R Programming and Bibliometrics

Introduction

  • This module is part of an R programming certificate series.
  • The course is beginner-friendly, no prior experience in bibliometric analysis or programming is required.
  • Taking notes is mandatory and must be submitted to project managers, as it will impact individual research project placements.

About the Instructor

  • Taylor Pape is the chief data officer at ThinkNeuro.
  • Bachelor's in Biology with minors in Computer Science and Data Analytics from New Jersey Institute of Technology (NJIT).
  • Played soccer for four years.
  • Pursuing a Master's in Advanced Computer Science at the University of Essex.
  • Planning to pursue a PhD in computational biology, focusing on neuroscience.

Research Experience

  • Began research at NJIT's Biosmart laboratory, building biosensors for pain quantification.
  • Transitioned to computational research at NJIT's STG laboratory of neuron and circuit dynamics.
  • Currently running a project computationally modeling neural circuits in crustaceans.

What is Bibliometrics?

  • Bibliometrics is the quantitative analysis of scientific publications.
  • It involves using numbers to study published science.
  • Includes analyzing:
    • Number of papers published on a topic.
    • Citation frequency of papers.
    • Common keywords.
    • Collaboration networks (researchers, institutions, countries).
  • Bibliometrics is "the science of measuring science."

Real-World Example: AI in Neuroscience

  • If curious about AI in neuroscience, bibliometrics can:
    • Analyze papers from the last 10 years to see research growth.
    • Identify popular keywords like deep learning and brain imaging.
    • Determine which journals publish the most on the topic.
    • Identify top authors and institutions.
  • This helps researchers, students, and policymakers understand the field's current state.
  • Guides literature reviews to identify research gaps and inform data-driven decisions.

Why Bibliometrics Matters

  • Tracks scientific trends:
    • Shows the growth or decline of research areas like AI in medicine over time.
  • Identifies influential research:
    • Highlights most cited articles and leading authors.
  • Reveals collaboration patterns:
    • Shows which researchers, institutions, or countries collaborate.
  • Guides literature reviews:
    • Helps find the most relevant and impactful papers.

Example: Impact of Data Sharing in Neuroscience

  • Research studied the impact of sharing digital reconstructions of neuromorphology via neuromorpho.org on article citations.
  • Two groups were compared:
    • Sharing articles (data publicly available): ~1,600 papers.
    • Unsharing articles (data not shared): ~3,000 papers.

Graph A: Yearly Citation Distribution

  • Papers grouped by citations per year (logarithmic scale).
  • Blue line: sharing articles.
  • Orange line: unsharing articles.
  • Sharing articles had a higher average citation rate.
  • Sharing articles averaged 8.91 citations per year vs. 6.19 for unsharing articles (44% increase).

Graph B: Citation Increase Due to Data Reuse

  • Distribution of citation increase due to data reuse.
  • Average increase: ~13.8%.
  • Data reuse explains less than one-third of the citation boost.
Takeaway
  • Sharing data boosts visibility, with increased exposure and credibility being the primary drivers of citation advantage.
  • Bibliometrics provides quantitative evidence of the impact of data sharing on research influence.

R vs. RStudio

  • R: The programming language that runs the code and performs calculations (the engine).
  • RStudio: An interface for writing, editing, and running R code (the dashboard).
  • Both are needed to get started.
  • RStudio automatically detects R.

Installation

  • Install R first.
  • Install RStudio second.
  • Default installation paths are acceptable.
  • In R installer, select the following:
    • Main files
    • 64-bit files
    • Message translations
  • Accept the defaults during R installation and name it as "r" for simplicity.
  • Create a desktop shortcut and save version number in registry during installation.

Important Notes

  • A Google form will be sent out every Wednesday in the Slack channel for questions to be brought to office hours; submit by the following Monday at 11:59 PM Eastern Standard Time.
  • Office hours: Wednesdays, 5:30 PM - 6:30 PM Eastern Standard Time on Zoom (link in Slack).
  • General questions: use the s u 25 questions channel on Slack.
  • Notes must be taken and submitted.
  • Citation for the research article used in the example is provided.

RStudio Environment

  • RStudio is organized into four main panes:
    • Console: where R runs code (like a calculator).
    • Script editor: where you write and save R scripts.
    • Environment: stores variables, data frames, and functions.
    • Files: manages files, plots, packages, etc.

Console

  • Type commands directly here and see the output right away.

Script Editor

  • Write and save R scripts.
  • Click run to send code to the console.

Environment

  • Stores variables, data frames, and functions.
  • History tab shows all previous actions.

Files

  • Manages files and shows plots.
  • Packages tab lists installed libraries.

Programming Basics: Assigning Variables

  • Assign the number 10 to a variable called citation:
citation <- 10
print(citation) # Output: 10
  • Assign a list or vector to a variable:
citation <- c(10, 20, 30)
print(citation) # Output: 10 20 30

Common Built-In Functions

  • length(): get the length of a vector.
length(citation) # Output: 3
  • sum(): sum the elements of a vector.
sum(citation) # Output: 60
  • mean(): find the mean of a vector.
mean(citation) # Output: 20
  • max(): find the maximum value in a vector.
max(citation) # Output: 30
  • sort(): sort a vector.
sort(citation, decreasing = FALSE) # Output: 10 20 30 (ascending)

Creating a Data Frame

  • Create a small data frame with titles, years, and citation counts:
titles <- c("paper a", "paper b", "paper c")
years <- c(2021, 2022, 2022)
citations <- c(5, 12, 8)
papers <- data.frame(title = titles, year = years, citation = citations)
print(papers)