Information Technology Reference
In-Depth Information
Table 2. Analysis core categories and catalyst questions for open coding
Zachman column
Core question
Detailed questions
Data
What?
• Is the data accessed for reporting purposes only? (Are related services type 1 or 2?)
• Is the data aggregation of other data? Is the data actually a category containing other catego-
ries or codes? (Related services are type 3)
• Is the data available only in non-electronic form? (Type 4)
• Is the data received directly or indirectly from any external source? (Related services are type
7)
Function
How?
• Is any data being accessed, processed, validated, generated or searched? How and which
data? (Details for types 1, 2, 3, 7)
• Is there a logic, which is atomic, independent and resulting in a concrete result? (Type 5)
• Is stateful interaction needed between human beings and the IT system? (Type 6)
• Is there any stateful communication with external parties? (Type 8)
• Is data format transformation needed in communication? (Type 8)
Network
Where?
(No detailed questions as none of the eight identified service types had any specific characteristics
related to where the service is used.)
People
Who?
• Who is using the system and with what kind of channel or device? (Reusability of type 6 services)
Time
When?
(No detailed questions as none of the eight identified service types had any specific characteristics
related to when the service is used.)
Motivation
Why?
• Why is the data processed in non-electronic form? (Is there need for type 1 or 2 services?)
into sub-categories. If different wording has been
used for the same phenomenon during the open
coding phase, then these codes should be merged,
as well. Additionally any relationships between
codes under the same categories should be identi-
fied. The questions in this phase are, for example:
in service elicitation, will probably benefit from
a set of pre-defined categories. The pre-defined
categories and their characteristics are needed to
identify the service types known to be potentially
reusable. Additional categories and sub-categories
can be created inductively, if needed.
The actual service candidates are identified
after the use case analysis and are generated from
the codes under the “Function” core category with
their relations to the codes and categories in the
“Data” and “People” core categories.
Service candidate types 1, 2, 3 and 7 (Table 1)
are identified from the codes under the “Access
information” categories. The granularity of the
candidate is determined with the relation to the
codes and categories in the “Data” core category.
The abstraction level is the one identified from the
business process analysis, if available. Codes in
the “Access non-electronic data” category can be
transformed into content services if the business
process itself is revised, as well.
Service candidate types 5 and 8 are identified
from the codes under equivalent categories. If
these codes have relations to other core categories,
then these relations can be used to determine the
•
Are there any codes which mean the same
thing and could be merged?
•
Are there similar codes which could be
grouped under a common category?
•
Can any categories be seen as an aggregate
of two or more categories?
In the last phase, selective coding is performed
on the data identified in the previous phases. In
this phase, the identified categories and codes are
sorted into core categories based on the Zachman
Framework and the sub-categories based on Table
1. The hierarchy of the categories is presented in
Figure 1.
The use of partly pre-defined categories as a
priori categories is against the inductive nature of
the grounded theory research methodology. How-
ever, the recurring analysis of similar data, such as
Search WWH ::
Custom Search