Subject outline
Assessment items
APA Style Guides

Applied Statistics
in the Social Sciences

A free, online, university-level course

James Neill
Last updated:
23 Aug 2003

Introduction

Welcome.  This is a free, online, applied course on the use of statistics in the social sciences.  Examples focus on psychology, education, and human movement (kinesiology).

This online course offers two unique advantages over a more typical university on-campus class:

  • the pace of learning can be more closely matched to the individual student

  • non-linear learning paths can be pursued, thus appealing to a broader range of learning styles

The course material has been developed by a university lecturer, James Neill, who has tutored and coordinated in about half a dozen different university research design and statistics courses, at both undergraduate and graduate level, in Australia and the USA.  However, I am far from a mathematician and take more of a right-brained approach -- intuitive, visual, & applied.  I am less interested in formulae and mathematical theory than in the practical, social utility of research methods and statistical analyses.

Students' Past Experiences of Statistics Courses

The more I taught statistics to students, the more I realized that many students brought a high level of fear and anxiety to statistics -- more so than in  any other subject, it seemed.  Physical education, perhaps, provokes similar levels of anxiety for many.

So, in the first class, I spend time in a circle having students describe what past statistics courses they've taken, if any, how it went for them, and their level of confidence / anxiety about learning statistics. 

To the extent that a warm, empathic approach is taken by the instructor, I have found deeper and deeper revelations of highly negative past experiences of learning statistics -- in short, nothing less than a series of educational horror stories. 

I don't try to provide post-mathematical trauma therapy in this course, but I do try to help students build a positive momentum of confidence in their capabilities, based on developing real skills, including for example analyzing data sets and interpreting analyses.

Advice for Students

Throughout the course I strongly encourage students to:

  • seek help early when they come up against problems -- this helps to minimize the risk of having anxiety build and interfere with learning

  • on the other hand, it is important not to get overly caught up minor concepts which aren't entirely understood

Thus, I think its important that as the course proceeds, that it is understood which concepts are critical before proceeding, and which concepts is it desirable, but not necessary to grasp, at least for the time being.

Sometimes I find that it is helpful sometimes to take side-routes off main topics, particularly to explore an underlying concept.  If the student is struggling to understand a topic (e.g., t-tests), it is often because he/she doesn't understand a more foundational concept (e.g., normal distribution).

I have found in teaching statistics that if students come to understand the basic elements really well, e.g., the normal distribution and correlations, then learning more advanced concepts such as multivariate analyses of variance and multiple regression is relatively easy.

So, in the teaching content and practical tasks in this course, a solid understanding of the basic concepts (univariate statistics) is the core, with it being desirable that students have at least an introductory understanding of the (multivariate statistics).


Textbook
It is advisable to purchase a statistics reference text which provides a solid introduction to at least these topics:

  • normal distribution theory

  • descriptive statistics (e.g., mean, standard deviation, etc.)

  • t-tests & ANOVAs (tests of mean differences)

  • correlation, regression & multiple regression (tests of association)

  • non-parametric tests of mean differences and association

  • effect sizes and power

  • confidence intervals

Desirable, but not essential text-book topics:

  • scale construction & reliability analysis

  • factor analysis

  • meta-analysis

Why have a hard-copy text for an online course?

Computers don't always work, neither is the web super-reliable.  Also, it can be helpful to have a text open alongside the computer and to read both the online course material and the material in the textbook.  In statistics, reading two or more different versions can greatly aid understanding.  Finally, this course does not provide extensive formulae, statistical tables, referencing, etc. which should be part of your hard-copy statistics text.

Software
SPSS is used for many exercises; some are in Excel.

Questions

Consider the following questions and email your responses to the instructor (or group):

  1. What research methods or statistics courses have you studied before?

  2. In general, how easy or hard is maths and statistics for you?

  3. How confident are you about this course?

  4. What kinds of information or exercises seem to work best for you in learning concepts such as statistics and skills such as statistical analysis.

  5. What specific areas of research, theory, or practice interest do you have?  The more detail you provide, the more instructor can tailor examples and extra readings to your areas of interest.