Occam's Razor

Created: Monday, 24 March 2008 Written by Gary Elsbernd

"When two or more explanations exist, the simplest one is probably true." - Occam's Razor - William Occam

Sometimes said as "When you hear hoofbeats, don't start looking for zebras," the actual quote is "Entities must not be multiplied beyond what is necessary," but this seems unnecessarily (and ironically) complicated. The sentiment, however, brings us back to simplification. Great design is not the result of nothing more to add, but nothing more to take away.

If we can equally implement a function with different controls, the one that is more common, generally accepted or more easily manipulated is the best choice. We cannot focus on the "gee-whiz" technology so much that instead of helping us simplify our jobs it makes them more difficult or complicated.


Fitt's Law

Created: Monday, 24 March 2008 Written by Gary Elsbernd

"The time to acquire a target is a function of the proximity and size of the target." - Fitt's Law

This is a generally known principle in physical activity. The larger the target is, and the closer it is to where you start, the easier it is to find. That's not rocket science.

This law can also be applied to interface design.

  • Making the desired task flow clear and prominent will help the user identify the next steps and encourage the appropriate behavior.
  • Making exceptions and tangential information smaller and isolating them will reduce the likelihood of error.


Gilbert's Behavioral Engineering Model

Created: Monday, 24 March 2008 Written by Gary Elsbernd

There are many possible reasons for poor performance. In the past, documentation or training was the only solution to these problems, as phrases like, "It's a training problem," or "We'll put it in the manual" were catch-all solutions to poor processes. Thomas Gilbert's Model analyzes performance deficits from six standpoints. The interventions to overcome performance barriers have the highest leverage (cheapest to implement for highest return) from box 1 to 6. In other words, if the problem can be solved through better communication of expectations, it is more effective, easier and cheaper to the organization than a training program to teach performers a task they don't understand.





Environment (organizational factors)

1. Data, information

Do performers know what is expected?

Interventions: Communication, clear statements of purpose and expectations

2. Resources, tools, environmental support

Do performers have what they need to perform?

Interventions: Open supervisor support, appropriate tools, applications

3. Consequences, rewards, incentives

Do performers get appropriate feedback?

Interventions: Consistent and immediate feedback of results, consequences must be linked to performance

Performer characteristics (personal factors)

4. Knowledge, skills

Do performers have the knowledge or skills to perform?

Interventions: Training, Job Aids

5. Capacity

Are performers capable of performing?

Interventions: Selection process

6. Motivation

Do the performers care about the job or their performance?

Interventions: Selection process

In analyzing the root causes for a performance issue, we often will identify issues and solutions that have nothing to do with documentation or training. Because of this, we are no longer limited to those solutions, but can design performance centered systems leveraging all of the tools at our disposal.

Grudin's Law

Created: Monday, 24 March 2008 Written by Gary Elsbernd

"If the person responsible for using the system does not benefit, the system is doomed to failure." - Grudin's Law, Jonathan Grudin

We need to recognize that pushing our pain, that is, our needs, on someone else to implement will at best give us grudging and resentful compliance. If we can ensure that the person receiving the benefit is the one doing the work, we will get a more motivated group of performers.For example, the marketing group for a retailer decides they want zip codes from all of their customers. It involves more work for the cashier, and is resisted by customers. If there is no clear incentive for the cashier, or consequences of failure, the cashiers will enter false data in an attempt to make their lives easier.What we need to do in cases like this is to show the clear benefits of the task to the user and get them "on our side". The other option is to monitor and correct issues, but the carrot is preferable to the stick.


More Articles ...