Tool Design




PeterT leading, aiming for July 2013


Tool Design Issues

Four key areas:
· Back-end: calculation schemes etc.
· Front end: interface design / outputs
· Implementation platform
· Process

Back-end

· Need to ensure consistency in scorings, i.e. 1 should always be low and 3 (or whatever) should always be high.
· Need to review ranges. These should be impaired (i.e. always a mid-point). Some variables need to be zero-based, others not. Recommend 1 – 5 or 0 – 4 is appropriate for most circumstances.
· We need to review and refine the calculation scheme and possibly conduct a comparative analysis of different calculation schemes to identify the most robust / predictable approach.
· We need to include sensitivity analyses when testing calculation schemes.
· The new filters etc. need to be integrated into the updated version.

Front-end

· Improve the user management (“idiot-proofing”).
· Embed outputs of the tool in narratives describing technologies (pros, cons, application etc.) or vice versa.
· Explore the use of graphical outputs for explaining outputs (e.g. relative contributions of land, labour etc. to the outcome of the analysis).
· Rank and sort lists of outputs.
· Automate reporting.
· Different outputs for different end-users (e.g. pictorial outputs for farmers’ consumption).

Implementation Platform

· Stick with Excel for now but review as complexity increases. May be a case for moving to a bespoke platform (PC, Android?).
· Deliver tool and manual over the web but no online implementation.

Process

· User evaluation. Review experiences of past users of the tool and feed this into proposed re-design.
· Conduct a user-requirement analysis with key informants.
· NT, VP, AD and PT to agree responsibilities for implementing the new version. Possible face-to-face programming-shop, late April. PT will follow up.
· Consider structural linkages to FEAST (VCA?).


Tool Design Issues

Four key areas:
· Back-end: calculation schemes etc.
· Front end: interface design / outputs
· Implementation platform
· Process

Back-end

· Need to ensure consistency in scorings, i.e. 1 should always be low and 3 (or whatever) should always be high.
· Need to review ranges. These should be impaired (i.e. always a mid-point). Some variables need to be zero-based, others not. Recommend 1 – 5 or 0 – 4 is appropriate for most circumstances.
· We need to review and refine the calculation scheme and possibly conduct a comparative analysis of different calculation schemes to identify the most robust / predictable approach.
· We need to include sensitivity analyses when testing calculation schemes.
· The new filters etc. need to be integrated into the updated version.

Front-end

· Improve the user management (“idiot-proofing”).
· Embed outputs of the tool in narratives describing technologies (pros, cons, application etc.) or vice versa.
· Explore the use of graphical outputs for explaining outputs (e.g. relative contributions of land, labour etc. to the outcome of the analysis).
· Rank and sort lists of outputs.
· Automate reporting.
· Different outputs for different end-users (e.g. pictorial outputs for farmers’ consumption).

Implementation Platform

· Stick with Excel for now but review as complexity increases. May be a case for moving to a bespoke platform (PC, Android?).
· Deliver tool and manual over the web but no online implementation.

Process

· User evaluation. Review experiences of past users of the tool and feed this into proposed re-design.
· Conduct a user-requirement analysis with key informants.
· NT, VP, AD and PT to agree responsibilities for implementing the new version. Possible face-to-face programming-shop, late April. PT will follow up.
· Consider structural linkages to FEAST (VCA?).