Cognitive Modeling and Human Computer Interaction

What is Cognitive Modeling?

"A cognitive model is an approximation to animal cognitive processes (predominantly human) for the purposes of comprehension and prediction. Cognitive models can be developed within or without a cognitive architecture, though the two are not always easily distinguishable." - Wikipedia


Use of Cognitive Modeling


As the definition above said that the cognitive modeling is used for the purpose of prediction. In relation to Human-Computer Interface development, we can use cognitive modeling to aid in design and evaluation of interface evaluations.


There are many models in the field of cognitive modeling which can be used to accomplish this. These are being discussed below:


I am discussing it in the order of the most practical to most theoritical


GOMS


I mentioned the GOMS method in my last blog also. Here, I will mention this from a cognitive science perspective.


GOMS analyzes a task in terms of


G - Goals


simple Operations used by a particular person performing a task


Methods for accomplishing a goal


Selection rule is required to chose.


GOMS model does not entail creativity or tasks which involve problem solving, etc. 


GOMS is well suited for analysis of routine, skilled performance as opposed to problem solving.


Post Completion Error


I will give a brief introduction to post completion errors. These will be applied later.


An adequate theory of error is one that enables us to predict where and under which conditions  the error will error. This phenomenon of errors after completion is called post completion error.


An example of Post Completion Error:


An example of post completion error would be forgetting the card inserted in the ATM machine after withdrawing money.
A similar example would be leaving the original document in the photocopy machine after making a xerox. 


Since these errors occur after completion of the tasks, they are rightly labeled as post-completion errors.


It is to be noted that these "mistakes" that I described in the above scenarios, namely leaving card in the ATM, forgetting the original document after taking xerox, happen only occasionally.


So, we conclude that post completion errors occur but they do not occur often.


Information Access


Remember when you last time spend days searching the phone number of an old friend in your old records? 


This exemplifies that information is useful only if we can find it when we need it.


We will apply this simple concept in human computer interface development.


3. Cognition - Artifact - Task Triad


Cognition refers to our thought process. Artifact refers to the tool we use and task is what we want to accomplish.


Traditionally, methodologies generally use cognitive, artifacts and task pairwise rather than altogether.


For example, the field of computer science is focused on developing artifacts. These artifacts are for doing certain tasks. However, cognitive concerns are ignored.


If we ignore any one from the cognition - artifact - task triad , the results will be errorprone.




However, Cognitive Modeling is bound to consider cognition, artifact and task as inter-related components. 


Cognitive Modeling and Cognitive Task Analysis


Cognitive Task Analysis specifies the cognitive steps required to perform a task using an artifact.


Cognitive Modeling goes further than cognitive task analysis. Each step in cognitive modeling is grounded in cognitive theory.


Dimensions of Cognitive Models


There are two:


1. Generative vs. Descriptive
2. Generality vs. Realism


Generative vs. Descriptive


Generative models are implemented as executable computer program that takes the same input and generate  the same output the people do.


Generative models run on computer, whereas descriptive models must be hand-simulated, increasing the chance of error.


Generality vs. Realism


Generality is the extent to which a model offers theoretical implications that extend beyond model's domain.


Realism, on the other hand, is the extent to which the modeled behavior corresponds to actual interactive behavior of a particular operator performing a particular task.


Summary


The cognitive models mentioned above are useful in predicting the user's action and are used to improve the interface design.


References:


http://mentalmodels.mitre.org/cog_eng/reference_documents/gray-cognitive%20modeling%20and%20hci.pdf