This site contains article references, online tutorials/calculators, and other sources that people have sent to me or that I've looked up to help students and colleagues. Rather than letting these resources sit in my e-mail folders and web favorites list, I thought it would be good to share. Some basic knowledge of statistics is assumed. My aim, therefore, is to help people branch out from their current statistical repertoire and/or trouble-shoot common roadblocks with various techniques. If you'd like to suggest other links and sources, please e-mail me via my faculty webpage.
LAST UPDATED: February 24, 2014
General Tools, Multiple Techniques |
Stat Pages (compilation
of hundreds of online statistical calculators, from John C. Pezzullo) |
Selecting statistical techniques based on number and types of variables |
Interactive Decision Tree Article "Choosing Statistical Tests" UCLA -- Tabular form, with advice specific to different software programs Pepperdine -- Tabular form Links to additional tables |
Data Management |
Data Screening/Cleaning Checklist
Skewness and kurtosis (note that skewness and kurtosis should
each be divided by its respective standard error to see if the traditional cut-off for
two-tailed .05 significance on the normal curve, +/- 1.96, is attained
[for simplicity, a ratio of 2 is often cited, such as
here and
here]. Some authors
recommend larger [absolute-value] cut-offs).
Guidelines for transforming variables to improve distributional
qualities
Index (composite variable) creation (SPSS; discusses situation of participants missing responses to some of the variables comprising the index) Missing Data -----Acock, A. (2005). Working with missing values. Journal of Marriage and Family, 67, -----UCLA (how missing data handled in SPSS) -----Missing data in SEM (with a focus on AMOS) Regression diagnostics (see also Outliers under Trouble-Shooting, below) Reverse scoring of variables -- On measures with a strongly disagree-strongly agree (Likert) format, where one or more items have an oppositely toned wording to the majority of items (e.g., "I dislike Restaurant A," where the other items are "I like Restaurant A," "I plan to keep going back to Restaurant A," "The food is great at Restaurant A," etc.), the scoring of the oppositely toned item(s) should be reversed before the items are combined into an overall scale. This first document explains how to use the compute command in SPSS, so that for example, on a 1-7 scale, taking 8 minus the original value converts a 1 into a 7, a 2 into a 6, etc. This second document shows how to reverse the relevant items manually through the SPSS recode technique (i.e., literally telling the computer to convert the 1 into a 7, the 2 into a 6, etc.). I always recommend giving the new variable a new name (e.g., "item4r" where "r" stands for "reverse"), so that the original variable "item4" remains unperturbed in case you need to go back to it for any reason. |
Overview Intros |
ANOVA:
one-way Canonical Correlation (here and here) Cluster Analysis ......Rapkin, B. D., & Luke, D. A. (1993). Cluster analysis in community research... American J. of Community Psychol., 21, 247-277. Correlation
analogues (e.g., biserial, point-biserial, phi)
-----40,
532-538.
Latent Class Analysis (here)
Mediation/Moderation (Kristopher
Preacher's
resources; Andrew Hayes's
resources and article "Beyond
Baron and Kenny") |
Freeware |
Meta-analysis (only some of the listed programs are free) |
Trouble- Shooting, Difficulties |
General: Cortina, J. M. (2002). Big things have small beginnings: An assortment of “minor” methodological misunderstandings. Journal of Management, 28, 339–362 (abstract; thanks to Austin Houghtaling). Reviews scenarios such as outliers; standardized regression Betas greater than 1; negative variances (Heywood Cases) and correlated errors in SEM; and more! Chi-square: Which cell(s) are the major contributors to an overall significant chi-square? (standardized residuals) Comparing correlations: There are different types of correlation comparisons. Do you want to compare correlations of the same variables (A & B) in two independent samples? Or, within the same sample, is the correlation of A & B different from that between A & C (dependent correlations)? This document explains the different types of comparisons and links to an online calculator (where it says "Go to procedure"). For more specialized issues, such as comparing partial correlations, see: ..........Preacher, K. J. (2006). Testing complex correlational hypotheses using structural equation modeling. Structural Equation Modeling, 13, 520-543.
Effect sizes with repeated
measures:
Dunlap,
W. P., Cortina, J. M., Vaslow, J. B., & Burke, M. J. (1996).
Meta-analysis of experiments with matched groups or repeated measures designs.
Psychological Methods, 1,
170-177. Interactions in multiple regression and related techniques: Resources from Preacher, Curran, and Bauer Moderation and Mediation: Resources from Kristopher Preacher Multicollinarity (here, here, here, and here) Outliers (suggestions from Peter Westfall) Presenting Data: "Just Plain Data Analysis" Spuriousness and Suppressor Effects/Suppressor Variables (here; suppression refers, for example, to when a relationship between variables switches from positive to negative, with the introduction of control variables) Syntax
(SPSS) |
Terminology |
Compendium of terms (each place it says "Category" opens to a set of
more specific links) Odds vs. probability (pertinent to logistic regression) |
Tips for Constructing Tables and Figures to Present Data |
Tables for multiple regression (Alan Acock) Lane, D. M. & Sandor, A. (2009) Designing better graphs by including distributional information and integrating words, numbers, and images. Psychological Methods, 14, 239-257. |
Power Analysis for Advanced Techniques |
Power for multiple
techniques (here) Chris Aberson's power-analysis resources ("downloadable resources include SPSS syntax for completing power analysis for a ton of designs. These do come from my text on power analysis but the files can be used without purchase of the book") Hertzog, C., Lindenberger, U., Ghisletta, P., & von Oertzen, T. (2006). On the power of multivariate latent growth curve models to detect correlated change. Psychological Methods, 11, 244-252. Preacher, K. J., & Coffman, D. L. (2006, May). Computing power and minimum sample size for RMSEA [Computer software]. |
Other Resource Pages | Online journals
offering free tutorial articles:
Practical
Assessment, Research, and Evaluation and
Frontiers in Quantitative Psychology and Measurement (more
advanced) Carolyn Anderson's syllabus/lecture notes for Applied Categorical Data Analysis and Multilevel Analysis/HLM" (lots of examples!) James Grice's Personality Research Lab resource page (discussion of advanced topics in methods and statistics) Kevin Grimm's computer scripts for advanced longitudinal analyses
Winfred Arthur's
vita (on conducting meta-analysis with different software
programs; see book [2001] and articles from 1990s)
Statistics Resources for Businesses and Educators (Thanks to
Stacy Kozak)
Dr. Reifman's Introductory Graduate Statistics Dr. Reifman's Structural Equation Modeling |
Lines of Argument
for Why Small Effect Sizes Can Still Be Meaningful |
1. Importance (i.e., life and death) of the outcome variable. |
Secondary Data Analysis | U. of Michigan (ICPSR)
archive of datasets Andersen, J., Prause, J., & Silver, R. C. (2011). A step-by-step guide to using secondary data for psychological research. Social and Personality Psychology Compass, 5, 56-75 (abstract). Trzesniewsk, K. , Donnellan, M. B., & Lucas, R. E. (Eds.) (2010). Secondary data analysis: An introduction for psychologists. Washington, DC: American Psychological Association (publisher page). |
Dr. Reifman's Other Website Compilations |
Summer statistics/methodology workshops around the world Questionnaire instruments (personality traits, social behavior) in the public domain |