"There are three kinds of lies: Lies, damned lies and statistics."
(Benjamin Disraeli [a politician] as attributed by Mark Twain]
“There are three kinds of liars: Liars, damned liars and politicians”
(anonymous statistician)
This practical course provides a user friendly introduction to the fundamental concepts of quantitative analysis in the social sciences. It will cover underlying principles, terminology, research design, sampling strategies, uncertainty and missing data, computerised data management and analysis and univariate and multivariate approach to data analysis. Its primary goal is to provide you with the basic tools that social scientists use to collect, organise and analyse quantitative data. Thus, the goal is not to transform you into a statistician, but to provide a foundation of quantitative skills that will enable you to make an informed choice regarding the analytical approach that you take to your own research.
Since the students in this module come from diverse backgrounds, we will begin with simple techniques and then present more advanced topics. For those of you who do not have statistical training, consider this a rare opportunity to undertake a gentle introduction to quantitative methods. For those of you who have had statistical training previously, please consider this an opportunity to reinforce the fundamentals of what you have already studied. Please note that advanced statistical training modules will be provided within departments for students who will rely heavily on quantitative methods in their own research.
Many students fear courses in statistics because they not feel they have the necessary mathematical background or ability. We understand this fear and we will offer remedial tutorials for students who desire them. Other students believe that their research does not require quantitative methods. For these students it is important to note that statistics permeate every aspect of our life, from school league-tables to sports trivia. Statistics are a part of our culture: they form a basis for decisions in business, government and science. It seems reasonable, therefore, that social scientists should be conversant in them.
- To introduce fundamental concepts of quantitative analysis
- To introduce basic statistical terminology
- To discuss how formal, quantitative analysis differs from “common sense”
- To describe the role that uncertainty plays in the statistical analysis of data
- To increase awareness of the limits and pitfalls of quantitative approaches
- To raise issues of research design, particularly with regard to power analysis and sampling strategy
- To prepare the student to be an educated consumer of quantitative analyses
- To introduce computerised data management and analysis
- To provide a working familiarity with common univariate tests
- To introduce the multivariate approach to data analysis
- To introduce the pitfalls of missing data and describe ways of minimising them
Students who perform well in this module will:
Demonstrate a knowledge of:
-
the advantages of quantitative analysis vs. “common sense”.
-
basic statistical concepts and terms.
-
the merits and limitations of various research designs.
-
role of quantitative analysis in behavioural and social science.
Have developed the competence to:
- recognise flawed statistical approaches & interpretations.
- perform power analysis, including estimate sample sizes required to detect a given magnitude of effect.
- formulate an appropriate sampling strategy for a given research question.
- manipulate databases with Excel and SPSS.
- perform and interpret statistical analyses using Excel and SPSS.
- present summaries of data in tables or charts using Excel and SPSS .
Assessment will be in the form of weekly practical tasks predominately completed within the timetabled class (for further details - see course outline below). . The course work will consist of 8 exercises designed to assess your knowledge of the concepts and methods that are presented in the course. The computer work for each exercise will be performed in the practical sessions of the class, but some exercises will entail work outside class.
Top
The modules will consist of 10 meetings. Approximately half of each meeting will be devoted to lectures, with the remaining half devoted to practical work and tutorials.
Session 1:
- Issues: What role does quantitative analysis play in the behavioural sciences? What are the benefits and costs of using numbers to describe social and behavioural phenomena? What considerations should be taken into account when using numbers to describe social and behavioural phenomena?
- Practical: Introduction to the computer system. How accurate is “common sense”? Self-assessment of mathematical and statistical background.
- Assessment: Statistics & our understanding of the world: Choice behaviour in a guessing task
Session 2:
- Issues: What considerations should be taken into account when formulating a research design that will be analysed quantitatively? How do I specify and operationalise the variables that should be measured? What types of sampling strategy are available? What are their strengths and weaknesses?
- Practical: Summary of self-assessments conducted in meeting 1: Which ways of displaying the data are useful? Small group discussion regarding the design and implementation of a hypothetical study.
- Assessment: Design of a questionnaire study
Session 3:
- Issues: How can numbers be used to summarise the characteristics of samples and populations? How can we use small samples to infer characteristics of larger populations? What information is lost when we use summary statistics to described a sample or population? What are the benefits of using summary statistics?
- Practical: Use of SPSS and Excel to summarise data. Exploring distributions with Excel.
- Assessment: Summary analysis and interpretation using SPSS
Session 4:
- Issue: How can we manage large data sets of both numerical and textual information?
- Practical: Databases and pivot tables in Excel. Data management in SPSS Converting, exporting and importing data into Excel and SPSS
- Assessment: Using Pivot Tables to manage and summarise data
Session 5:
- Issues: How can we infer whether an apparent pattern in a sample is likely to arise from a given population? How can we detect unusual cases (sometimes called outliers) in a numerical data set?
- Practical: Visualisation and numerical determination of outliers in Excel
- Assessment: Calculation and interpretation of standardised scores in MS Excel
Session 6:
- Issues: How can we draw conclusions about relationships among variables from tables of frequencies? How can we draw conclusions about relationships among two continuous variables?
- Practical: Contingency table analysis in SPSS Correlation in SPSS and Excel
- Assessment: Power analysis using G*Power
Session 7.
- Issues: How can we determine whether an apparent differences among means of two or more samples reflect differences in the corresponding populations?
- Practical: ANOVA in SPSS
- Assessment: Calculation and interpretation of correlation and ANOVA in SPSS. Comparison and contrast of correlation and ANOVA
Session 8.
- Issues: How can we make quantitative predictions based on the characteristics of a given sample?
- Practical: Regression in SPSS. Curve fitting in Excel
- Assessment: Fitting curves rather than lines
Session 9.
- Issues: What are the advantages and disadvantages of analysing more than one variable at a time? How can we use multiple variables to make quantitative predictions?
- Practical: Partial correlation in SPSS. Multiple regression in SPSS
- Assessment: Interpretation of the results of multiple regression.
Session 10:
- Issues: The world is messy: How do we analyse incomplete data sets?
- Practical: Comparison of approaches to missing data using SPSS
- Assessment: Analysis and interpretation of statistical analysis of an incomplete data set.