<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta />
    <article-meta>
      <title-group>
        <article-title>AnnL: The Tool for Planning for Viable Enterprises</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <string-name>Edward Lewis</string-name>
          <email>e.lewis@adfa.edu.au</email>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <aff id="aff0">
          <label>0</label>
          <institution>University of New South Wales, Australian Defence Force Academy</institution>
          ,
          <addr-line>Canberra</addr-line>
          ,
          <country country="AU">Australia</country>
        </aff>
      </contrib-group>
      <fpage>121</fpage>
      <lpage>130</lpage>
      <abstract>
        <p>This paper presents the Analysis of Networked Links (AnnL) tool that supports the principles and practices for the planning for viable enterprises, through such disciplines as Enterprise Engineering (including Enterprise Architecture, Governance, and Service Management). It shows how the software can be used throughout the lifecycle of Enterprise Engineering, providing synchronized reports not readily available in other tools.</p>
      </abstract>
      <kwd-group>
        <kwd />
        <kwd>Enterprise Engineering</kwd>
        <kwd>Systems Planning</kwd>
        <kwd>Analysis of networked Links</kwd>
        <kwd>Work systems</kwd>
      </kwd-group>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>Introduction</title>
      <p>This paper describes the use of a tool that can be used to plan an enterprise so that it is
viable. This tool – Analysis of networked Links (AnnL) – can be used from strategic
through to tactical planning; from enterprise-wide portfolio management to detailed
service design. The design of the tool is based upon the experience of using it for
over 15 years in preparing more than 50 strategic plans, Business Cases, tender
evaluations, architecture risk analyses, and policy developments.</p>
      <p>
        Every enterprise consists of a set of processes that convert inputs into outputs as
products or services. These processes require people using resources to do their job
within socio-technical systems, which we will call ‘work systems’ [
        <xref ref-type="bibr" rid="ref1">1</xref>
        ]. If the
enterprise is to be viable then these work systems must be planned (decisions made and
resources allocated) so that the enterprise as a whole can act in time to avoid the
effect of risks or to take up opportunities.
      </p>
      <p>
        We will consider how the AnnL tool supports general managers (as they are
designers too [
        <xref ref-type="bibr" rid="ref2">2</xref>
        ]) or specialists such as strategic planners, Business Process Managers/
Decision managers, Enterprise Architects, or Service Managers - gathered under the
label of Enterprise Engineering in either the European [
        <xref ref-type="bibr" rid="ref2">2</xref>
        ] or US [
        <xref ref-type="bibr" rid="ref3">3</xref>
        ] meaning- in their
planning of work systems that make up viable enterprises.
      </p>
      <p>
        There are five risks in the existing techniques and tools for Enterprise Engineering:
• Incomplete consideration of all resources – Until recently the most important
resources, Who and Whom, were overlooked in the commonly used Enterprise
Architecture frameworks. Recently, we have seen the emergence of Human Views,
thankfully [
        <xref ref-type="bibr" rid="ref5">5</xref>
        ]. There is almost no mention of Worth in any framework, including
Zachman (as recently pointed out by [
        <xref ref-type="bibr" rid="ref6">6</xref>
        ]). Only one tool, Abacus [
        <xref ref-type="bibr" rid="ref7">7</xref>
        ], helps in
determining the cost of ownership of architectures – despite the mention of Economic
views in GERAM [
        <xref ref-type="bibr" rid="ref8">8</xref>
        ] (or ISO 15704, if you prefer). Even Archimate, which does
provide a notation schema that applies to most of the layers (resources) in a work
system, misses skills, worth, facilities, and time. We need techniques and tools
that cover all of the resources.
• Insufficient consideration of alignment – Currently, Enterprise Architecting looks
the wrong way. It concentrates upon integrating resources across processes - along
the rows in Table 1. We do have Business Motivation Models or Enterprise
Visions to give the ‘big picture’. We do have diagrams that show the relationship
between capabilities and functions (business services) and activities and, ultimately,
data or physical resources. Unfortunately, most viewpoints cover only two layers
in an architecture. It is hard to align the resources over all of the levels. There are
some exceptions: service maps used in Service Management do show the
top-tobottom alignment of resources to objectives, with [
        <xref ref-type="bibr" rid="ref9">9</xref>
        ] giving such an example using
Archimate notation. We need to be able to align all resources in this way.
• Lack of unified approach – There is confusion in the use of different notation,
terms, and approaches. There are attempts to develop standards in notation
UML, SysML or Archimate [
        <xref ref-type="bibr" rid="ref4">4</xref>
        ] - but we still need notations that address all of the
layers of planning; from business motivation to implementation plans, such as
physical data models or network blueprints.
• Documentation rather than design – Although some tools, such as Systems
Architect or Abacus, provide simulation, risk management, or costing support for the
design of architectures, most tools are merely documentation aides (as noted also by
[
        <xref ref-type="bibr" rid="ref10">10</xref>
        ]). They do little to help in planning, providing only views of the intended
design for buyers and blueprints for builders.
• Addressing the wrong audience – Most Enterprise Engineering tools are intended
for use by specialist enterprise modelers rather than by general managers. They
produce detailed visual representations of possible systems to be considered by
verbally skilled senior managers as part of their decisions about the acquisition of
resources. We need to support the decision-making process of these managers
rather than just feeding them incomprehensible models.
      </p>
      <p>AnnL is designed to remedy all of these risks.
2</p>
    </sec>
    <sec id="sec-2">
      <title>Use of AnnL Tool</title>
      <p>All that is necessary to use AnnL is to list items (resources) then link them. AnnL
uses the categories of the items, the initial estimates of their parameters, and the
nature of the links between the items to prepare graphical, numerical, or verbal reports
to decision makers so they can judge best what should be done.</p>
      <p>These items and the nature of their links come from the data model in Figure 1.
The data model is derived from the need for AnnL to support the planning practices
and the Principles of Planning that are given on the Systems Planning Mentor website
at www.layrib.com, from where the software and its manual also can be downloaded.</p>
      <p>Through this data model, AnnL enables a planner to consider the consequence of
pressures that are of concern to the various key points-of-view; who are then willing
to pay to have resources with requisite value (capability, capacity, constraints) to use
to avoid the negative consequences and enhance the positive; to measure the risks of
the options that have been generated from combinations of alternative resources
thought to have the requisite capability; to determine the price risk of the options
through a cost model of the resources needed to carry out the tasks that implement the
options; and to describe the Action Plan for carrying out the tasks using these
resources, according to blueprints (‘viewpoints’) that guide those people who are
implementing that option with the least price and performance risk.</p>
      <p>
        It is this list-link operation that is the key difference between AnnL and other
Enterprise Architecture or data modeling tools. AnnL involves building a database (a
linked list, of course) from which the diagrams or tables are generated rather than
starting with a diagram and then building an ‘encyclopedia’. This approach was also
advocated by [
        <xref ref-type="bibr" rid="ref11">11</xref>
        ], independently from, and well after, the inception of AnnL. Their
approach does not cover the range of resources or reports as AnnL.
      </p>
      <p>
        Abacus from Avolution [
        <xref ref-type="bibr" rid="ref7">7</xref>
        ] also uses this approach, although mostly when building
an architectural description from existing lists of resources rather than during the
initial design process. As an aside, AnnL could be developed into a library within
Abacus, using it as the engine for producing the Reports rather than the current use of
Excel macros, but there are sufficient differences in approach, outlined below, to
warrant AnnL standing alone.
      </p>
      <p>List Items. So, the first action in AnnL is to list the items to be linked. Figure 2
shows an extract from the Input sheet. The planners describe each resource or action
or stakeholder in the left most column. They classify each item, according to the type
of resource or action, in the middle columns. Then they give initial estimates (using
low-likely-high values, if necessary) in the right most columns.</p>
      <p>Item Description</p>
      <p>Item Type
Grow business
Increase cash flow for
BUP
Win new business</p>
      <p>Values: capable
Values: capable
Values: capable</p>
      <p>Way</p>
      <p>Business service</p>
      <p>Resource
Way
Worth</p>
      <p>Sub-resource
Business service
Income - Earnings</p>
      <p>Low
1000
500</p>
      <p>Likely
1500
700</p>
      <p>High
2500</p>
      <p>In order to ensure semantic consistency, the planner classifies each item by using a
pick list that draws upon the checklist of resources in the data model. The pick list in
the successive columns alters according to what has been picked for the higher
classification. For example, if ‘Way’ has been chosen as the relevant resource for a value
then the sub-resources in the next column are only those part of the ‘Way’ checklist.</p>
      <p>Of course, this step is easier said than done. There are many planning practices,
described on the Systems Planning Mentor website, that need to be used by planners
to make sure this list is correct.</p>
      <p>Link Items. The other main action in building AnnL is to link the items. These
links show the nature and extent of the relationship between the different items.</p>
      <p>The links are made through pick lists, as shown in Figure 2. The resource at the
head of the link is picked; then the resource forming the tail is picked from a list
tailored to the head. AnnL then automatically moves to the rows for the tail items.
Finally, the link is entered in the rows corresponding to the tails of each link, using
more pick lists to ensure that the nature of the link is semantically consistent for the
items being linked.</p>
      <p>As for listing the items, the planners need to draw upon their expertise and the
planning practices to determine what these links should be.</p>
      <p>One of the differences between AnnL and other Enterprise Engineering tools is
that AnnL can use, encourages the use of, verbal input. The extent of the links can be
numeric (0.1 – 1) or verbal (very weak, vw, – very strong, vs), for example.
Similarly, as shown later, AnnL can report numerically or verbally; whatever the
recipients of the Reports prefer.
Fig. 3. List of Links, showing the pick lists that are used to link head items and row items (to
the left), with the extent of the link inserted into the three columns to the right
That is all that has to be done by the planners. AnnL does the rest: carrying out the
variety of analyses and producing the models that the planners need, in whatever
format that they wish.
3</p>
    </sec>
    <sec id="sec-3">
      <title>Applications for Enterprise Engineering</title>
      <p>AnnL produces many different Reports, according to the application of planning. The
Reports contain various versions of the trees, tables (‘matrices’ or ‘catalogs’ in
TOGAF terms), or words that are the result of the calculations carried out by AnnL,
using the type of items and the links between them.</p>
      <p>One of the major advantages of AnnL is that it enables all of the many documents
that make up a Business Case or the design of an Enterprise Architecture to be
synchronized easily. Each time a planner changes an item or its link, AnnL will
automatically update all of the Reports containing that item and link. That is, the new
reports can reflect not only changes in editing of items, such as names, but also
changes in the logic underlying the relationships between the components. This
ability is rare in most Enterprise Architecture tools.</p>
      <p>Examples of these Reports are given below. These examples show the variety of
formats and results that can be produced by AnnL. The underlying algorithms used in
the analyses are available from the Mentor site.</p>
      <p>The following examples are taken from a case study (loosely) based upon an actual
project. Anonymity is protected - as is the author. The case study involves a
highlevel strategic decision by Business Utility Providers (BUP) to take an initiative
forming a new business line in hiring out specialist but redundant staff. There are a
number of decisions to be made at the strategic, operational, and tactical level. After these
decisions (whether to centralize or decentralize the business line or to outsource the
IT, for example) have been made then AnnL can prepare the blueprints for the project
managers or the systems builders.</p>
      <p>
        There is not enough space to show the full portfolio of reports. The following
viewpoints [
        <xref ref-type="bibr" rid="ref12">12</xref>
        ] were chosen because they show the results of techniques that only
AnnL provides or bring out the versatility of AnnL or they are more compact than the
diagrams that might be preferable in the actual case and so can fit in the page limit.
      </p>
      <p>Consequence Chain. One of the techniques used in strategic planning is the
modeling of the consequences of external or internal pressures upon the enterprise. The
intention of this viewpoint is to assist planners in finding points of intervention that
break the chain leading to risks or amplify the chain leading to opportunities. These
interventions are values (descriptions of the required resource, such as a procedure for
winning more business or a facility in a less risky location). The set of values that
intervene in the chains describe the capability of the strategic initiative.</p>
      <p>
        Figure 4 shows a consequence chain in the N2 or Design Structure Matrix [
        <xref ref-type="bibr" rid="ref13">13</xref>
        ]
viewpoint. Of course, this example is but a very small subset of the full analysis. This
viewpoint can be easier to use than the equivalent diagrammatic map for the more
complicated circumstances in the usual architectural description [
        <xref ref-type="bibr" rid="ref14">14</xref>
        ]. This example
shows how the value of ‘Win new business’ breaks the link to the risk of ‘lose more
staff’; nothing gets past this intervention. The numbers at the head of each column
are a result of the modeling of the consequences from the cost of the risk back to the
initial drivers of the risk (considering other drivers not shown in this example). So
‘win new business’ is an important capability.
      </p>
      <p>Influence of Points of View. Figure 5 shows the influence between the
points-ofview (stakeholders). This analysis starts with estimates of the initial power of the
stakeholders according to their position, entered on the list of items. AnnL combines
this estimate with the extent of the influences, shown in the links, to determine who
inherits the most power.</p>
      <p>The viewpoint uses strings to represent the influence from one point-of-view to the
next. The ditto (“) marks represent a repetition of the item on a previous line.</p>
      <p>
        The diagram shows that the PS Association representative influences the
Politicians who influence the Government, and so on. There is a branch at Government,
who influences both the Department of Labour and the BUP Board. The Suppliers
start their own string, which intersects with the other at the BUP People CEO.
PS Prof
Association
[
        <xref ref-type="bibr" rid="ref10">10</xref>
        ]
"
"
Suppliers of
services[
        <xref ref-type="bibr" rid="ref5">5</xref>
        ]
"
      </p>
      <p>
        Politicians[
        <xref ref-type="bibr" rid="ref8">8</xref>
        ]
      </p>
      <p>
        Government[
        <xref ref-type="bibr" rid="ref6">6</xref>
        ]
"
"
"
"
      </p>
      <p>
        Dept of
Labour[
        <xref ref-type="bibr" rid="ref2">2</xref>
        ]
"
      </p>
      <p>The numbers in brackets are the results of the calculations of cascading influence,
which can be used to determine which point-of-views are key (Politicians and
Suppliers in this case). AnnL does produce other tables with this information in more detail.
It is dangerous politically to show this report to the actual points-of-view.</p>
      <p>
        This analysis is essential for the proper design of a system [
        <xref ref-type="bibr" rid="ref15 ref2 ref5">2, 5, 15</xref>
        ] but it is not
available in most Enterprise Engineering tools. Nor is the Strings viewpoint.
      </p>
      <p>Values Statement. AnnL can generate values trees (or means-ends tree or
Functional Decomposition Diagrams), as can most tools. More usefully, AnnL also
generates verbal descriptions of values, for the senior decision-makers who are more
comfortable with words than with diagrams, as shown in Table 1, AnnL uses the links
between values and sub-values, and the extent to which the key stakeholders are
willing to pay for each value, to determine the words (‘must’, ‘should’, ..) expressing the
criticality of each value. It generates this table and the equivalent visual tree.
In order to contribute to meeting the goal of winning new BUP business, the
BUP CEO
a. Must increase awareness of staff capabilities
b. Should improve exposure to market
c. Might increase expertise of BUP staff
whilst meeting the constraints:
a. Must comply with corporate policy about …</p>
      <p>etc</p>
      <p>
        Design Analysis Display. Although Enterprise Engineering should be about
design, most Enterprise Engineering tools are useful for documenting a design rather
than for designing. They rarely help in creating ideas for options. AnnL extends the
powerful creativity technique of (General) Morphological Analysis [
        <xref ref-type="bibr" rid="ref16">16</xref>
        ] through the
use of Design Analysis Displays (see www.layrib.com).
      </p>
      <p>Figure 6 shows an example of a Design Analysis Display. The diamonds are
design decisions formed by each of the capability values that intervene in the
Consequence Chain. The boxes are alternative solutions for these design decisions. A path
through all of the alternatives is an option. In the example, there are 13 possible
options.</p>
      <p>AnnL can use the judgment of the planners (shown in the links representing the
assessment of the alternatives against the pertinent values) to show the ‘best’ paths.</p>
      <p>Business case report. The design of work systems involves finding the option
with the least performance and price risk. As shown in Table 2, AnnL produces a
“Risk Picture” for decision-makers to use to trade off the performance risks of options
against their total cost of ownership, adjusted for the time of expenditure. It shows the
output in words or numbers, to fit to the cognitive style of the audience.
h
t
o
n
oD ign
Improve exposure nice to have might not might not</p>
      <p>The planner can use different financial metrics. They include cost-benefit ratios,
Return on Investment, Internal Rate of Return, Nett Present Value, and (preferably)
Risk-Adjusted Price – where the costs of failing to meet the values are added to the
price. Abacus seems to be the only other tool to carry out such calculations.</p>
      <p>Action Plan. AnnL can produce detailed Action Plans, as shown in Table 4. They
list the tasks for implementing processes, who is responsible for the tasks, the assets
needed and the budget for them, and the measurement of the quality of these tasks.
This Report is in a format more familiar to managers.</p>
      <p>AnnL also produces associated Reports. They include the list of roles, grouped
over tasks; total budget for each task; and consolidated list of qualities (performance
measures). These other Reports are generated at the same time as the Action Plan.
The AnnL tool supports all of the planning needed to ensure an enterprise is viable. It
uses a comprehensive checklist (WHAT) to provide a complete and consistent
consideration of the resources at the various steps in the planning process.</p>
      <p>All the planners need to do is to list and link items. AnnL then produces the
Reports that support the design of architecture, in the formats that are most acceptable to
the various audiences. These Reports are readily updated and synchronized.</p>
      <p>If Enterprise Engineers use AnnL when planning work systems then they are free
to use their insight and experience, systematically and with full analytical support,
without being caught up in the drudgery of documentation.</p>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          1.
          <string-name>
            <surname>Alter</surname>
            ,
            <given-names>S.:</given-names>
          </string-name>
          <article-title>The Work System Method: Connecting People, Processes, and IT for Business Results</article-title>
          . Work Systems Press, Larkspur, CA (
          <year>2006</year>
          )
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          2.
          <string-name>
            <surname>Boland</surname>
            <given-names>R.</given-names>
          </string-name>
          and
          <string-name>
            <surname>Collopy</surname>
            ,
            <given-names>F</given-names>
          </string-name>
          . (eds): Managing as Designing, Stanford University Press, Stanford CA (
          <year>2004</year>
          )
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          3.
          <string-name>
            <surname>Hoogervorst</surname>
            ,
            <given-names>P</given-names>
          </string-name>
          : Enterprise Governance and Enterprise Engineering. Springer (
          <year>2009</year>
          )
        </mixed-citation>
      </ref>
      <ref id="ref4">
        <mixed-citation>
          4.
          <string-name>
            <surname>Rebovich</surname>
          </string-name>
          , G.:
          <article-title>Engineering the Enterprise</article-title>
          .
          <source>1st Annual IEEE Systems Conference, Honolulu, HA</source>
          ,
          <fpage>9</fpage>
          -
          <issue>13</issue>
          <year>April</year>
          , p1-
          <fpage>6</fpage>
          (
          <year>2007</year>
          )
        </mixed-citation>
      </ref>
      <ref id="ref5">
        <mixed-citation>
          5. The Open Group: ArchiMate 1.0 Specification. http://www.archimate.org (
          <year>2009</year>
          ), viewed
          <issue>25 Jun 2010</issue>
        </mixed-citation>
      </ref>
      <ref id="ref6">
        <mixed-citation>
          6.
          <string-name>
            <given-names>Systems</given-names>
            <surname>Engineering</surname>
          </string-name>
          and
          <article-title>Assessment: The Human View Handbook for MODAF</article-title>
          . Bristol, UK,
          <string-name>
            <surname>Crown</surname>
          </string-name>
          (
          <year>2008</year>
          )
        </mixed-citation>
      </ref>
      <ref id="ref7">
        <mixed-citation>
          7.
          <string-name>
            <surname>Simsion</surname>
          </string-name>
          , G.:
          <article-title>What is wrong with the Zachman Framework?. The Data Administration Newsletter, www</article-title>
          .tdan.com/view-articles/5279/, (
          <year>2005</year>
          ), viewed
          <issue>17 Mar 2011</issue>
        </mixed-citation>
      </ref>
      <ref id="ref8">
        <mixed-citation>
          8. Avolution:
          <article-title>Abacus Products page</article-title>
          . http://www.avolution.com.au/products.html, (
          <year>2011</year>
          ), viewed 20 May 2011
        </mixed-citation>
      </ref>
      <ref id="ref9">
        <mixed-citation>
          9.
          <string-name>
            <surname>Bernus</surname>
            ,
            <given-names>P.</given-names>
          </string-name>
          :
          <article-title>Generalised Enterprise Reference Architecture and Methodology</article-title>
          . http://www.cit.griffith.edu.au/~bernus/taskforce/geram/versions/geram1-6-3/v1.6.3.html, (
          <year>1999</year>
          ), viewed
          <issue>26 Jun 2010</issue>
        </mixed-citation>
      </ref>
      <ref id="ref10">
        <mixed-citation>
          10.
          <string-name>
            <surname>Holschke</surname>
            ,
            <given-names>O.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Narman</surname>
            ,
            <given-names>P.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Flores</surname>
            ,
            <given-names>W.</given-names>
          </string-name>
          , Eriksson, Evaline, and
          <article-title>Schonherr: Using Enterprise Architecture Models and Bayesian Belief Networks for Failure Impact Analysis</article-title>
          .
          <source>J. Enterprise Architecture</source>
          ,
          <volume>6</volume>
          (
          <issue>2</issue>
          ),
          <fpage>7</fpage>
          -
          <lpage>18</lpage>
          (
          <year>2010</year>
          )
        </mixed-citation>
      </ref>
      <ref id="ref11">
        <mixed-citation>
          11.
          <string-name>
            <surname>Martinez</surname>
            ,
            <given-names>C.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Cane</surname>
            , Sheila, Salwa,
            <given-names>A.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Smith</surname>
            ,
            <given-names>K.</given-names>
          </string-name>
          , and
          <string-name>
            <surname>Lee</surname>
          </string-name>
          , Kristin:
          <article-title>Application of Network Visualization to Identify Gaps in Complex Information System Architectures</article-title>
          .
          <source>Tracking Number 08-0109</source>
          , MITRE Corporation,
          <source>VA</source>
          (
          <year>2008</year>
          )
        </mixed-citation>
      </ref>
      <ref id="ref12">
        <mixed-citation>
          12. ISO/IEC 42010 FCD:
          <article-title>Systems and</article-title>
          software engineering - Architecture
          <string-name>
            <surname>Descriptions</surname>
          </string-name>
          , International Organization for Standardization. Geneva, Switzerland (due
          <year>2011</year>
          )
        </mixed-citation>
      </ref>
      <ref id="ref13">
        <mixed-citation>
          13. Design Structure Matrix home page: http://dsmweb.org (
          <year>2011</year>
          ), viewed 20 May 2011
        </mixed-citation>
      </ref>
      <ref id="ref14">
        <mixed-citation>
          14.
          <string-name>
            <surname>Ghoniem</surname>
            ,
            <given-names>M.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Fekete</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          , and
          <string-name>
            <surname>Castagliola</surname>
            ,
            <given-names>P.</given-names>
          </string-name>
          :
          <article-title>A Comparison of the Readability of Graphs using Node-Link and Matrix_Based Representations</article-title>
          .
          <source>IEEE Symposium on Information Visualization, Oct</source>
          <volume>10</volume>
          -12, Austin, TX (
          <year>2004</year>
          )
        </mixed-citation>
      </ref>
      <ref id="ref15">
        <mixed-citation>
          15.
          <string-name>
            <surname>Clegg</surname>
            ,
            <given-names>C.</given-names>
          </string-name>
          :
          <article-title>Sociotechnical principles for systems design</article-title>
          .
          <source>App Ergon</source>
          ,
          <volume>31</volume>
          ,
          <fpage>463</fpage>
          -
          <lpage>477</lpage>
          (
          <year>2001</year>
          )
        </mixed-citation>
      </ref>
      <ref id="ref16">
        <mixed-citation>
          16.
          <string-name>
            <surname>Ritchey</surname>
          </string-name>
          <article-title>Consulting: Swedish Morphological Society home page</article-title>
          . http://swemorph.com (
          <year>2011</year>
          ), viewed 20 May 2011
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>