Please check this website periodically for announcements regarding PSYCH 710.

"Without replication, all results should be taken as preliminary." -- Gary Marcus, Cleaning Up Science, The New Yorker News Desk, December 24, 2012

• Review of basic aspects of descriptive and inferential statistics: PSYCH 710 assumes that students have taken an undergraduate course in inferential statistics. Some of the important concepts taught in such a course are reviewed in the following documents.
  • Maxwell & Delaney's basic statistics tutorial. This file also can be found on the CD that came with the textbook.
  • Using p-values to test a hypothesis by Daniel Lakens.
  • PJB's notes on z and t-tests [updated 10:55AM, September 9, 2014]
  • A brief handout describing how to extract subsets of data from R variables can be found here and here.

• The following items review important issues regarding significance testing:
  • The ASA's Statement on p-Values
  • Goodman, S (2008). A Dirty Dozen: Twelve P-Value Misconceptions, Seminars in Hematology, 45(3), 135-40.
  • Cohen, J (1994). The earth is round (p<.05). American Psychologist, 49(12), pp. 997-1003.
  • Lykken, D.T. (1968). Statistical significance in psychological research. Psychological Bulletin, 70(3), pp. 151-159.
  • Gelman, A (2013). P values and statistical practice. Epidemiology, 24(1), 69-72.
  • Loftus, G. (1996). Psychology will be a much better science when we change the way we analyze data. Current Directions in Psychological Science, 5, 161-71.
  • Krantz, D (1999). The null hypothesis testing controversy in Psychology. J. American Statistical Association, 44(448), 1372-81.
  • Significant
  • Please don't ever try to boost your results.

• R links
  • The Comprehensive R Archive Network (CRAN)
  • Introductory Tutorials Covering Miscellaneous Topics in R
  • R For Beginners
  • R for Data Science
  • Some Free R Manuals, Tutorials, etc.
  • Jonathan Baron's R reference card
  • tidyverse: A collection of several useful R packages that share common philosophies and are designed to work together.
  • R bloggers
  • Quick-R
  • NY Times article discussing why R is taking over the world
  • Edinburgh Psychology R-users
  • A blog post illustrating how to do a paired-sample t test is here.
  • R Tutorial:
    • Learn how to do a chi-squared test of independence here.
    • A comparison of two proportions.
    • t tests for paired and independent samples.
    • Easy alternatives to bar charts in native R graphics.

• Interesting websites that deal with statistical issues
  • The following items review important issues regarding replication:
    • Five ways to fix statistics.
    • Cleaning Up Science [by Gary Marcus]
    • Reproducibility Project (Psychology)
    • How Reliable Are Psychological Studies (The Atlantic, Aug 27, 2015)
    • The Replicability of Cognitive Psychology in the OSF-Reproducibility-Project
    • The Bayesian Reproducibility Project
    • Mini Meta-Analysis of Your Own Studies
  • The 20% Statistician [Daniel Lakens blog]
  • Statistical Modeling, Causal Inference, and Social Science [An interesting and sometimes-entertaining blog about statistics]
  • Flowing Data [Data Visualization]
  • Decision Science News
  • The Hardest Science
  • Peg's Blog [Blogs on on psychology, psychometrics, and statistics that help me to remember what I've read and done and think]

Course Notes
Much of the following material is based on the 2nd and 3rd editions of Designing Experiments and Analyzing Data, by Maxwell and Delaney. The book can can be purchased here.

• Review of descriptive statistics: [Notes (PDF), Rscript]
• Review of inferential statistics (t tests; equivalence tests): [Notes (PDF), Rscript]
  • I created two R scripts that illustrate the central limit theorem. The scripts show the distribution of sample means when the scores are drawn from a uniform distribution and a log-normal distribution. To use the scripts you must download them to your computer and then execute them using R's source command. More detailed instructions, as well as a description of what you will see when the script executes, are provided in the script files (which can be read with any text editor or word processor). If you do open/edit the files with a word processor, make sure to save them as ascii text files.
  • R tutorial for conducting t tests for paired and independent samples.
  • This on-line chapter by Daniel Lakens discusses philosophical and methodological issues concerning the use of p values to test hypotheses.
  • This R script demonstrates some coverage properties of 2-tailed and 1-tailed confidence intervals.
  • Fact: p values are distributed uniformly when the null hypothesis is true. [R script demo]
• >One-way Between-Subjects ANOVA [Rscript]
  
• Linear Contrasts [Rscript]
  
  • ANOVA + Contrasts in R [A blog post from U of Oregon's R Club]
• Multiple Comparisons of Means [Rscript]
  
  • I discuss an issue with the way the Scheffe test is done in the R package, emmeans, here
  • Illustrations of potential problems associated with doing multiple comparisons can be found here, here, and here.
  • Why you don't need to adjust your alpha level for all tests you'll do in your lifetime. The 20% Statistician, 14-Feb-2016.
• Trend Analyses [Rscript] [updated 25-Sep-2021 11:33 AM]
  
  • A brief comparison of trend analysis and regreression is presented here.
• Between-Subjects Factorial Designs [Rscript] [updated 09-Oct-2021 09:37 PM]
  [Added Sections 7.8.1, 7.10.1, and 7.11.2.]
  • Here is a paper describing common erroneous analyses of interactions.
  • Here is an illustration of a problem that can occur with a linear model that violates marginality.
  • This handout shows how factorial ANOVA can be thought of as a set of orthogonal contrasts performed on a one-way design.
  • Some useful R commands for summarizing data from factorial designs.
  • This handout gives an example of analyzing 3-way interactions.
• Designs with Covariates [Rscript]
  
• Designs With Random or Nested Factors [Rscript]
  [Fixed minor typesetting errors.]
  • Examples of random and mixed model analyses
• (One-way Within-Subjects Designs: Univariate Approach [Rscript]
  
  • Examples of mixed-model analyses. [HTML]
• Higher-Order Designs with Within-Subjects Factors [Rscript]