to teach science to new learners is by not sounding scientific. If science people can absolve themselves from the burden of data overload, eLearning professionals should do the same. However, this is not to say that information and data are not important – they are. But the old adage that too much focus on the trees can make us not see the forest, rings true here.”

The degree of fidelity and the learning stage of the learner are mutually interdependent variables that must be considered when designing instruction intended for transfer of learning (Noble, 1998). Novice learners are easily overloaded by detail while advanced learners need it (Clark et al, 2006).