jrosen48 / airs-ngss-survey

Home Page:https://jrosen48.github.io/airs-ngss-survey/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

airs-ngss-survey

Background

This is a repository for an analysis of the extent to which science teachers across the United States consider themselves to be teaching the Next Generation Science Standards.

This is an example repository for a project that involves the analysis of summary data from a large-scale assessment. It uses data from the following source:

The project uses a number of tools to address a number of challenges and/or to leverage certain opportunities:

  • git/GitHub for sharing code publicly in a way that promotes collaboration and transparency
  • Open Science Framework (OSF) for
  • OSF for sharing data (even large datasets) in either a private or public repository
  • OSF for sharing products, including reports and pre- and post-prints
  • {osfr} for linking between these efficiently

How to reproduce this analysis

1. Access this repository

Clone or download this repository from its homepage.

2. Open the R Project in this repository

This step requires having R and RStudio installed first.

Open the project by clicking on airs-ngss-survey.Rproj. This opens the project in RStudio.

Install the required packages by running the following function in your R console:

install.packages(c("here", "tidyverse", "readxl", "janitor", "rmarkdown", "osfr", "fs", "targets"))

3. Running the analysis

To run the project pipeline, enter and run the following function in the RStudio console:

targets::tar_make()

This will run all of the analyses and create the output, including an R Markdown report.

This includes accessing the data from OSF (available here) via the {osfr} package.

If any part of the analysis that a later parts depends upon changes, then running the above command will re-run all of the analyses from that point forward.

Note on data. While private data is often stored in OSF repositories, public data may also be; the targets plan will download the two required files (in the /data directory). If this data were a part of a private repository, you would first need to be granted access to the repository. Then, you would need to authenticate your identity via osfr.

About

https://jrosen48.github.io/airs-ngss-survey/


Languages

Language:R 100.0%