Reference: Schotter, A.,Buchel, O. & Vashchilko (Lukoianova), T.(2017). Interactive visualization for research contextualization in international business.
Journal of World Business. DOI: 10.1080/01639374.2012.682254


The IB Visualization Toolbox provides a set of tools to build interactive visualizations of IB phenomenon in multiple contexts. In How To... we describe the major packages and tools that we use to build IB Interactive Visualizations of IB Network Explorer, IJVs and Political Risk, IJVs and Political Events, Hexabinning IJV Cluster Analysis, and Movement Vectors (Vector Models and Interactive Lenses). To get started and build more extensive interactive visualizations, we explain the main steps for network visualizations, map visualizations, and charts/graphs along with the required web-tools (e.g., a JavaScript visualization library D3.js, Google Maps API), R packages, and their multiple combinations in Advanced Tools.

Japanese MNEs in China
Hexbin Map of Japanese MNEs in China
International Joint Ventures 1985-2010
Economic cores and peripheries
International Joint Ventures bubble map
Magic lense zooming into Shanghai MNEs
Adjacency matrix of International Joint Ventures
International Joint Ventures 1985-2010
Standard Deviational Ellipses International Joint Ventures 1985-2010
Chord digram of International Joint Ventures

Getting Started with R

1) Download RStudio Desktop from this link.

2) Learn how to set up a working directory for storing and retrieving your datasets: link.

3) Things you need to know to upload files to R: link.

4) Things you need to know before you start working on a network visualization: link.

IBPackage for visualization

Our IBPackage has a wide selection of R packages for visualization: "shinydashboard","shiny","d3heatmap","corrplot","ggplot2","plotGoogleMaps","sna","network","igraph","statnet","visNetwork","networkD3","leafletR","plotly".

All of these packages have some graphical capabilities: they can produce interactive and non-interactive representations.

The package can be downloaded from this link and installed in RStudio with the following command: install.packages("IBPackage_0.1.tar.gz", repos = NULL, type="source"). Not all packages work under R version 3.2.3, however.

To load the packages, submit the following command in R ipak() after you install and load the IB package.

Network Visualizations

For basic non-interactive network visualization we suggest you to use the following R packages: igraph, sna, and network. These packages include many functions available in UCINET software. Use network package to create and modify network objects and vertex/edge/graph attributes from relational data types. Use igraph and sna for modeling and analyzing networks. If you want to implement recent advances in the statistical modeling of networks based on Exponential family Random Graph Models you may be interested to use statnet R package. Unlike UCINET or Pajek visualizations, statnet is more suitable for modeling than visualization or statistical analysis.

For network visualization, check out ntdv, visNetwork, and networkD3 R packages. ndtv can help you with temporal network visualizations, networkD3 can help visualize networks with D3.js, and visNetwork can help you animate your network visualizations.

The following tutorials will help you get started with igraph, sna and interactive visualization packages:


Sources of geographic data (shapefiles):

Tool for coverting geographic files from one format to another: Mapshaper

There are several packages in R that allow to create interactive maps. These are: RGoogleMaps, leafletR, plotGoogleMaps. Primers how to use these packages can be found at:,,

Special types of maps:


Interactive charts and graphs can be elegantly drawn with the following R packages:

An R-package Shinydashboard allows easily create interactive dashboards, whereas an R-package googleVis helps to create interactive charts based on R data frame by providing an interface between R and Google Charts API.

To visualize a correlation matrix or any matrix Corrplot R package could be extremely helpful, and could be complemented by ggplot2) to further explore patterns in data through 2- and 3-dimensional graphics.

Visualizing R with D3.js

Once you learn how to visualize with R packages, you may want to do something more elegant and visualize your data using amazing D3.js library. This tutorial will help you to get started with D3.js and R.

You may also use plotly R packages to visualize your data as graphs, charts, and other visual representations. A cheatsheet for how to use plotly package can be found here.

If you want to develop an interactive visualization with R that continously communicates with R packages, you should use Shiny package (documentation is available at or set up your own R server with OpenCPU package.

IB Network Explorer

interactive network visualization

Examining dyadic data by means of traditional statistical and network visualization allows researchers to investigate networks only partially. This is particularly the case when dyadic data are at the country level or other more general location level. The IB Network Explorer offers an improved exploratory view of dyadic datasets. It augments network analysis with interactive spatial and visualization techniques and tells a story of the development of the network under investigation over time without forcing researchers to pre-select representations or to re-run analytics when adjusting certain subsets of the data. Additional examples/insights gained from this visualization are described in the online manual ( ).

Link to IBNetwork.

International Joint Ventures (IJVs) and Political Risk

The visualization shows a relationship between the formation of IJVs and the levels of political risk in different countries of the world. All graphs in this dashboard are connected: changes in one graph affect representation of other graphs. Using these graphs, we found that after 1995, there were no IJVs formed in countries with the political risk greater than 70 even though political context characterized by political risk above 70 was present after 1995.

Link to International Joint Ventures (IJVs) and Political Risk dashboard.

International Joint Ventures (IJVs) and Political Events

The visualization shows the spatio-temporal distribution of IJVs (which are represented as blue bubbles) and political violence events* (represented as a green-and-red heat map layer). The map has an interactive time slider that controls aggregation of IJVs and political events on the map. Reducing or increasing the time range on the slider changes the representations of blue bubbles and the heat map, which either increase or decrease in size or intensity depending on the selected time period. The two histograms above the time slider juxtapose the temporal distribution of IJVs in the world (blue color) in 1985-2010, and the total number of political violence events in the world across in 1985-2005 (red color). The histograms allow researchers to examine changes in political violence and the number of IJVs at the global level. The data indicates that in the middle of the 1990s, the total number of political event in the world was in a minimal range, and at the same time the total number of IJVs has reached its overall maximum.

* We define political violence as part of “contentious politics” or collective political struggle, which includes such things as revolutions, civil war, riots and strikes, but also more peaceful protest movements (O’Neil, 2015).

Link to International Joint Ventures (IJVs) and Political Events.

Hexabinning IJV cluster analysis

Hexabinning is an efficient data grouping technique that aggregates data in the form of hexagons. It is particularity feasible for representations of large datasets. Hexabinning reveals densities, highlights real-time empirical contours, and visualizes hot spots of activities. Dynamically zooming in and out on the map allows researchers to explore how aggregations change geographically.

Hexbins reveal densities, highlight real-time empirical contours, and show hot spots of activities. Zooming interaction on the map allows researchers to observe how aggregation depends on geographic scale.

Movement Vectors (Vector Models and Interactive Lenses)

The visualization shows Japanese MNEs as sets of footprints consisting of multi-colored segmented lines and points. Each color represents a certain sequence in which an MNE entered the Chinese market and then expanded. However, when looking at the map it becomes apparent that patterns created by footprints are too cluttered and hard to understand or make sense of without filtering. The visualization has filtering by sequences and interactive lenses. Applying filtering to sequences allowed us to identify several hub locations which we call FDI cores from which MNEs expand in China. The visualization also showed us to observe that MNEs do not expand homogenously across all of China, despite several institutional changes that the government implemented to promote inland investments. With the help of the interactive lens we were able to explore MNEs that started outside of the regions in finer detail, leaving only those that started within the selected region.

Link to Movement Vectors (Vector Models and Interactive Lenses).

How To ...

Most Useful for IB Visualization Tools

Visualization Visualization Use Application in IB Example Link to R-Resource or other software
1. Map Geographic mapping - mapping of objects on a geographic world map Visualization of any business entities and the characteristics of their international business environment on a geographic world map MNE subsidiary networks, MNE supplier networks, MNE IJV networks, MNE affiliates networks, FDI network, international trade network, and international business environment characterized by various politico-economic between or within countries Mapbox API, Google Maps JavaScript API
2. Vector Filtering (e.g., (1) by categories (e.g., values of a variable or its attributes), (2) by sequences, (3) by groups (for example, clustered with spinglass algorithm)

(1) Filtering by categories:

visualization of only chosen categories from all the categories provided;

(2) Filtering by sequences: visualization of a sequential movement of an entity or object from one geographic point to another (or between geographic points);

(3) Filtering by groups (for example, clustered with spinglass algorithm): visualization of only chose groups from all the groups derived after the application of spinglass algorithm

Visualization of sequential opening of subsidiaries in different geographic locations (by sequence filtering);

Visualization of different categorical variables that take different values (by categories filtering)

Sub-samples of the MNEs, their subsidiaries or networks, host or home countries, regions,etc.

Google Maps API

JavaScript (groups clustered with spinglass algorithm; categories)

jQuery UI API - slider

3. Interactive Lenses Interactive technique for filtering any type of markers (e.g., bubbles) on a map to select a group of entities To demonstrate variations in the spatial distribution of a phenomena (interaction with the number of IJVs hosted by every country Sub-samples of the MNEs, their subsidiaries or networks, host or home countries, regions, etc. Circle in Google Maps with JavaScript filtering
4. Raster Models

Visualization of agglomerations of objects to identify cores and peripheries of their linked activities

Identification of cores and peripheries of business activities MNE subsidiary networks, MNE supplier networks, MNE IJV networks, MNE affiliates networks, FDI and international trade network, and any other politico-economic conditions between or within countries characterizing international business environment Heatmap in QGIS transformed into a grid file, which is visualized as a custom-made overlay on Google Maps, contoured with conrec algorithm in JavaScript
5. Graduate Symbols To denote the groupings of objects with similar sizes by corresponding graduate symbols Differentiate business entities across their sizes MNEs and their affiliates; home and host countries; regions Google Maps API, JavaScript (custom-made)
6. Heat Maps To visualize the spatial diffusion of objects

To compare business entities by the densities of some of their characteristics

Spread of MNEs around the world; intensity of political violence characterizing business environment

Google Maps API



To visualize the distribution of phenomena across time

To examine the temporal distribution of any phenomena at different levels of analysis

Distribution of the IJVs, FDI inflows, MNEs over time

JavaScript, D3.js bar chart custom-made layout


Standard Deviational Ellipses

To visualize spatial distribution of entities on a geographic map

To demonstrate the direction of the current and future trends of a phenomenaÕs spatial spread

MNEs and their affiliates or headquarters, IJVs, mergers and acquisitions


Mapbox API. JavaScript with the computation of the standard deviational ellipses


Hexbinning (hexagon maps)

To visualize densities, real-time empirical contours, and hot spots of activities by employing hexagon binning, the most efficient and compact division of two-dimensional spaces.

To overcome the limitations of scatterplots and maps in visualization of large datasets

MNEs and their affiliates or headquarters, IJVs, mergers and acquisitions

Leaflet API with a plugin from


Space-Time-Cube Model

To represent the flow of time effectively, space and time are considered in a three-dimensional representation, with space on the x and y axes and time on the third dimension

To visualize business activities in time to reveal additional temporal patterns of spatially mapped firms

FDI inflows, MNE operations

Uncharted Software Inc. GeoTime


Adjacency Matrix

To visualize more or less tightly connected nodes of the network with more or less intense non-black color, if the nodes belong to the same group (based on any clustering algorithm, which is in IJV Network case is spinglass algorithm); and to visualize the links between nodes clustered in different groups with more or less intense black color depending on the link intensity.

Visualizing the within- and out- of groups links between any entities clustered into groups within any business network

MNE subsidiary networks, MNE supplier networks, MNE IJV networks, MNE affiliates networks, FDI network, international trade network

Adjacency matrix layout in D3.js


Chord Diagram

To simultaneously visualize (1) the distribution of a phenomena across objects (e.g., the sizes of arcs corresponding to each host country represent the number of hosted IJVs by that country) and (2) the magnitude of the relationships of each pair of objects (e.g., the width of the connecting lines between the arcs represents the number of IJVs opened in a host country by investors from a particular home country) To show symmetries and asymmetries in the relationships among business entities along with the simultaneous demonstration of the distribution of a phenomena across business entities MNE subsidiary networks, MNE supplier networks, MNE IJV networks, MNE affiliates networks, FDI network, international trade network Chord diagram layout in D3.js

Multilayered Pie Charts

To visualize relationship of parts out of a whole as percentages and proportions between categories of a phenomena by splitting a circle into the corresponding segments;

Pie charts might have different layers to simultaneously visualize different attributes of an object with different colors of the slices, different coloring of a circumference of a pie chart, etc.

To compare categorical variables with 2-3 categories across multiple contexts if multilayered pie charts are employed (e.g., simultaneous visualization of in-, out- degree centrality as well as alpha-centrality)

MNE subsidiary networks, MNE supplier networks, MNE IJV networks, MNE affiliates networks, FDI network, international trade network

Custom-made overlay for Mapbox Map written in D3.js
valign="top">14. Timelines (or temporal filtering) To filter and visualize objects by time range Visualization of temporal distribution of phenomena MNE subsidiary networks, MNE supplier networks, MNE IJV networks, MNE affiliates networks, FDI network, international trade network HTML5 slider controlled with JavaScript

jQuery UI API -- slider

D3.js brush with JavaScript filtering

15. Polygons Selection of an area on a geographic map instead of a point for further employment of various interactive visualization tools Identification of geographic objectsÕ neighbors that could facilitate the detection of spatial movement and other related trends for a subsample selected with a polygon MNE subsidiary networks, MNE supplier networks, MNE IJV networks, MNE affiliates networks, FDI network, international trade network Google Maps Visualization API Leaflet Drawing API, Mapbox API