Quantitative methods for social sciences
Luke Sloan, University of Cardiff
Published on Jorum February 2012, copied to Economics Network June 2015
These files were made available on Jorum under a Creative Commons Attribution-Noncommercial-Share Alike 2.0 UK: England & Wales licence
Asking Good Questions (Powerpoint)
This presentation introduces students to research questions, hypotheses and data collection questions. Specifically it deals with the problems of falsification, access, ambiguity and perception (concentrating on research questions). The part of data collection questions takes the form of an interactive activity which can be run in a lecture of 90-120 mins. Encouraging students to design 'The Worst Question' is a great way of getting them to think about the rules they need to break.
Using and Sourcing Secondary Data (Powerpoint)
This presentation covers the advantages and disadvantages of using secondary data. It also identifies multiple sources of quality secondary data.
Coding Surveys (Powerpoint)
This presentation outlines the process of coding data from paper to electronic formats. Specifically we introduce levels of measurement, research design, 'other please specify', missing data and resolving difficult variables. As well as providing example of coding we also offer real-life comedy answers received in a recent survey.
Coding with NVivo8 (Powerpoint)
This presentation introduces coding in NVivo8. It briefly covers the generation of codes in qualitative data and provides numerous examples of why NVivo is a useful tool (opposed to doing it by hand). Towards the end is a step by step screenshot guide to importing external documents and creating free nodes in NVivo8.
Mixed Methods (Powerpoint)
This presentation provides and overview of mixed methods as a methodology in the social sciences. Crucially the slides are concerned with issues of research design - not analysis. The presentation ends with some seminar questions to encourage students to think about methodological plurality.
Sampling and Selecting (Powerpoint)
This presentation covers probability and non-probability sampling in the social sciences. It also addresses the issue of selecting case studies and the problems of non-response bias and how big a sample should be. There are two group activities at the end which are suitable for a 50 min seminar.
Hypotheses and Chi-Square (Pearson) (Powerpoint)
This presentation covers the formulating of hypotheses in relation to research questions and the application and interpretation of the chi-square test for independence. It can be delivered in 50 mins.
Hypotheses, Probability, Chi-Square and T-Tests (Powerpoint)
This presentation covers hypothesis generation, introduction to probability (p-values), the chi-square test for independence and the independent and related samples t-tests. There is a strong focus on interpretation of the test outputs from SPSS and the assumptions of the t-tests.
Correlation and Simple Linear Regression (Powerpoint)
This presentation covers correlation and simple linear regression. It contains many graphs and images to illustrate points and begins to introduce the generalised linear model through teaching students 'how to draw a line'. Particular attention is given to interpreting SPSS output.
Cross-Sectional and Longitudinal Designs (Powerpoint)
This presentation looks in-depth at cross-sectional and longitudinal designs. It also looks at the problem of establishing causality (rather than correlation) in social science and uses a video from GapMinder to show the power of temporal data. It uses major social science surveys to demonstrate the types of topics and questions that can be addressed by the different research designs. The activity at the end of the presentation is suitable for a seminar group - it's harder than it sounds!
Questionnaire Design (Powerpoint)
These slides cover the basics of good questionnaire design including: measuring social concepts, validity, reliability, hypotheses and questions, examples of quantitative surveys and formatting. This takes a holistic approach focusing not just on asking good questions but also on operationalising concepts and making surveys 'look attractive' to respondents.
These slides introduce the concepts of: levels of measurement, frequency tables, mean, median, mode, range and standard deviation. The presentation starts with a brief recap of the nature of data in the social world. At the end are a series of headlines that use statistics - students are encouraged to think about how the data used to generate these 'facts' may have been manipulated or misconstrued. This lecture can be delivered in 50 mins.
Research Designs for Quantitative Methods (Powerpoint)
These slides provide an overview of how research designs inform quantitative methods including experimental, cross-sectional, longitudinal, case study and mixed-methods (hybrid) designs. Although the examples are quantitative in nature it should be stressed that there is nothing inherently quantitative about any research design strategies.
Logistic Regression (3 Powerpoints)
These presentations cover both binary and multinomial logistic regression and use examples from the Social Capital Teaching Dataset (ESDS). The first presentation discusses the theory behind logistic regression, why it is so important to the social sciences and how to select and specify variables for a model. The second presentation covers how to run and interpret a binary logistic regression model in SPSS. The third presentation covers how to run and interpret a multinomial logistic regression model in SPSS. Each presentation can be covered in 50 mins if the class is confident with the material. All three presentations should be supported with a following computer workshop.
This booklet provides a guide to the basic functions of the online survey tool SurveyMonkey (www.surveymonkey.com) including guides on using the most common question types.
Student Guide to NVivo (PDF)
This booklet provides a guide to basic functions in NVivo (circa version 8) including: creating a project, importing different source types, nodes and coding, exploring data.
Student Guide to SPSS (PDF)
This document provides a beginner's guide to using SPSS (circa version 18). It is very basic but useful for informing students which buttons to push (it does not cover interpretation of output in any great depth). It includes screenshots and step by step guides for: opening an SPSS file, creating a database, entering data, data screening, central tendency, frequencies, modifying variables, crosstabulations and chi-square, independent samples t-test, paired samples t-test, simple linear regression, multiple linear regression, exploratory factor analysis.