“I keep six honest serving men (They taught me all I knew) Their names are What and Why and When and How and Where and Who” –  Rudyard Kipling

Control Area Analysis and Design

There is nothing new under the sun as they say, and CAAD is mostly a cut-down version of many existing methods and the critical aspect of the approach is very tight scoping of the project (and therefore clear definitions for the user requirements). Clear and concise report writing is also fundamental to CAAD.

CAAD also has a couple of additions that are not normally seen in analysis and design methods. A leaf has been taken out of the approach to double-entry accounting technique and helps to provide an audit trail for the system. There is also detailed analysis applied to data ‘ownership’ when designing the system.

The main components of CAAD are:

  1. Detailed Scoping
  2. Data Ownership
  3. Business Rules
  4. Data Integrity
  5. Processes
  6. Constraints
  7. Perspective
  8. Double Entry

The above is a list of subject areas.  A brief summary of each follows. Clear user requirements and a glossary of terms (data dictionary) are required in addition to this list.

Detailed Scoping

It is assumed that the reader has knowledge regarding the general issues of scoping an area of business, a project or even an application. Scoping is extremely important to CAAD since it provides a point of measurement for the veracity of the exercise. How data is structured depends entirely on whether the data originated inside or outside the scope of the exercise as described in the following section.

Data Ownership

This subject is key to data design and understanding how CAAD is a measurable process.

Data is compartmentalized as belonging to one of three distinct groups or ‘boxes’; White, Grey and Black. White data is that which is newly created by a process that falls within the scope of the exercise. Black data is ‘black boxed’ and the data has been verified by a source outside the scope of the project. An example of black-box data is postal addresses that have been provided by whatever organization is responsible for the veracity of the data, such as the Post Office or some government body. You cannot change the data because you did not create it and you do not ‘own’ it. Grey data is essentially black-box data that may need a bit of scrubbing to get it cleaned up to be in a fit state to use. If your business records postal addresses from other parties such as a home owner, then this is ‘grey’ data – it’s still not your data but you may need to give it a bit of a scrub for it to be usable.

The main point here is how you model and subsequently manage the different types of data. If you look at a management information system (MIS) you will see a lot of data repetition, and this is because you cannot change the data since it was created elsewhere – this is the black-box data that the user has no jurisdiction over. Grey data is trickier, for instance you may well obtain the corresponding black-box data in order to validate and correct the grey data (as in the case of postal addresses), so your system needs to track this and it is a decision that needs to be supported by the business rules in this case.

Both black-box and grey-box data contents belong to ‘domain sets’. Their values are not subject to any particular constraints since they are a given by virtue of the fact that they were supplied to the Control Area and were not created within the Control Area.

From applying the rules of data ownership the data structure may look quite different than if the data was all treated in the same way through a common approach to normalization. Data repetition is not an issue providing that it is controlled and does not threaten the data integrity – so normalization should not be a consideration here.

Data Integrity

Integrity of data is managed without obsessive application of normalization rules and concerns about data duplication – both of which are red herrings as can easily be demonstrated.

This subject requires a data dictionary and a clear understanding of the meaning of the data, and its meaning depends entirely on the context defined by the business rules.

Black-box data requires either minimum or no validation since the guarantee of its veracity is simply the source of the data, such as that provided to a management information system from the operational area of the business responsible for the creation of the data. Should these values change then there needs to be a controlled replication of the changes. It should be noted that historical data must not be changed since it may well have been propagated well outside of the existing system and should be kept consistent even if it is now inaccurate since the original transaction has been completed. A couple of simple examples are where someone has changed their name either through marriage or deed poll – you need to know that you are dealing with the same person but you cannot retroactively modify transactions that took place under their original name. To accommodate linking the two different names for reporting purposes is a system/database design issue.

Where a transaction is completed then it is locked out. Any changes necessary will be made in the fashion of an accounting system where a subsequent entry effectively ‘reverses out’ the original entry. If there are instances where this approach is deemed unsuitable then specific rules will apply to that particular situation. Business rules regarding changes to data that has already been committed are defined in the data dictionary. The section titled Double Entry helps explain how data integrity is applied in certain situations.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: