Survey research and design in psychology/Tutorials/Correlation

Tutorial - Correlation

Resource type: this resource contains a tutorial or tutorial notes.

Completion status: this resource is considered to be complete.

This tutorial introduces statistical and graphical examination of linear relations between two variables. Non-parametric and parametric correlational techniques are explained. The tutorial builds on understanding of univariate descriptive statistics and graphing, and levels of measurement. Non-parametric and parametric linear correlational analyses are conducted in SPSS, with an emphasis on the importance of visualising the data, particularly scatterplots. Interactive correlation guess exercises are used to help make intuitive connection between correlations and scatterplots. In addition, the effect of outliers and non-linear relations are considered.

Contents

1. General steps
2. Types of correlation and level of measurement
3. Types of correlations - Exercises
4. Scatterplot correlation guess
5. Effects of outliers
6. Correlations and non-linear relations

Readings

1. Howitt & Cramer (2014)
   1. Part 2: Descriptive statistics (particularly Chapters 8 to 11)
2. Allen & Bennett (2010)
   1. Chapter 12 "Correlation"
   2. Chapter 16 "Non-parametric procedures"
      1. 16.3 Chi-square for contingencies
      2. 16.10 Cramer's V
      3. 16.11 Spearman's Rho and Kendall's Tau-B
3. Francis (2007)
   1. Section 3.1 "Relations Between Metric Variables"

Recordings

1. Virtual tutorial recording, 2018
2. Screencasts

See also

1. Introduction (Previous tutorial)
2. Psychometrics (Next tutorial)
3. Instructor notes
4. Correlation
5. Correlation (Lecture)
6. Outliers and restricted range (Advanced correlation tutorial exercise)

External links

• Linear correlation coefficient (Tutorial; stattrek.com)
• Scatterplots: The correlation coefficient, r (University of Wollongong)