Critique of User/Task Analysis
 
Automated Waitron Environment (AWE)
CS 6751, Summer_94, Class Project
 
Terence D. Hughey
25 Jan 99

Project Purpose

This project, with a team of 5 people, is described as an interface for the restaurant dining experience.  It is more accurately described as an ordering interface with entertainment possibilities while you are waiting on your food.  The basic problem is defined as:

The project determines that the dining experience consists of several user tasks which include choosing a restaurant, making reservations, traveling to the establishment, arriving, being seated, ordering, waiting eating, paying and leaving.  The restaurant’s tasks include taking reservations, meeting/greeting, seating, taking orders, serving, receiving payment and cleaning up.  The effort seeks to automate some of these processes.
The design team decided that a product would be built to enhance the following:
User Analysis

The user analysis sought to define the “general world dining population with at least a 5th grade reading level and no physical or mental disabilities.  They used questionnaires to obtain user data of both the diner and the waitron. The knowledge of the users on the project team appeared to frame most of the opinions about users.
 

Task Analysis

“Task Decomposition” was performed initially using the Smart Guy/Gal approach since they had waitron and customer experience on the project team. They did not use a hierarchical structure but simply listed the tasks in time order.  Most of the task analysis appears to be of the group’s observation of their own tasks as users and their waitron at the Pizza Hut restaurant they visited.  I could not determine if the tasks were representative of other than their group and observations at the one resturant.

There was no formal use of  “Knowledge Based Analysis” or “Entity-Relationship Based Techniques” analysis.
 

What They Did Well

They clearly defined and understood who a certain type of user was.  They did a good job listing the tasks that were performed by this user (customer and waitron) and the way they would go about the dining experience.

The project group used the principal of observation in their Pizza Hut restaurant visit to gather data well.  They captured mostly time data and clearly had a good dining experience as evidenced by their comments and tip.
 
I personally thought the simplicity of their web pages, without a lot of graphics and other asides, made it easier to read and pick out the “meat” of the project ideas.  It sure made it faster to print.
 

What They Did Poorly

They missed an important HCI part of the issue by not addressing, in a more general and formal way, the memory and sensory issues of how we relate to food and the dining experience.  What cues, stimuli, and mental aids do we use when ordering our meal and interacting with a waitron?  What makes the environment pleasant?

The project team did not go outside its group or social ranking to perform its user and task analysis.  Perhaps, someone who rarely eats out, would have different expectations and tasks.  They assume that everyone is like them and comfortable with replacing the waitron’s time of interaction with the diner. Possibly, many come to the restaurant for interaction with a waitron and the socialization that occurs. The Hooter’s crowd might revolt, if you replaced any of their time with their waitron with a computer.

The assumption, that many have entertainment needs, other than stimulating conversation and interaction with other diners, may be erroneous. The Smart Guy/Gal approach used in gathering data has disadvantages, which they fail to acknowledge.  In short, more observation of different diners in a different restaurant would have given additional insights.  Encountering a bad dining experience, which the project seeks to prevent with its AWE, would have been helpful.

A “Hierarchical Task Analysis”, rather than just listing the tasks, would have benefited the interactive organization of the device.  Some use of a “Knowledge Based Analysis” or “Entity-Relationship Based” analysis might have been helpful.

The user/task analysis insights should have been mapped to formal HCI terms.  For example, what HCI concepts are involved in how a person thinks about food on a menu?  What is the mental model of the customer when the waitron explains the “rich creamy sauce”, with a to-die-for expression on their face? How do you duplicate the recommendation by the waitron of their favorite dish, describing it in detail?

I found no analysis of a number of competitive systems in existence at the time of this project.

There appears to be a lack of closure in evaluating the original goals of the system with anything measurable.

Back to Terence D. Hughey's cs6751b Classwork Page
Back to Gromit Home Page