Statistics

[also see miscellaneous references]

GENERAL TOOLS

jamovi

• • Also see Courses

JASP

• • jamovi versus JASP: Part 1 and Part 2.

R Scripts

R Tutorial

Web Interfaces for Quick, Simple R Coding

• • TutorialsPoint

  • OneCompiler (This is nice, but has become inconvenient to use unless you create an account, which is free.)

SPECIAL-PURPOSE STATISTICAL CALCULATORS

Statistical Power

• https://webpower.psychstat.org/wiki/models/index

  • • Effect Size: Conversion

    • Effect Size: Classification (rules of thumb for "small," "medium," and "large")

    • An argument against using classification: Correll, J., Mellinger, C., McClelland, G. H., & Judd, C. M. (2020). Avoid Cohen’s ‘small’, ‘medium’, and ‘large’ for power analysis. Trends in Cognitive Sciences, 24(3), 200–207. https://doi-org.ezproxy.bgsu.edu/10.1016/j.tics.2019.12.009

• Statistical Power for Chi Square and Proportion Tests

• G*Power3 (Assists in calculating a priori and other forms of statistical power.)

Test of Whether Two Independent Samples have Identical Distributions

• The Kolmogorov-Smirnov Two-Sample Test

Chi Square

• One-Variable Chi Square test (also called a Chi Square goodness-of-fit test; for when there is just one row of observed frequencies).

  • • • Another, Spreadsheet-Based, One-Variable Chi Square Test

• Chi Square contingency test.

• McNemar's Change Test (site 1)

• McNemar's Change Test (site 2)

Combinations and Permutations Calculator. Calculates the number of possible combinations or permutations (does not conduct statistical tests). Conceptually useful, though crash prone on some computers.

Binomial Probability Calculator

Correlation Calculation

• Comparing Two Correlations

  • • Correlation Comparison Calculator at Vasser Stats

    • More Variations: Correlations and Correlation Comparisons.

    • Notes on comparing two correlations (includes some web links)

    • Comparison of tests of equality of dependent correlation coefficients

      • • Myers, L, & Siriois, M. J. (2006). Spearman correlation coefficients, Differences between. In Encyclopedia of statistical Sciences. In S. Kotz, S., C. B. Read, N. Balakrishnan, & B. Vidakovic (eds.), Encyclopedia of statistical sciences. New York: Wiley. doi: 10.1002/0471667196.ess5050.pub2

          • Spearman correlation coefficients, Differences between

    • https://www.ibm.com/support/pages/differences-between-correlations

• Confidence interval for Pearson correlation coefficient (video and calculator link)

• Estimating Pearson's correlation coefficient with a bootstrap

• Finding the least significant value of the Pearson Correlation Coefficient

Signal Detection

• ROC Curve Generator

• Also see Calculation of Signal Detection Theory Measures

Miscellaneous Calculators:

• APerm II (A newer, more flexible, but slower permutation tester. Requires Office 2010 or later.)

• Anderson's Permutation Tester (Windows Only. Requires Microsoft Office 2010.)

• A Brief Overview of the Logic of Permutation Testing

Miscellaneous TUTORIALS

• An Online Introductory Statistics Textbook . . .

  • • See the "Table of Contents" at http://vassarstats.net/textbook/

• Another, more advanced Online Statistics Textbook

• Multiple Regression: Multi-Part Video Tutorial

• Hierarchical Curvilinear Regression

• Applied Regression Modeling. Iain Pardoe. Published 2013:

Full online textbook on multiple Regression (BGSU login required)

• Bayes Theorem Tutorial

• The "New Statistics"

• • • • New Statistical Rules for APS Journals

      • APS Video Presentations

  • Main effects and interactions: Demo for practice

  • Missing Data

    • https://www.tandfonline.com/doi/full/10.1080/00223891.2018.1530680

    • http://www.stat.columbia.edu/~gelman/arm/missing.pdf

MISCELLANEOUS TIPS

• jamovi and R

Logical Operators

equal? ==

not equal? !=

less than? <

greater than? >

are both true? and {jamovi} & {R}

is at least one true? or {jamovi} | {R}

is it false? not {jamovi} ! {R}

• R: To create a vector (i.e., an array) of shuffled consecutive integers, where n is the number of consecutive integers, and m is the number of elements to in the new vector (and is equal to n), use: mynewvector <- sample.int (n, size = m, replace = FALSE, prob = NULL)

• GGPLOT2/R: To horizontally offset elements that have already been dodged:

aes(x = as.numeric(myXvar) + 0.2)

• Verbal Bayes-factor categories from Kass & Raftery (1995) [also see the "footnotes" of the present web page]:

______________________________________________________________________________

Evidence Against the Null Hypothesis

BF > 1 through BF <= 3 "Not worth more than a bare mention."

BF > 3 through BF <= 20 "Positive"

BF > 20 through BF <= 150 "Strong"

BF > 150 "Very Strong"

______________________________________________________________________________

• Data Trimming Advice

• Excel: Creating random normal deviates: =NORM.INV(RAND()+0.0000000001,0,1)

Footnotes

A snippet from Kass and Raftery: Robert E Kass, R. E., & Raftery, A. E. (1995). Bayes Factors. Journal of the American Statistical Association, 90, 774-795.