4.4 Worked Example of Framework
4.4.4 Phase 4: Application
The application phase outlines how reports should be created using the infor- mation encoding framework. Contrary to the first three phases, it is recom- mended that the steps in this phase should be repeated every time information is reported. To illustrate this phase in the worked example, a hypothetical sce- nario will be created wherein fictional datasets will have to be reported to a cluster of the managers. All of the information regarding the datasets are fictional and can be thought of as information related to three asset intensive factories that produce some kind of product. The 20 managers consulted will represent the employees of this fictional company along with their attributes (departments and management levels) and their information presentation pref- erences. Each step in the application phase of the framework will be followed closely to illustrate how the responses collected from the managers will aid in selecting formats in which to encode information in a fictional report.
4.4.4.1 Step 8: Design Report
According to the framework, the first step in the application phase is designing the report. This means that designers should decide on what they want to
convey in the report and which datasets can be used to support their message. A summary of the datasets which will be used in the report is thus the output of this step. Table4.11shows information about five fictional datasets to which the steps in the application phase of the framework will be applied.
Table 4.11: Summary of datasets used in worked example.
No. Dataset Name Period Source Datasets Data Type 1 Total output per factory 2014 Multiple Single Nominal 2 Factory A five-year output 2010-2014 Factory A Single Interval 3 Monthly output Factory A 2014 Factory A Single Interval 4 Sources of downtime per factory 2014 Factory A Multiple Nominal 5 Critical HS&E incidents per factory 2010-2014 Multiple Multiple Interval
4.4.4.2 Step 9: Identify Data Tasks in Report
Once the datasets that need to be included have been identified and sum- marised, the main data task that needs to be facilitated by each needs to be identified. This will normally be decided by the designers of the report as they can match the main reason for including the dataset in the report to one of the data tasks discussed in Section 3.2.2. Furthermore, the task code of that data task in the framework also has to be determined and Table 3.3 (also in Section 3.2.2) shows the task codes as a function of the task name, number of datasets and data types. Table4.12shows how each dataset has been assigned a task code based on its number of datasets, data type, and task name.
Table 4.12: Datasets with task names and task codes assigned to each.
No. Dataset Name Datasets Data Type Task Name Task Code 1 Total output per factory Single Nominal Compare Discrete Value TC3 2 Factory A five-year output Single Interval Identify Trend TC15 3 Monthly output Factory A Single Interval Characterise Distribution TC1 4 Sources of downtime per factory Multiple Nominal Find Extreme Values TC12 5 Critical HS&E incidents per factory Multiple Interval Identify Trend TC16
Now that the task codes have been identified, the most preferred informa- tion presentation format for each can be retrieved.
4.4.4.3 Step 10: Identify Target Audience
This step is critical to the effectiveness of the recommendations made by the framework. Designers should pay special attention when identifying the tar- geted recipients of the reporting process as including or excluding the wrong
employees in the clusters could greatly decrease the effectiveness of the frame- work. The target audience of this fictional report has been chosen so as to include the most prominent department and the management level. Table 4.1
shows that twelve of the 20 managers indicated that they worked in the en- gineering departments of their organisations. Of these twelve managers, ten were middle managers with one of the remaining two being a high level man- ager and the other a low level manager. The target audience for this fictional report was thus the ten middle managers in the engineering department. 4.4.4.4 Step 11: Match Audience Preferences to Tasks
This step combines all of the information collected in the previous steps to recommend an information presentation format for each dataset identified in Step 8 (Section 4.4.4.1). These recommendations will attempt to satisfy as many of the information presentation preferences of the target audience as possible. In order to do that, the preferences of the target audience need to be isolated from those of the rest of the managers. The preferences of the ten managers who did not form part of the target audience were filtered out in the Microsoft ExcelTM spreadsheet in which they were stored. Table 4.13 shows the combined encoded P, S and A responses of the ten middle managers in the engineering department - the target audience of the fictional report.
Table 4.13: Combined P, S and A preferences of target audience after encoding.
Task Code Table Poin
t Graph (Dot Plot) Radar Graph Bar Graph (Simple) Bar Graph (Group ed) Bar Graph (Stac ked) Column Graph (Simple) Column Graph (Group ed) Column Graph (Stac ked) Bo x and Whisk er Histogram Histogram (Multiple) Line Graph Line (Area) Graph TC1 -5 -1 -4 0 15 -8 3 4 0 TC3 -2 -3 -4 5 17 TC12 11 1 -6 0 11 TC15 -8 1 -7 0 7 5 11 -4 TC16 -8 0 -6 -4 -2 -1 -2 8 13 3
The cell with the highest value for each task code has been shaded green to indicate the presentation format which is recommended for that particular task code. Notice that row of task code TC12 has two cells which contain the value 11 which is also the highest value in that row. This means that there are two formats which are considered to be equally appropriate for task code TC12 by the framework, namely tables and grouped column graphs. Section 3.5.4
mentions three steps which should be followed sequentially to determine which one of the tied formats should be used. These steps are:
1. The presentation format with the smallest number of -2 ’s (avoid) re- sponses should be chosen.
2. If there is still a tie, the format with the most 1 ’s (secondary preferences) should be used.
3. If there is still a tie, it means that the number of P, S and A responses for the tied formats are the same. The designer can then use his or her own discretion in deciding on one of the tied formats.
Table4.14shows the number of primary, secondary and avoid responses for task code TC12 (not encoded) as indicated by the ten managers in the target audience.
Table 4.14: Number of P, S and A preferences of target audience for TC12.
Task Code Table Poin
t Graph (Dot Plot) Radar Graph Bar Graph (Simple) Bar Graph (Group ed) Bar Graph (Stac ked) Column Graph (Simple) Column Graph (Group ed) Column Graph (Stac ked) Bo x and Whisk er Histogram Histogram (Multiple) Line Graph Line (Area) Graph TOT AL Primary TC12 4 2 0 0 4 10 Secondary TC12 3 1 0 2 3 9 Avoid TC12 0 2 3 1 0 6
Neither one of the first two steps recommended to break the tie are ap- plicable as none of the managers requested that tables or grouped column graphs be avoided when performing task code TC12 and both formats were the secondary choices of three managers. Since the formats are tied and both have the same number of S and A responses, it means that both formats have the same number of P responses as well. Consequently, as indicated by the third step recommended when there is a tie, the designers of the report can use their own discretion when deciding whether to use a table or a grouped column graph when encoding the information for task code TC12.
The column on the far right shows the total number of primary, secondary and avoid preferences that have been indicated by the ten managers in the target audience. Notice that of the ten managers, only nine indicated a sec- ondary preference for task code TC12. If the manager who did not indicate a secondary preference had indicated either a table or a grouped column graph as his or her secondary preference, that format would have been selected. This
illustrates the importance of validating the responses to ensure that a single primary and a single secondary response is collected for each task code for each manager.
Table 4.15 shows the recommended format for each dataset identified in Step 8 of the framework. These recommended information presentation for- mats are based on the preferences indicated by the target audience in Step 4 of the framework. All of the datasets, except the Sources of downtime per factory dataset, have a single recommended presentation format. The designers of the report can decide whether they would like to use a grouped column graph or a table to depict the information in the Sources of downtime per factory dataset.
Table 4.15: Condensed example of the framework output showing the ideal pre- sentation format for each task.
No. Dataset Name Task Code Presentation Format 1 Total output per factory TC3 Column Graph (Simple) 2 Factory A five-year output TC15 Line Graph
3 Monthly output Factory A TC1 Column Graph (Simple)
4 Sources of downtime per factory TC12 Column Graph (Grouped) or Table 5 Critical HS&E incidents per factory TC16 Line Graph
4.4.4.5 Step 12: Create Report
The last step in the framework is not only creating the report, but also creating the appropriate presentations for each dataset. Contextualising the framework (Phase 1) ensures that all of the recommended presentation formats are pro- ducible by the organisation. Appendix Dshows information of each dataset to be reported as well as an example of the format recommended by the frame- work.