# 04.02.20 Call with Luke

## General
- Time zone mixup -> write time zones when scheduling
- Awesome reading reports and books (yes)
- Depends on project e.g. World Bank probably uses all of the internal presentation analytical stuff in report, was one of the more analytical processes, questions: 1. Where do you expect decision making to happen, 2. when something goes wrong, who do you talk to? Government should make decisions, but local community should answer to faults

- [ ] schedule regular calls
- Might come in march
- Week of 16th all around?
- [x] Send Luke and Emma a schedule with timing

## Development Process
- One series of visualization, what does that take to implement in lively? Prototype, iterate 
- One as PO, internal customer, how much velocity we get out 
- End of march: one visualization a day, 
- Luke: an hour a week being an analyst and trying out

- Send schedule 
- Milestones, who is working at what rate, who is there when
- [x] Plan chart

## Persona
- Mixture of several people at AVF
- People who speak languages, labelling, translating -> qualitative people
- Quantitative analysis -> other people
- One tool for both, break down walls between these

- kavice can’t write python, Lukas can (two people working at AVF)

- Tom : Increase qualitative experience, decrease development experience
- Might not exist but might be plausible

- AVF only 5 years old

- Qualitative analysis researchers, not necessarily going to be programmers
- Some coding, but primarily excel system, people around who can help with python and R

## Lukes opinion
- More interested in recommendation process comprehensible
- Compare similar diagrams
- Correlation in data
- Important correlations

- Not : good recommendations, full potential out of excel, Elon musk
- Computer making suggestions less important

- [x] Generating 7000 random data points for spikes

## Researcher call
- try, may not get interesting answers, or better answers that he says
- Another project at the moment, customer is interested in this and that question, looking into literature, figuring out questions to ask on radio show, collecting data, reading, taking context into analysis
- No algorithm (looking at data, reading)

- Introduce us to people doing research: 
- Study performing on them, set of questions that we want to ask them
- [ ] Write out a script

## How do qualitative and quantitative analysts work together?
- Thematic labelled things, set of descriptions
- Then send to quant: significant effects between attributes, correlations, effect variation, statistical 
- Qualitative: interesting relatable quotes
- Overall project lead combines these two, producing charts + example messages

- Going from makro-behaviour searching for individual who supports that view distorts view
- So we’re trying to do the other way around for qualitative analysts going from message to bigger overview

## Visualizations 
#### (9) 
- Taking two perspectives, putting together
- Where does this work and how?
- Can we do this easily and when? With what data

#### (23) 
- adjustments of dimensional structures
- Interesting view of external structuring

#### (27) 
- historical narrative for exploration route, annotate into powerpoint format
- Structure narrative differently than structure of analysis (hopefully not three days)

- What happens if we fork history?
- Question, following steps, record history, rewind and go back, fork
- Story board with branching and backtracking
- Should view dead ends as learning opportunities, sign post what they look like

- Rational design process, why it’s impossible and why you should fake it

- rather Input text field for feedback than scale from 1 to 10, because AVF doesn’t believe in simple numbers

## Storyboard
- Plot distribution of age
- Which fraction of the data are we showing or not showing
- Same thing with themes, distribution of people reporting which theme, indicated by number above theme

- Could be build, give to researcher, interactive user story

- If anything is ready enough to take to researchers to play with, he needs to have some time to prepare and find people beforehand

- Storyboard versions : merging visualizations together, use of exploration history

## Luke's Interactive chart Prototype
- Important to curate list of themes
- Dot structure for multiple themes is difficult to read
- One diagram to aggregate all points
- Introduces separations too quickly
- Stakeholders liked it, engaged with it, but got really fast into detailed designing errors

## Schedule
- Halfway through February with more storyboards, check in
- Working implementation in march

## Retro
- Agenda funktioniert gut, besser strukturiert, Timeboxes funktionieren aber interessieren ihn gar nicht so anscheinend 
- Zeitzonen mitschreiben bei Meeting ausmachen
- Moderator beibehalten
- Luke konnte manchmal nicht so frei reden, Modus vielleicht nicht der beste mit “hast du dir das angeguckt und jetzt sag mal was dazu”, genaueren Plan machen, wir uns auch welche rauspicken, zu denen wir Feedback haben wollen

## Review
- Sollen jetzt wieder mehr programmieren
- Trotzdem Visualisierungsbar noch voll machen
- Oft komplett andere Ideen spannender, aber auch “übliches” Storyboard sehr gut angekommen
- Lively oder Google Forms? An mehrere Google Forms, erstmal Lively


## ToDo

- Call-Deliverables:
    - Projektmanagement Gantt-Chart (mit Milestones)
    - Woche 16.März -> vorbeikommen kann?
    - Neuer Call 19 - 29 Februar
    - Storyboards/Spikes:
        - Merge diagrams
        - Use of exploration history
- Next steps
    - Visualisierungen
        - From individuals to context / Macrozusammenhänge
        - Message focus
    - Storyboards
    - Technological spikes
        - 7000 points in lively
        - Real data in lively anzeigen 
            - Data compression
            - Pipeline data into d3
    - Mail 
        - -> In der Woche am 16ten zwei Leute nicht da, ansonsten schon. 
    - Skript überlegen für Researcher Call
    - Learn the javascript
    - How to build framework

## Vision

**Was wir denken, was Luke möchte**

- Qualitative Researchers sollen beim Explorieren der Messages Diagramme machen können
    - Die können nicht programmieren, sollen das aber mit (wenig) Unterstützung von Leuten, die programmieren können, können
- Diagramme sollen interaktiv / explorierbar sein

**Unsere Vorstellung**
- Zuerst:
    - Diagramme programmieren, die beim Messages lesen erstellt werden können
    - Dann davon abstrahieren und das ohne Programmierkenntnisse (aber mit Unterstützung von Programmierer:innen)  möglich machen
- Dann wäre cool, wenn Diagramme einfach (iterativ) veränderbar wären
- Und es wäre cool, wenn interaktiv -> Diagramme für Policy Maker