In mandel94/easyAnalysis: What the Package Does (One Line, Title Case)

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)

easyAnalysis

WHAT?

The goal of easyAnalysis is to provide an interface for implementing different types of analysis on qualitative data. The focus is on making the analysis as ease as possible, by abstracting away any low-level implementation and giving the analyst visual control over the analysis pipeline.

WHY?

One: code-free! easyAnalysis makes the analysis independent of the analysts' programming skills; Two: 'intelligence' is king! easyAnalysis helps to focus on the 'intelligence' part of the problem, using advanced analysis techniques without the need to worry about complex technical implementations; Three: try and retry! Lengthy scripts tend to hinder the analysis flexibility and reproducibility. Since data analysis is a process of trial and error, reproducibility of the pipeline is a fundamental driver for extracting good insights. The visual interface of easyAnalysis provides all the reproducibility you need to elevate your skills as an insight chaser!

Installation

You can install the released version of easyAnalysis from CRAN with:

install.packages("easyAnalysis")

Things you should know

• The identifier should be the first column of the imported data frame;
• easyAnalysis supports only .csv files;
• As for the data, easyAnalysis currently supports two types of data, coherently to the only type of analysis it can perform (MCA):
  • binary data (in the '1/0' or 'yes/no' formats);
  • multi-categorical data.

How to use easyAnalysis

Let's explore the different sections of the user interface.

Import

Select CSV file

here you can browse your local system for a CSV file to load.

Import

Click here to import into the application the CSV file you previously uploaded.

Select variables

Once the data has been imported, the list of variables will be available here. Select all the variables you are interested in.

Include selected variables

Click here and a table with imported data for the variables you selected will display on the left side of the screen.

Clear

Click here to clear the currently displayed table (only the ID_CODE will display)

Remove rows with NA's

Check this box if you want to remove all rows with at least one missing value. This is a pretty brutal approach, as you are likely to lose a lot of usable information. We recommend that you consider imputation methods for filling missing values (easyAnalysis does not currently support imputation methods in its interface).

Glossary

MCA: Multiple Correspondence Analysis is a technique of analysis widely used in a variety of research fields (typically in social and behavioral sciences), as a way to find associations in categorical data.

A typical application is with survey responses, in which each feature corresponds to one survey question, each one admitting a limited set of responses (just 'yes/no' for binary questions). It allows summarizing most of the variation in data with a limited set of factors called dimensions (like PCA, it is a way for reducing the dimensionality of the problem at hand). Each dimension is influenced by a subset of the original variable pool and represents a hidden phenomenon, that manifests itself in the form of an empirical association between the variables belonging to that subset. In brief, MCA can be a useful tool if we want to understand how responses from a survey are associated, and to extract a relevant insight from that association. For example, MCA could find an association in the response Sex: female and the response Entertainment: theater, and so on.

Active variable:

in the context of multiple correspondence analysis, a variable is said to be active when it influences the resulting biplot coordinates. Active variables are the "construction materials" of the MCA factors.

Supplementary variables:

in the context of multiple correspondence analysis, a variable is said to be supplementary when it does not influence the resulting biplot coordinates. We might wish to include a variable as supplementary when we want to study how that variable can be explained in terms of the factors being extracted, without influencing those factors.

Cos2

In the context of MCA, cos2 is a value in the interval [0, 1] associated with a given entity (a variable, or an individual), that measures how important a particular dimension is for that entity. Each entity is assigned multiple measures of cos2, one for each dimension extracted from the MCA. Cos2 can be useful for interpreting the dimensions of an MCA. In general, if variables are the entities, we can interpret a dimension by looking at which entities have the strongest cos2.

Example

You'll still need to render README.Rmd regularly, to keep README.md up-to-date. devtools::build_readme() is handy for this. You could also use GitHub Actions to re-render README.Rmd every time you push. An example workflow can be found here: https://github.com/r-lib/actions/tree/master/examples.

mandel94/easyAnalysis documentation built on Jan. 2, 2022, 9:52 a.m.