Mixed Methods Tools for Content and Sentiment Analysis

Details:

The AQMeN research team based at the University of Strathclyde researched the impact of widely accessed social media discussion threads on people's sentiments and opinions towards Scottish Independence (and other constitutional options), nations and nationality (and regions), political leaders and their parties. 

Sentiment analysis (or opinion mining) refers to the use of natural language processing, text analysis and computational linguistics to identify and extract subjective information in source materials. This event aimed to introduce content and sentiment analysis and some of the mixed methods tools available for conducting this type of research.

This event was designed for doctoral students, academic staff and non-academic researchers with an interest in carrying out content and sentiment analysis through both qualitative and quantitative approaches. 

Course Details:

Part 1: Introduction to Qualitative Data Analysis using NVivo

NVivo software supports qualitative and mixed methods research and allows you to collect, organise and analyse content from a range of sources including social media. This half-day workshop will introduce the basics of Qualitative Data Analysis and demonstrate the use of NVivo as a tool for content and sentiment analysis. The session will highlight issues that should be addressed before beginning analysis and demonstrate core features of NVivo. This workshop will be a mixture of theoretical and hands-on practical sessions and will cover the following topics:

•Introduction to Qualitative Data Analysis

•Different approaches to content analysis: qualitative vs quantitative

•Different tools for content analysis: NVivo vs R-tm

•Issues to consider prior to beginning analysis with NVivo:

◦data type and issues with NVivo

◦types of content analysis and mixed methods

◦coding issues and coding steps

◦reliability and tests for reliability

•Core features of NVivo:

◦Label and import data

◦Explore theory through Nodes

◦Query words and connect to Nodes

◦Apply data visualisation through models

Part 2: Introduction to R workshop

R is an open-source (i.e. free) suite of packages that enables a very wide range of statistical calculations to be performed. It can import data from other software (such as SPSS, Stata and Excel). The training will show how to use R to carry out basic and intermediate statistical methods that are common in social science, for example t-tests, non-parametric tests, analysis of variance, linear regression, and logistic regression. Participants in the training must be familiar with these methods, because the attention of the training will be on how to use R to implement them, not on the methods themselves. Attention will also be given to using R for basic statistical summary and presentations – for example, constructing tables and drawing graphs.

This session will provide pre-requisite training in preparation for the third workshop in this event which will focus on text mining using R (see below). Applicants with experience in R may wish to miss out this part of the event but would be required to demonstrate existing experience of using R in their application prior to attending Part 3.

Part 3: R-tm Sentiment Analysis workshop

This half-day workshop will build on the previous workshops and will teach participants how to conduct text data mining analysis and content analysis using the text mining package R-tm. This workshop will be a mixture of theoretical and hands-on practical sessions and will cover the following topics:

•R-tm: the text mining package for R

◦Data retrieval from SPSS and Twitter

◦Pre-processing and analyses

•Tm-plugin-sentiment

◦Introduction

◦Example runs

•General Inquirer coding

•Machine learning approaches to text coding with RTextTools

◦Training with coded data

◦Applying to un-coded data

Request training materials
View all courses available