Skip to content
This repository was archived by the owner on Apr 7, 2022. It is now read-only.

Feature: Longitudinal tracking#4

Merged
avisionh merged 9 commits intomainfrom
feature/long-tracking
Jan 18, 2021
Merged

Feature: Longitudinal tracking#4
avisionh merged 9 commits intomainfrom
feature/long-tracking

Conversation

@avisionh
Copy link
Contributor

@avisionh avisionh commented Jan 3, 2021

Summary

This branch visualises the movement how students have responded to the wellbeing survey over time.

The main script to review in this branch is:

  • notebooks/longitudinal-tracking/analysis.R

Note: *There are scripts, specifically those below, are rebased from #3, so can be ignored in this script.

  • surveyanalysis.Rproj
  • src/utils/read_responses.R
  • src/.Rbuildignore

Checklists

General checks

  • Code runs
  • Developments are secure and ethical
  • You have made proportionate checks that the code works correctly
  • Test suite passes
  • Assumptions, and caveats log updated (see docs/aqa/assumptions_caveats.md), if necessary
  • Minimum usable documentation written in the docs folder

Project-specific checks

These are additional checks determined necessary according to the analytical quality assurance plan
(see docs/aqa/aqa_plan.md).

  • The visualisations are appropriate for the task in hand.
  • There are no code errors which lead to dodgy visualisations.

This is so we can link up the data.
This is so we can get in a format ImpactEd are expecting.
This is inspired by Duncan's code.
This is so we can pivot wider without creating lists.

Also, export to csv for sharing.
This is to update packages used in scripts.
This is because the second script cannot call the first script (causes RStudio to abort). One suspects it is to do with Google Authorisation not being able to apply across scripts.

Move function to separate script so it can be tested more easily and to modularise code.
This is so we can see how respondents answered over time.
This is so we can see how people have responded over time.
This is to show how the share of responses change between each month.
@avisionh avisionh marked this pull request as ready for review January 5, 2021 01:45
@avisionh avisionh mentioned this pull request Jan 10, 2021
9 tasks
Comment on lines +88 to +97
# no. of surveys they complete
n <- 4

# no. of levels (likert)
lvls <- list(1:5)

l <- rep(x = lvls, n)

# get no. of unique permutations
combos <- expand.grid(l)
Copy link
Contributor Author

@avisionh avisionh Jan 10, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Never got round to completing this so feel free to ignore.

Essentially, i wanted all the possible combinations that people can answer questions.

For instance:

  1. Jan -> 'Very likely', Feb -> 'Likely', Mar -> 'Likely', June -> 'Very likely'
  2. Jan -> 'Very likely', Feb -> 'Very likely', Mar -> 'Likely', June -> 'Likely'
  3. ...

Then i wanted to calculate the percentage of people who responded in these way out of all those who have responded. I would then visualise each response as a node and the percentages as values for the edges.

@avisionh
Copy link
Contributor Author

Merging and closing as Owen seems happy with it.

@avisionh avisionh merged commit 6751d64 into main Jan 18, 2021
@avisionh avisionh deleted the feature/long-tracking branch January 18, 2021 08:38
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant