Outline Scope for ICPI Analytic Work Spaces

[bowdenj]

Per Inbrief 2/20, schedule meeting with -ICPI Analytic Work Space participants -DIV BI Subgroup -Cluster Leads

to discuss -Analytic questions to be answered in the work spaces -Use of site level data in the work spaces (given Amb priorities for RPM and moving forward)

[KTmonkey]

I think we might want to consider prioritizing OVC and PP_PREV as that seems to be causing some challenges at RPM...

[shapaklyak]

hi @bowdenj, were there any notes after the meeting the week of RPMs?

[bshook]

Hey Katya,I have notes and an outline/SOO for next steps I’m just waiting on the code book before proceeding- it will affect the time line. I spoke w jasmine yesterday and am updating based on our discussion. Will be ready as soon as the codebook

[KTmonkey]

I emailed on Tues asking for a status update on the code book and haven't gotten a response. During Monday's inbrief Jess said that there was going to a meeting sometime this week and they should know more about the code book...

[shapaklyak]

@bshook @KTmonkey what is the codebook? Starting to review all the materials sent for workspaces after RPMs...

[bowdenj]

The codebook I believe is referring to a "dictionary" or some other documentation that identifies what data can be pulled directly from the PDH via the work spaces. Right now, the datasets are labelled in a way that is not exactly human-readable and is confusing for analysts used to using fact view datasets.

My meeting with the developers starts in a few minutes. We'll be reviewing exactly this, to talk through how these could be labeled differently or some kind of dictionary provided so users can identify the right data for their analysis in the work spaces. I will respond back here when we have a timeline for updates.

[KTmonkey]

what @bowdenj Jess said!

[bowdenj]

@KTmonkey - The meeting went well yesterday, but more follow-up is needed. The developers are interested in exploring the option of having some cubes or datasets available that mimic the structure of the fact views, so we were talking through that option a little bit yesterday (very much still in the exploratory phase there). We have a follow-up scheduled next Wednesday afternoon to sit and walk through a code-book or dictionary.

[bshook]

@bowdenj @shapaklyak @KTmonkey submitted some draft initial documents to leads.

1) Re: Jess's last comment- if the end goal is to hopefully help pano developers transition the tools more easily then wouldn't it be best to try and recreate the visuals using the same data structure as they would be? Just making sure the exercise is useful to the developers.

2) After discussing further with Jasmine she wanted the first step of the project to be staff recreating the analytic dashboards in the workspaces and the second phase to have staff create new visuals for analytic questions (per our initial meeting with Jess during RPMs).

I have updated a proposed SOW (attached). ICPI Analytic Workspace SOW_12MAR2018.xlsx It also assigns the participants a dashboard- let me know if you have any feedback on current assignments.

Since DIV also wanted a catalogue of visuals I think we can use this opportunity to start the log. Analytic Workspace Log_20180306.xlsx. Each participant would record the names of all the visuals in a tool, determine if they are in pano or not (and if they are functional), and then records issues with recreating them (with the idea that we wouldn't recreate visuals already in Pano). Hopefully this helps to streamline feedback and kills a few birds with one stone.

[KTmonkey]

Hi friends!

Question on the log- where @bshook said

Each participant would record the names of all the visuals in a tool, determine if they are in pano or not

is that just referring to the visuals in our assigned tool or is that asking something more broad?

[shapaklyak]

@bshook,, just talked w/ @jb3436. Can you please swap out the assignments? OVC dashboard for me and then @kimisatoPC for DREAMS dashboard? Thanks!

[bowdenj]

Met with developers yesterday, and it's not going to be quite as simple as creating a code-book, unfortunately. What they are doing is essentially re-creating fact view files as options for users to select within the work spaces (starting with sitexIM for a few countries). Will keep you guys updated as this moves forward. One thing they did mention that I thought was interesting is that if you go into "PEPFAR project," you have the option for datasets or visuals. The developers indicated that the visuals should essentially reproduce the dashboards available on Pano. Some may look a little different, but the data should be the same. I agree that it would help to reproduce tools we're familiar with as a way to train up in using the data, but it could be a good check to put together one dashboard and then, in a separate dossier, pull in a copy of the Pano one. Otherwise, you may be comparing yourselves to a standard that can't really be met, because some functionality is slightly different between work spaces and Pano dashboards.

[bowdenj]

@bshook - do you feel as though scope has been outlined for work spaces? If yes, I'd like to close this ticket and open new ones for tracking the work being done in the spaces. If you feel like there is more planning/outlining to be done with the group, we can keep this open and track it here. Let me know your preference.

[bshook]

@bowdenj I think we can close this

[shapaklyak]

@bshook @bowdenj can we put the analytic workspace SOW in a wiki page and the log for reference/use?

[bowdenj]

Katya and I looked into linking documents on wikis, and we thought it might make more sense to put this in SharePoint. Here's the link: https://www.pepfar.net/OGAC-HQ/icpi/Shared%20Documents/Forms/AllItems.aspx?RootFolder=%2FOGAC%2DHQ%2Ficpi%2FShared%20Documents%2FClusters%2FDIV%2FProjects%2FAnalytic%20Work%20Spaces&FolderCTID=0x012000C815322C717A7E4B8164EA374FA254EC002682B939F9BED347BD49E43D77D3C691&View=%7B94C838B2%2DE166%2D4122%2DB8B4%2D7BEB9E1BC12B%7D&InitialTabId=Ribbon%2ERead&VisibilityContext=WSSTabPersistence

Both documents are there. If this works for people, I'll close this ticket.