Key Features Cognitive Model

The second type of cognitive model is called key features. The key features cognitive model is used when the attributes or features of a task are systematically combined to produce meaningful outcomes across the item feature set. The use of constraints ensures that the relationships among the features yield meaningful items. The key features model is most suitable for measuring the examinees' ability to assemble and apply key features within a domain, as well as to solve problems using these key features. The defining characteristic of this modelling approach is that the content for the item can vary, and the key concept varies across the generated items due to the meaningful combination of features. In contrast to the logical structures model which focuses on representing algorithmic problem solving, key features models are organized based on the definition and relationship between features that are used to generate items. The systematic combination of permissible features is defined by the constraints specified in the feature panel (i.e., bottom panel) of the cognitive model.

Table 2.3 contains a parent item for diagnosing either the cold or the flu. As with the logical structures model, we present a relatively simple example that will be used throughout the book to illustrate principles and applications using key features. This example was selected because it is straightforward and does not require extensive medical knowledge while still demonstrating the logic of the key features model. Table 2.4 provides a list of cold and flu symptoms. The variables on this list serve as the key features that can be used to differentiate cold and flu symptoms. Figure 2.4 contains a cognitive model based on the parent item in Table 2.3 using the symptoms in Table 2.4. The top panel identifies the problem and its associated scenarios. The SME began by identifying the general medical problem (i.e., respiratory illness) specific to the parent test item. Two types of scenarios associated with this problem are the common cold and seasonal flu. The middle panel specifies the sources of information related to the problem. In this simple example, two sources

Table 2.3 Parent Item Related to Diagnosis of Common Cold and Seasonal Flu

A 22-year-old female sees her doctor and reports that she's been experiencing a mild cough and slight body aches that have developed over a few days.

Upon examination, she presents with an oral temperature of 37°C. What is the most likely diagnosis?

(A) Hay fever

(B) Ear infection

(C) Common cold*

(D) Acute sinusitis

(E) Seasonal influenza

* correct option

Table 2.4 Symptoms of Common Cold and Seasonal Flu

Common Cold

Seasonal Flu

Fever Is Rare


Mild Cough, Chest Discomfort

Severe Cough, Chest Discomfort

Mild Body Aches and Pains

Severe Body Aches and Pains



Mild headache

Severe Headache

Sore Throat

Sore Throat

Stuffy, Runny Nose

Stuffy, Runny Nose

Adapted from Public Health Agency of Canada "Cold or Flu: Know the Difference".

are identified: history and examination. These sources of information are used to describe the content that will be manipulated in the generated items. The bottom panel highlights the salient features, which include the elements and constraints, for each source of information. For the example in Figure 2.4, five features (i.e., age, cough type, body aches, onset, temperature) were identified for the history and examination sources of information. Recall, each feature specifies two nested components: the elements and constraints. For instance, the cough type feature contains three values: mild, hacking, severe. The model is constrained so that mild and hacking coughs are associated with a cold, while a severe cough is associated with the flu. In other words, the SME specified that the features mild and hacking coughs can only be paired with a cold. The feature of severe cough, by comparison, can only be paired with the flu in our example. These requirements are presented in the constraint section of the feature panel. Mild and hacking are coded as CC, while severe is coded as SF. The key features model is characterized by its reliance on specific constraints in the features panel to produce generated items that contain logical combinations of values. Hence, Figure 2.4 serves as a cognitive model for AIG because it specifies and coordinates the knowledge, skills, and content required to generate items that, in turn, can be used to evaluate examinees' reasoning and problem-solving skills to diagnose respiratory symptoms associated with a cold and the flu. Examples of key features cognitive models for AIG can be found in the content areas of abdominal injury (Gierl, Lai, & Zhang, 2018b), infection and pregnancy (Gierl & Lai, 2016b), chest trauma (Gierl &

Lai, 2018), hernia (Cierl & Lai, 2013), post-operative fever (Gierl, Lai, & Turner, 2012), jaundice (Gierl & Lai, 2017), and dentistry (Lai, Gierl, Byrne, Spielman, & Waldschmidt, 201 6).

< Prev   CONTENTS   Source   Next >