Bug Detection: Adventures in Intuition and Tacit Knowledge Iain McCowatt ...

14
Bug Detection: Adventures in Intuition and Tacit Knowledge Iain McCowatt http://exploringuncertainty.com [email protected] @imccowatt imccowatt © Can Stock Photo Inc. / fr

Transcript of Bug Detection: Adventures in Intuition and Tacit Knowledge Iain McCowatt ...

Bug Detection:Adventures in Intuition and Tacit Knowledge

Iain McCowatthttp://exploringuncertainty.com

[email protected]@imccowattimccowatt

© Can Stock Photo Inc. / frenta

An Unhappy Engagement

“They don’t notice obvious bugs”

“The tests aren’t relevant”

“We’re not allowed to ask questions”

“They don’t have theDomain knowledge”

“The KT didn’t make sense”

“We get thrown specs andtold to get on with it”

How We Detect Bugs: Two Modes

The analytic:• Deduction• Explicit

Models• Hypothesis

Testing (Checking)

The intuitive:• Induction• Experience• Pattern

Recognition

How We Detect Bugs: Two Modes

The analytic:• Deduction• Explicit

Models• Hypothesis

Testing (Checking)

The intuitive:• Induction• Experience• Pattern

Recognition

Tacit & Explicit Knowledge

Explicit Knowledge: that which has been explicated

Weak Tacit Knowledge: not explicated, and for no deep reason

Strong Tacit Knowledge: not been explicated, rooted in society

If you haven’t done so yet, make Tacit & Explicit Knowledge by Harry Collins the next thing you read

Dependency on the Tacit

Solve: 10+10(no RST grads please!)

20?100?32?

1010?False?True?

Two Modes Revisited

The analytic:• Seemingly

explicit• Yet tacit-laden

The intuitive:• Predominantly

tacit

Turning to Expertise

Contributory Expertise: Possesses TK and can “do”

Interactional Expertise: Possesses TK, can “speak the language” but not “do”

Interactive Ability: The ability to interact, socialize in a domain, and gain TK

Implications: Looking for Interactive Ability

• Interactive Ability, the ability to develop interactional expertise, may be THE primary attribute of a good tester

• Possible indicators:• Active listening• Relating to target context• Bootstrapping from neighboring

domains• Mirroring language

Implications: Location, Location, Location

• The greater the cultural distance, the larger the required message size

• At some point even very large messages will fail

• Remote teams have fewer opportunities for socialization

• Remote teams are more reliant on formal communications

• A double whammy: bigger messages, narrower bandwidth

Implications: Independence

IndependenceInterdependence

More likely to share biases?

Less likely to have requisite tacit

knowledge

BUT: Possessing TK in one domain does not precludepossessing TK in another

Implications: Diversity

• If these are the domains of what we know, what we would recognize as a problem, then which offers the greatest bug recognition potential?

• This applies to both the individual and the team.

Last Word

“No matter how it looks at first, it's always a people problem”-Gerald M. Weinberg