Critical Thinking

REVISED: 2020

Author: Ed Nelson
Department of Sociology M/S SS97
California State University, Fresno
Fresno, CA 93740
Email: 
ednelson@csufresno.edu

© The Author, 2014. 2016, 2020. Last Modified, May 2020.

The goal of this exercise is to describe a data-based approach to teaching critical thinking.

These exercises use various data sets to teach critical thinking skills.  They include the General Social Survey and various Pew Research Center surveys. Critical thinking in this class refers to the ability to answer why questions such as why some people favor and others oppose gun control and why people vote the way they do. We focus on using the scientific approach to answer such questions although not all questions are best answered with this approach. Critical thinking also refers to the ability to develop organized and logical arguments and to test hypotheses using the scientific approach. At CSU Fresno, critical thinking is a GE requirement and is offered in a number of departments. In Sociology it’s also a requirement for the major. In my class the papers revolve around the analysis of survey data.  There are five main sections in the class:

  • scientific approach,
  • argumentation,
  • causal arguments and claims,
  • analysis of quantitative data and
  • fallacies or errors of reasoning.

In the first section of the course we discuss the scientific approach as one way of answering why questions.  During this part of the course we talk about such topics as concepts, measures, hypotheses, theories, and sampling.  There are many texts that could be used.  My choice is Making Sense of the Social World by Daniel Chambliss and Russell Schutt (Sage). 

Any course in critical thinking must deal with argumentation so that's the second section of the course for which I use The Elements of Reasoning by Ronald Munson and Andrew Black (Cengage Learning).  We cover topics such as recognizing, analyzing, and evaluating arguments, as well as valid and invalid argument forms (e.g., modus ponens, modus tollens). 

The third section of the course is where the first two sections come together – causal arguments and causal claims.  There are sections in both texts which we use here.  The focus is on testing causal claims which takes us into a discussion of experimental design and survey design.

It's in the fourth section of the course that students get the opportunity to put into practice what we have been discussing.  I try to choose topics that I think students will be interested in such as the legalization of marijuana, gun control, and same-sex marriage.  There are numerous sources of good data many of which are available at no cost.  Recently I have used the General Social Survey and various Pew Research Center's surveys.  Students write papers that ask them to formulate hypotheses, develop arguments to support their hypotheses, use the data to test their hypotheses, and decide if the data support their hypotheses.  Data analysis proceeds from univariate to bivariate to multivariate analysis where we discuss spuriousness. 

Here are links to the paper instructions and data sets for the last three semesters.

Paper instructions

  • Spring 2019 – opinions on homosexuality and same-sex marriage
    Paper 1     Paper 2  
  • Fall 2019 – opinions on gun issues
    Paper 1    Paper 2  
  • Spring 2020  – opinions of the legalization of marijuana

Data sets

  • Spring 2019 -- 2014 Pew Research Center's Religious Landscape Survey
  • Fall 2019-- 2016 Pew Research Center's Political Survey
  • Spring 2020 – 2018 General Social Survey (NORC); no data set is attached since we used the full 2018 GSS data set online

The papers start by asking students to explore some topic such the legalization of marijuana, gun issues, or same-sex marriage.  It typically asks students to run a frequency distribution for a variable that describes how respondents feel about the topic in that semester.  This variable becomes the dependent variable for their paper.  Then the paper turns to a discussion of possible independent variables.  Once they understand dependent and independent variables, then they select particular independent variables to crosstabulate with their dependent variable. The only statistics that students use in the first paper are frequencies and two-variable crosstabs with percentages. 

The second paper is a continuation of the first paper.  In recent years, the second paper has focused on religion and their dependent variable.  Religion includes both religiosity (i.e., how religious they are), religious preference (i.e., what religion they are), and other aspects of religion.  In addition to percentages, the second paper also uses Chi Square.  Towards the end of the second paper, the idea of elaboration is introduced where students add a third variable into their analysis as a control variable.  This allows us to discuss such ideas as spuriousness.  This link will take you to a short discussion of spuriousness which uses an example from the General Social Survey that focuses on the control of pornography.

Using a statistical package means that the students don’t have to worry about how to compute the statistics.  Moreover, they learn a little about the use of statistical packages.  I use SPSS for the statistical analysis but you could substitute any statistical package of your choice.

During the 2020 spring semester the pandemic occurred and classes became virtual.  Most of us had no experience teaching virtually and had to scramble to adapt to this new environment.  SPSS was only available in computer labs on campus and was not available to students online.  That left me in a quandary.  Fortunately, I was familiar with SDA (Survey Documentation and Analysis) which is an online statistical package that is freely available online.  Combined with a data set such as the General Social Survey which is also freely available online, I was able to make use of this option.  I also combined the two papers into one paper which was easy to do.

I don't use the entire data set provided by Pew and NORC. These are all very large surveys with many variables. I choose a subset of the variables and use all the cases. Some variables are recoded: composite variables are created out of existing variables.This is important because students are often overwhelmed by large data sets. I did use the full 2018 GSS due to time constraints as we transitioned to a virtual environment.

If you are interested in other topics and data sets, take a look at the Textbooks and Exercises page on the Social Science Research and Instructional Council's website.

The final section of the course is a short one-week discussion of various errors and fallacies using a chapter from The Elements of Reasoning. I try to use current events to illustrate these fallacies.

You are free to use the assignments and data sets as they are or to modify them in any way you wish.   Please feel free to contact me by email at ednelson@csufresno.edu.