IHRIS Collect Design Overview: Difference between revisions

From IHRIS Wiki
No edit summary
 
(10 intermediate revisions by 2 users not shown)
Line 1: Line 1:
iHRIS Collect is intended as a one-time data collection survey tool built on the I2CE Framework.
iHRIS Collect is intended as a one-time data collection survey tool built on the I2CE Framework.
==Motivation==
==Motivation==
*See the following [http://www.capacityproject.org/hris/blog/index.php/2009/09/ihris-collect/ HRIS Blog Entry] for a general overview
*Moving from a paper based system directly to iHRIS Qualify or iHRIS Manage often involves difficulty as the data is not cleaned.  So this is really a three step process:
*Moving from a paper based system directly to iHRIS Qualify or iHRIS Manage often involves difficulty as the data is not cleaned.  So this is really a three step process:
**collect the data  
**collect the data  
Line 7: Line 8:
*the data collected may not already be organized into the structure expected by iHRIS Manage. For example, a few places have been  interested in using iHRIS Manage.  They did not have a Job/Position structure already defined.  As a result, they imported/entered a lot of data into iHRIS Manage without a well thought out Job structure.
*the data collected may not already be organized into the structure expected by iHRIS Manage. For example, a few places have been  interested in using iHRIS Manage.  They did not have a Job/Position structure already defined.  As a result, they imported/entered a lot of data into iHRIS Manage without a well thought out Job structure.
*Introduces administrators/developers to the I2CE framework so that they have familiar systems as the move from the data collection process to the routine data entry
*Introduces administrators/developers to the I2CE framework so that they have familiar systems as the move from the data collection process to the routine data entry
*If data entry clerks are maintained between the one-time data collection process and the routine data collection, they will have a similar system which reduces training
*If data entry clerks are maintained between the one-time data collection process and the routine data collection, they will have a similar system which reduces training.  This is the process for the iHRIS Manage implementation in Zanzibar,
*we have lot of similar functionality and design architecture already built into the system via the custom reporting tool.
*we have lot of similar functionality and design architecture already built into the system via the custom reporting tool.


Line 13: Line 14:
There are several possibilities for a staged implementation.  The stages do not necessarily need to be completed in order.  They more reflect a break down into functional units that will ultimately benefit iHRIS Manage, Qualify and Plan.
There are several possibilities for a staged implementation.  The stages do not necessarily need to be completed in order.  They more reflect a break down into functional units that will ultimately benefit iHRIS Manage, Qualify and Plan.
===Stage 1:Branding/Single Form===
===Stage 1:Branding/Single Form===
In this stage, we develop the branding and provided a basic implementation for iHRIS Collect:
In this stage, we develop the branding and provided a minimial implementation for iHRIS Collect:
*Settle on the name (iHRIS Collect)
*Settle on the name (iHRIS Collect)
*Choose color scheme and update CSS.  I vote for beige.
*Choose color scheme and update CSS.  I vote for beige.
*Build off of iHRIS Common -- may need to break up iHRIS Common into some sub-modules.
*Build off of iHRIS Common -- may need to break up iHRIS Common into some sub-modules but probably not.
*Assume each page to edit only contains one form.  Assume the developer can manually [[Defining Forms|define a form]] in [[Configuration (Magic) Data|magic data]]
*Assume each page to edit only contains one form.  Assume the developer can manually [[Defining Forms|define a form]] in [[Configuration (Magic) Data|magic data]]
*The html template files are build by hand by the developer for each form.
*The html template files are built "by hand" by the developer for each form and one which links to all the forms in the survey.
*A survey page: is a form and its associated template file.  Each such page should be packaged as a module and assigned a task to view/edit the data.
*A survey page: is a form and its associated template file.  Each such page should be packaged as a module and assigned a task to view/edit the data.
*To enable custom reporting,  for each form one needs to define the relationship which contains just the form as a primary form.
*To enable custom reporting,  for each form one needs to define the relationship which contains just the form as a primary form.
*perhaps we can write a script to automate this once the fields (and their types) for a survey are defined.
*perhaps we can write a script to automate this once the fields (and their types) for a survey are defined.
===Stage 2:Custom Pages===
===Stage 2:Custom Pages===
Implements [[Custom Pages]] to allow the system administrator to easily create pages based on existing forms in the system.
Implements [[Custom Pages]] to allow the system administrator to easily create pages based on existing forms in the system.
Line 28: Line 30:
===Stage 4:Custom Page Displays===
===Stage 4:Custom Page Displays===
As the custom pages will general follow a similar architecture to the custom reporting tools, we can extend the functionality of the custom pages by adding in custom displays.  Here are some possible displays:
As the custom pages will general follow a similar architecture to the custom reporting tools, we can extend the functionality of the custom pages by adding in custom displays.  Here are some possible displays:
*create a PDF version of the survey for paper-based collection
*create a PDF version of the survey for paper-based collection.  Is this the priority here?  Should not be too difficult to do.  A few potential pitfalls:
**what to do with long lists of data, should we display them all and have a check box? or should we leave it as free text and let the data entry people deal with it?
**what about tiered data lists such as geography (district, region, country)?
*exporting survey to something be imported to make PDA/hand-held survey tools.  also provide need to provide a means to sync the data from the hand-held
*exporting survey to something be imported to make PDA/hand-held survey tools.  also provide need to provide a means to sync the data from the hand-held
**[http://www.episurveyor.org/user/index epi surveyor]
**[http://www.episurveyor.org/user/index epi surveyor]
Line 37: Line 41:
==Other Documents==
==Other Documents==
*[[File:IHRIS_Collect.pdf]]
*[[File:IHRIS_Collect.pdf]]
*[http://www.capacityproject.org/hris/blog/index.php/2009/09/ihris-collect/ HRIS Blog Entry]


[[Category:Design Document]]
[[Category:Blueprints]][[Category:iHRIS Collect]]

Latest revision as of 12:50, 8 October 2013

iHRIS Collect is intended as a one-time data collection survey tool built on the I2CE Framework.

Motivation

  • See the following HRIS Blog Entry for a general overview
  • Moving from a paper based system directly to iHRIS Qualify or iHRIS Manage often involves difficulty as the data is not cleaned. So this is really a three step process:
    • collect the data
    • put the data into electronic format
    • clean up the data
  • the data collected may not already be organized into the structure expected by iHRIS Manage. For example, a few places have been interested in using iHRIS Manage. They did not have a Job/Position structure already defined. As a result, they imported/entered a lot of data into iHRIS Manage without a well thought out Job structure.
  • Introduces administrators/developers to the I2CE framework so that they have familiar systems as the move from the data collection process to the routine data entry
  • If data entry clerks are maintained between the one-time data collection process and the routine data collection, they will have a similar system which reduces training. This is the process for the iHRIS Manage implementation in Zanzibar,
  • we have lot of similar functionality and design architecture already built into the system via the custom reporting tool.

Staged Implementation

There are several possibilities for a staged implementation. The stages do not necessarily need to be completed in order. They more reflect a break down into functional units that will ultimately benefit iHRIS Manage, Qualify and Plan.

Stage 1:Branding/Single Form

In this stage, we develop the branding and provided a minimial implementation for iHRIS Collect:

  • Settle on the name (iHRIS Collect)
  • Choose color scheme and update CSS. I vote for beige.
  • Build off of iHRIS Common -- may need to break up iHRIS Common into some sub-modules but probably not.
  • Assume each page to edit only contains one form. Assume the developer can manually define a form in magic data
  • The html template files are built "by hand" by the developer for each form and one which links to all the forms in the survey.
  • A survey page: is a form and its associated template file. Each such page should be packaged as a module and assigned a task to view/edit the data.
  • To enable custom reporting, for each form one needs to define the relationship which contains just the form as a primary form.
  • perhaps we can write a script to automate this once the fields (and their types) for a survey are defined.

Stage 2:Custom Pages

Implements Custom Pages to allow the system administrator to easily create pages based on existing forms in the system.

Stage 3:Custom Forms

Implements Custom Forms to allow the system administrator to easily create custom forms. This can then be used to build custom pages.

Stage 4:Custom Page Displays

As the custom pages will general follow a similar architecture to the custom reporting tools, we can extend the functionality of the custom pages by adding in custom displays. Here are some possible displays:

  • create a PDF version of the survey for paper-based collection. Is this the priority here? Should not be too difficult to do. A few potential pitfalls:
    • what to do with long lists of data, should we display them all and have a check box? or should we leave it as free text and let the data entry people deal with it?
    • what about tiered data lists such as geography (district, region, country)?
  • exporting survey to something be imported to make PDA/hand-held survey tools. also provide need to provide a means to sync the data from the hand-held

Other Documents