Most of the work in the US today is mental labor as opposed to manual labor. As individuals get more skilled in their work and become experts, their value to your organization grows. Once an employee reaches relative expertise, their absence (or loss) becomes a significant detriment. But capturing their knowledge allows businesses to reduce the impact of absence substantially. However, effective knowledge capture is difficult because experts tend to have a high omission rate when attempting to detail and record their decision-making. Cognitive task analysis (CTA) is an empirically proven method of producing the most complete knowledge capture possible.

CTA is a method of interviewing experts to get fine-grained knowledge from them. It is particularly well-suited for safety-critical, sensitive tasks and large-scale training initiatives. Before proceeding further, we need to learn more about expertise, automaticity, and the human cognitive structure.


Knowledge levels in a subject run from novice to expert. While novice is easy to understand, expert is a more ambiguous term. Compared to novices, experts have not only more knowledge but also a better quality of knowledge. They perceive large patterns within their domain, primarily because their knowledge is so well organized. This pattern recognition allows large amounts of information to be perceived and processed quickly that it almost appears to be intuition.


Cognitive psychology operates on the premise that humans have a severely limited amount of conscious memory (working memory), almost unlimited long-term memory, and an organizational system that allows the memories to function together and facilitate learning. Working memory is referred to as one’s focus, attention, or “consciousness.” This memory is where the mental work is done. Because it is so limited, we must use it efficiently. Chunking is one such way.

This image has an empty alt attribute; its file name is image-10.png

A chunk (or element) is anything that requires a slot in working memory. Probably different for each person, chunks vary in the amount of information they hold. This amount is a function of the individual’s acquired knowledge in schemas, with one schema taking up a single chunk in working memory. More complex schemas contain and integrate more information into a single slot of working memory.

Consider, for example, a chessboard set at the starting position, except that the white King and Queen have been place-switched. For someone who has never seen a chessboard, the information present could easily overload available working memory because each piece’s location occupies one chunk. However, many or all the piece locations could occupy a single chunk for someone well-acquainted with the game. This increase in the information held in each chunk is possible because of the way information is associated with long-term memory over time.

This image has an empty alt attribute; its file name is image-11.png

Development of Expertise

After extensive practice, a schema can become automated so that a task is carried out quickly, effortlessly, and with few errors without occupying any working memory. This leaves space available to address more difficult or unfamiliar tasks.

This image has an empty alt attribute; its file name is image-12.png

Expert performance of cognitive tasks is not due to innate ability but is acquired through deliberate practice. Deliberate practice is tailored to a task, provides immediate feedback, and is repeated. It requires conscious, focused exertion to improve performance and is so fatiguing that it cannot last long.

In the workplace, getting the deliberate practice necessary for the development of expertise is not straightforward. The time spent on other work duties takes away from deliberate practice. However, training derived from CTA is a proven method for accelerating expertise development.

Cognitive Task Analysis (CTA) is a generic name for various techniques used to elicit the information, knowledge, and method of performing a given task from experts. As an extension of time and motion studies of manual labor, CTA addresses an observable action by determining the procedural (how-to) and declarative (facts) knowledge, decision points, and strategies necessary for its appropriate completion.

CTA Process

Several authors describe a general procedure for CTA that encompasses roughly the same steps. These steps assume that a preliminary needs analysis has been performed. It indicates a need for the type of in-depth knowledge garnered from CTA, complex Tasks, have high consequences,  or whose performance impacts the safety or large amounts of capital or operating expenditures are good candidates. The process proceeds as follows:

  1. Preliminary Phase: Experts are selected, and the analyst takes time to learn the vernacular of the experts’ knowledge domain.
  1. Identify Knowledge Representations: Establish the framework of the type of knowledge needed, e.g., procedural, declarative, etc. Map it into tasks, sub-tasks, and supporting knowledge.
  1. Elicit Knowledge: Knowledge elicitation is typically in the form of structured or semi-structured interviews, live or recorded observation, self-report through, autonomic response, or some combination of these techniques.
  1. Analysis and Verification of Data: Transcriptions of the interviews are formatted and verified by the experts. The use of multiple experts produces better results.
  1. Formatting and Use: This information may become the primary source document during the instructional design of a class or training session. It may become the basis for a checklist or questionnaire for an evaluation or assessment.


When describing the procedures necessary for a task, experts omit up to 70% of the decisions and vital information required to complete the task. However, through the iterative interviews, a compilation of procedures, and expert feedback, the procedural knowledge gathered through CTA far exceeds individual free recall.

From an instructional viewpoint, the procedures’ completeness is irrelevant unless those procedures can be utilized to provide students with higher-quality instruction. If properly applied, the instruction developed using CTA-derived information will lead to improved performance from the cognitive perspective. When learners are given complete information, their cognitive resources are available for developing schemas and learning instead of expended in the instructional gaps and struggling to grasp the content.

In multiple studies comparing CTA-based instruction and traditional instruction, students receiving CTA-based instruction have demonstrated superior performance to those receiving traditional instruction in medicine, radar system troubleshooting, landmine detection, and computer spreadsheet usage. They have even demonstrated improvements in undergraduate biology coursework. Additionally, a meta-analysis found that, overall, instruction based on CTA offers an improvement over traditional instructional methods with a large effect size.  The same meta-analysis also revealed that studies in military and university settings demonstrated CTA-based instruction to be highly effective.

In Practice

The cost of performing a CTA is primarily a factor of time (of practitioner and expert) and the expert’s opportunity cost. If the task is crucial or safety-critical, the cost will be negligible compared to the impacts of ineffective training. Furthermore, if the training is delivered to many learners, the cost per person trained diminishes rapidly. Radiant Digital is ready to help you with your advanced training by performing CTA-based instructional design —enabling you to build and maintain operational excellence.