├── .Rbuildignore ├── .github └── .gitignore ├── .gitignore ├── DESCRIPTION ├── LICENSE.md ├── NAMESPACE ├── R └── exercises.R ├── README.Rmd ├── README.md ├── _pkgdown.yml ├── classes ├── 2021_grenoble │ ├── .gitignore │ ├── data │ │ └── sim-data-ind-samples.csv │ ├── faux.Rmd │ ├── fixed-effects.Rmd │ ├── mixed-effects.Rmd │ └── sim1.csv └── mixed2.Rmd ├── data-sim-workshops.Rproj ├── docs ├── 404.html ├── LICENSE.html ├── apple-touch-icon-120x120.png ├── apple-touch-icon-152x152.png ├── apple-touch-icon-180x180.png ├── apple-touch-icon-60x60.png ├── apple-touch-icon-76x76.png ├── apple-touch-icon.png ├── articles │ ├── calories.html │ ├── faux.html │ ├── faux_files │ │ └── figure-html │ │ │ ├── unnamed-chunk-13-1.png │ │ │ ├── unnamed-chunk-14-1.png │ │ │ ├── unnamed-chunk-15-1.png │ │ │ ├── unnamed-chunk-16-1.png │ │ │ ├── unnamed-chunk-17-1.png │ │ │ ├── unnamed-chunk-20-1.png │ │ │ ├── unnamed-chunk-21-1.png │ │ │ ├── unnamed-chunk-22-1.png │ │ │ ├── unnamed-chunk-23-1.png │ │ │ ├── unnamed-chunk-24-1.png │ │ │ ├── unnamed-chunk-28-1.png │ │ │ ├── unnamed-chunk-37-1.png │ │ │ ├── unnamed-chunk-4-1.png │ │ │ ├── unnamed-chunk-40-1.png │ │ │ ├── unnamed-chunk-41-1.png │ │ │ ├── unnamed-chunk-42-1.png │ │ │ ├── unnamed-chunk-43-1.png │ │ │ ├── unnamed-chunk-44-1.png │ │ │ └── unnamed-chunk-45-1.png │ ├── fixed.html │ ├── fixed_files │ │ └── figure-html │ │ │ ├── ind-sim-fig-1.png │ │ │ ├── pair-sim-fig-1.png │ │ │ ├── rnorm-plot-1.png │ │ │ ├── runif-hist-1.png │ │ │ ├── sample-prob-1.png │ │ │ ├── sample-replace-1.png │ │ │ └── sim-p-fig-1.png │ ├── index.html │ ├── mixed.html │ └── mixed_files │ │ └── figure-html │ │ ├── ex2-1.png │ │ ├── plot-dv-1.png │ │ ├── plot-ixn-1.png │ │ ├── plot-stim-ranef-1.png │ │ ├── plot-sub-ranef-1.png │ │ ├── rslope-plot-dv-1.png │ │ ├── unnamed-chunk-11-1.png │ │ └── unnamed-chunk-7-1.png ├── authors.html ├── deps │ ├── bootstrap-5.3.1 │ │ ├── bootstrap.bundle.min.js │ │ ├── bootstrap.bundle.min.js.map │ │ └── bootstrap.min.css │ ├── data-deps.txt │ └── jquery-3.6.0 │ │ ├── jquery-3.6.0.js │ │ ├── jquery-3.6.0.min.js │ │ └── jquery-3.6.0.min.map ├── favicon-16x16.png ├── favicon-32x32.png ├── favicon.ico ├── index.html ├── link.svg ├── logo.png ├── pkgdown.js ├── pkgdown.yml ├── reference │ ├── Rplot001.png │ ├── exercise.html │ ├── figures │ │ └── logo.png │ └── index.html ├── search.json └── sitemap.xml ├── inst └── stubs │ ├── calories-stub.Rmd │ ├── faux-stub.Rmd │ ├── fixed-stub.Rmd │ └── mixed-stub.Rmd ├── man ├── exercise.Rd └── figures │ └── logo.png ├── pkgdown └── favicon │ ├── apple-touch-icon-120x120.png │ ├── apple-touch-icon-152x152.png │ ├── apple-touch-icon-180x180.png │ ├── apple-touch-icon-60x60.png │ ├── apple-touch-icon-76x76.png │ ├── apple-touch-icon.png │ ├── favicon-16x16.png │ ├── favicon-32x32.png │ └── favicon.ico └── vignettes ├── .gitignore ├── calories.Rmd ├── data └── sim-data-ind-samples.csv ├── faux.Rmd ├── fixed.Rmd └── mixed.Rmd /.Rbuildignore: -------------------------------------------------------------------------------- 1 | ^LICENSE\.md$ 2 | ^.*\.Rproj$ 3 | ^\.Rproj\.user$ 4 | ^_pkgdown\.yml$ 5 | ^docs$ 6 | ^pkgdown$ 7 | ^classes$ 8 | ^README.Rmd$ 9 | ^\.github$ 10 | -------------------------------------------------------------------------------- /.github/.gitignore: -------------------------------------------------------------------------------- 1 | *.html 2 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | .Rproj.user 2 | .Rhistory 3 | .RData 4 | .Ruserdata 5 | inst/doc 6 | -------------------------------------------------------------------------------- /DESCRIPTION: -------------------------------------------------------------------------------- 1 | Package: dsw 2 | Title: Data Simulation Workshops 3 | Version: 0.0.0.9007 4 | Date: 2024-02-16 5 | Authors@R: c( 6 | person( 7 | given = "Lisa", 8 | family = "DeBruine", 9 | role = c("aut", "cre"), 10 | email = "debruine@gmail.com", 11 | comment = c(ORCID = "0000-0002-7523-5539") 12 | )) 13 | Description: Materials for data simulation workshops. 14 | Depends: 15 | R (>= 4.1.0) 16 | Imports: 17 | lme4, 18 | dplyr, 19 | tidyr, 20 | ggplot2 (>= 3.3.0), 21 | faux (>= 1.1.0), 22 | lmerTest, 23 | afex, 24 | broom, 25 | broom.mixed, 26 | MASS, 27 | emmeans, 28 | patchwork 29 | Suggests: 30 | testthat (>= 2.1.0), 31 | knitr, 32 | rmarkdown 33 | RoxygenNote: 7.1.1 34 | Encoding: UTF-8 35 | LazyData: true 36 | URL: https://github.com/debruine/data-sim-workshops, 37 | https://debruine.github.io/data-sim-workshops/ 38 | BugReports: https://github.com/debruine/data-sim-workshops/issues 39 | License: CC BY 4.0 40 | VignetteBuilder: knitr 41 | -------------------------------------------------------------------------------- /LICENSE.md: -------------------------------------------------------------------------------- 1 | Attribution 4.0 International 2 | 3 | ======================================================================= 4 | 5 | Creative Commons Corporation ("Creative Commons") is not a law firm and 6 | does not provide legal services or legal advice. Distribution of 7 | Creative Commons public licenses does not create a lawyer-client or 8 | other relationship. Creative Commons makes its licenses and related 9 | information available on an "as-is" basis. Creative Commons gives no 10 | warranties regarding its licenses, any material licensed under their 11 | terms and conditions, or any related information. Creative Commons 12 | disclaims all liability for damages resulting from their use to the 13 | fullest extent possible. 14 | 15 | Using Creative Commons Public Licenses 16 | 17 | Creative Commons public licenses provide a standard set of terms and 18 | conditions that creators and other rights holders may use to share 19 | original works of authorship and other material subject to copyright 20 | and certain other rights specified in the public license below. The 21 | following considerations are for informational purposes only, are not 22 | exhaustive, and do not form part of our licenses. 23 | 24 | Considerations for licensors: Our public licenses are 25 | intended for use by those authorized to give the public 26 | permission to use material in ways otherwise restricted by 27 | copyright and certain other rights. Our licenses are 28 | irrevocable. Licensors should read and understand the terms 29 | and conditions of the license they choose before applying it. 30 | Licensors should also secure all rights necessary before 31 | applying our licenses so that the public can reuse the 32 | material as expected. Licensors should clearly mark any 33 | material not subject to the license. This includes other CC- 34 | licensed material, or material used under an exception or 35 | limitation to copyright. More considerations for licensors: 36 | wiki.creativecommons.org/Considerations_for_licensors 37 | 38 | Considerations for the public: By using one of our public 39 | licenses, a licensor grants the public permission to use the 40 | licensed material under specified terms and conditions. If 41 | the licensor's permission is not necessary for any reason--for 42 | example, because of any applicable exception or limitation to 43 | copyright--then that use is not regulated by the license. Our 44 | licenses grant only permissions under copyright and certain 45 | other rights that a licensor has authority to grant. Use of 46 | the licensed material may still be restricted for other 47 | reasons, including because others have copyright or other 48 | rights in the material. A licensor may make special requests, 49 | such as asking that all changes be marked or described. 50 | Although not required by our licenses, you are encouraged to 51 | respect those requests where reasonable. More considerations 52 | for the public: 53 | wiki.creativecommons.org/Considerations_for_licensees 54 | 55 | ======================================================================= 56 | 57 | Creative Commons Attribution 4.0 International Public License 58 | 59 | By exercising the Licensed Rights (defined below), You accept and agree 60 | to be bound by the terms and conditions of this Creative Commons 61 | Attribution 4.0 International Public License ("Public License"). To the 62 | extent this Public License may be interpreted as a contract, You are 63 | granted the Licensed Rights in consideration of Your acceptance of 64 | these terms and conditions, and the Licensor grants You such rights in 65 | consideration of benefits the Licensor receives from making the 66 | Licensed Material available under these terms and conditions. 67 | 68 | 69 | Section 1 -- Definitions. 70 | 71 | a. Adapted Material means material subject to Copyright and Similar 72 | Rights that is derived from or based upon the Licensed Material 73 | and in which the Licensed Material is translated, altered, 74 | arranged, transformed, or otherwise modified in a manner requiring 75 | permission under the Copyright and Similar Rights held by the 76 | Licensor. For purposes of this Public License, where the Licensed 77 | Material is a musical work, performance, or sound recording, 78 | Adapted Material is always produced where the Licensed Material is 79 | synched in timed relation with a moving image. 80 | 81 | b. Adapter's License means the license You apply to Your Copyright 82 | and Similar Rights in Your contributions to Adapted Material in 83 | accordance with the terms and conditions of this Public License. 84 | 85 | c. Copyright and Similar Rights means copyright and/or similar rights 86 | closely related to copyright including, without limitation, 87 | performance, broadcast, sound recording, and Sui Generis Database 88 | Rights, without regard to how the rights are labeled or 89 | categorized. For purposes of this Public License, the rights 90 | specified in Section 2(b)(1)-(2) are not Copyright and Similar 91 | Rights. 92 | 93 | d. Effective Technological Measures means those measures that, in the 94 | absence of proper authority, may not be circumvented under laws 95 | fulfilling obligations under Article 11 of the WIPO Copyright 96 | Treaty adopted on December 20, 1996, and/or similar international 97 | agreements. 98 | 99 | e. Exceptions and Limitations means fair use, fair dealing, and/or 100 | any other exception or limitation to Copyright and Similar Rights 101 | that applies to Your use of the Licensed Material. 102 | 103 | f. Licensed Material means the artistic or literary work, database, 104 | or other material to which the Licensor applied this Public 105 | License. 106 | 107 | g. Licensed Rights means the rights granted to You subject to the 108 | terms and conditions of this Public License, which are limited to 109 | all Copyright and Similar Rights that apply to Your use of the 110 | Licensed Material and that the Licensor has authority to license. 111 | 112 | h. Licensor means the individual(s) or entity(ies) granting rights 113 | under this Public License. 114 | 115 | i. Share means to provide material to the public by any means or 116 | process that requires permission under the Licensed Rights, such 117 | as reproduction, public display, public performance, distribution, 118 | dissemination, communication, or importation, and to make material 119 | available to the public including in ways that members of the 120 | public may access the material from a place and at a time 121 | individually chosen by them. 122 | 123 | j. Sui Generis Database Rights means rights other than copyright 124 | resulting from Directive 96/9/EC of the European Parliament and of 125 | the Council of 11 March 1996 on the legal protection of databases, 126 | as amended and/or succeeded, as well as other essentially 127 | equivalent rights anywhere in the world. 128 | 129 | k. You means the individual or entity exercising the Licensed Rights 130 | under this Public License. Your has a corresponding meaning. 131 | 132 | 133 | Section 2 -- Scope. 134 | 135 | a. License grant. 136 | 137 | 1. Subject to the terms and conditions of this Public License, 138 | the Licensor hereby grants You a worldwide, royalty-free, 139 | non-sublicensable, non-exclusive, irrevocable license to 140 | exercise the Licensed Rights in the Licensed Material to: 141 | 142 | a. reproduce and Share the Licensed Material, in whole or 143 | in part; and 144 | 145 | b. produce, reproduce, and Share Adapted Material. 146 | 147 | 2. Exceptions and Limitations. For the avoidance of doubt, where 148 | Exceptions and Limitations apply to Your use, this Public 149 | License does not apply, and You do not need to comply with 150 | its terms and conditions. 151 | 152 | 3. Term. The term of this Public License is specified in Section 153 | 6(a). 154 | 155 | 4. Media and formats; technical modifications allowed. The 156 | Licensor authorizes You to exercise the Licensed Rights in 157 | all media and formats whether now known or hereafter created, 158 | and to make technical modifications necessary to do so. The 159 | Licensor waives and/or agrees not to assert any right or 160 | authority to forbid You from making technical modifications 161 | necessary to exercise the Licensed Rights, including 162 | technical modifications necessary to circumvent Effective 163 | Technological Measures. For purposes of this Public License, 164 | simply making modifications authorized by this Section 2(a) 165 | (4) never produces Adapted Material. 166 | 167 | 5. Downstream recipients. 168 | 169 | a. Offer from the Licensor -- Licensed Material. Every 170 | recipient of the Licensed Material automatically 171 | receives an offer from the Licensor to exercise the 172 | Licensed Rights under the terms and conditions of this 173 | Public License. 174 | 175 | b. No downstream restrictions. You may not offer or impose 176 | any additional or different terms or conditions on, or 177 | apply any Effective Technological Measures to, the 178 | Licensed Material if doing so restricts exercise of the 179 | Licensed Rights by any recipient of the Licensed 180 | Material. 181 | 182 | 6. No endorsement. Nothing in this Public License constitutes or 183 | may be construed as permission to assert or imply that You 184 | are, or that Your use of the Licensed Material is, connected 185 | with, or sponsored, endorsed, or granted official status by, 186 | the Licensor or others designated to receive attribution as 187 | provided in Section 3(a)(1)(A)(i). 188 | 189 | b. Other rights. 190 | 191 | 1. Moral rights, such as the right of integrity, are not 192 | licensed under this Public License, nor are publicity, 193 | privacy, and/or other similar personality rights; however, to 194 | the extent possible, the Licensor waives and/or agrees not to 195 | assert any such rights held by the Licensor to the limited 196 | extent necessary to allow You to exercise the Licensed 197 | Rights, but not otherwise. 198 | 199 | 2. Patent and trademark rights are not licensed under this 200 | Public License. 201 | 202 | 3. To the extent possible, the Licensor waives any right to 203 | collect royalties from You for the exercise of the Licensed 204 | Rights, whether directly or through a collecting society 205 | under any voluntary or waivable statutory or compulsory 206 | licensing scheme. In all other cases the Licensor expressly 207 | reserves any right to collect such royalties. 208 | 209 | 210 | Section 3 -- License Conditions. 211 | 212 | Your exercise of the Licensed Rights is expressly made subject to the 213 | following conditions. 214 | 215 | a. Attribution. 216 | 217 | 1. If You Share the Licensed Material (including in modified 218 | form), You must: 219 | 220 | a. retain the following if it is supplied by the Licensor 221 | with the Licensed Material: 222 | 223 | i. identification of the creator(s) of the Licensed 224 | Material and any others designated to receive 225 | attribution, in any reasonable manner requested by 226 | the Licensor (including by pseudonym if 227 | designated); 228 | 229 | ii. a copyright notice; 230 | 231 | iii. a notice that refers to this Public License; 232 | 233 | iv. a notice that refers to the disclaimer of 234 | warranties; 235 | 236 | v. a URI or hyperlink to the Licensed Material to the 237 | extent reasonably practicable; 238 | 239 | b. indicate if You modified the Licensed Material and 240 | retain an indication of any previous modifications; and 241 | 242 | c. indicate the Licensed Material is licensed under this 243 | Public License, and include the text of, or the URI or 244 | hyperlink to, this Public License. 245 | 246 | 2. You may satisfy the conditions in Section 3(a)(1) in any 247 | reasonable manner based on the medium, means, and context in 248 | which You Share the Licensed Material. For example, it may be 249 | reasonable to satisfy the conditions by providing a URI or 250 | hyperlink to a resource that includes the required 251 | information. 252 | 253 | 3. If requested by the Licensor, You must remove any of the 254 | information required by Section 3(a)(1)(A) to the extent 255 | reasonably practicable. 256 | 257 | 4. If You Share Adapted Material You produce, the Adapter's 258 | License You apply must not prevent recipients of the Adapted 259 | Material from complying with this Public License. 260 | 261 | 262 | Section 4 -- Sui Generis Database Rights. 263 | 264 | Where the Licensed Rights include Sui Generis Database Rights that 265 | apply to Your use of the Licensed Material: 266 | 267 | a. for the avoidance of doubt, Section 2(a)(1) grants You the right 268 | to extract, reuse, reproduce, and Share all or a substantial 269 | portion of the contents of the database; 270 | 271 | b. if You include all or a substantial portion of the database 272 | contents in a database in which You have Sui Generis Database 273 | Rights, then the database in which You have Sui Generis Database 274 | Rights (but not its individual contents) is Adapted Material; and 275 | 276 | c. You must comply with the conditions in Section 3(a) if You Share 277 | all or a substantial portion of the contents of the database. 278 | 279 | For the avoidance of doubt, this Section 4 supplements and does not 280 | replace Your obligations under this Public License where the Licensed 281 | Rights include other Copyright and Similar Rights. 282 | 283 | 284 | Section 5 -- Disclaimer of Warranties and Limitation of Liability. 285 | 286 | a. UNLESS OTHERWISE SEPARATELY UNDERTAKEN BY THE LICENSOR, TO THE 287 | EXTENT POSSIBLE, THE LICENSOR OFFERS THE LICENSED MATERIAL AS-IS 288 | AND AS-AVAILABLE, AND MAKES NO REPRESENTATIONS OR WARRANTIES OF 289 | ANY KIND CONCERNING THE LICENSED MATERIAL, WHETHER EXPRESS, 290 | IMPLIED, STATUTORY, OR OTHER. THIS INCLUDES, WITHOUT LIMITATION, 291 | WARRANTIES OF TITLE, MERCHANTABILITY, FITNESS FOR A PARTICULAR 292 | PURPOSE, NON-INFRINGEMENT, ABSENCE OF LATENT OR OTHER DEFECTS, 293 | ACCURACY, OR THE PRESENCE OR ABSENCE OF ERRORS, WHETHER OR NOT 294 | KNOWN OR DISCOVERABLE. WHERE DISCLAIMERS OF WARRANTIES ARE NOT 295 | ALLOWED IN FULL OR IN PART, THIS DISCLAIMER MAY NOT APPLY TO YOU. 296 | 297 | b. TO THE EXTENT POSSIBLE, IN NO EVENT WILL THE LICENSOR BE LIABLE 298 | TO YOU ON ANY LEGAL THEORY (INCLUDING, WITHOUT LIMITATION, 299 | NEGLIGENCE) OR OTHERWISE FOR ANY DIRECT, SPECIAL, INDIRECT, 300 | INCIDENTAL, CONSEQUENTIAL, PUNITIVE, EXEMPLARY, OR OTHER LOSSES, 301 | COSTS, EXPENSES, OR DAMAGES ARISING OUT OF THIS PUBLIC LICENSE OR 302 | USE OF THE LICENSED MATERIAL, EVEN IF THE LICENSOR HAS BEEN 303 | ADVISED OF THE POSSIBILITY OF SUCH LOSSES, COSTS, EXPENSES, OR 304 | DAMAGES. WHERE A LIMITATION OF LIABILITY IS NOT ALLOWED IN FULL OR 305 | IN PART, THIS LIMITATION MAY NOT APPLY TO YOU. 306 | 307 | c. The disclaimer of warranties and limitation of liability provided 308 | above shall be interpreted in a manner that, to the extent 309 | possible, most closely approximates an absolute disclaimer and 310 | waiver of all liability. 311 | 312 | 313 | Section 6 -- Term and Termination. 314 | 315 | a. This Public License applies for the term of the Copyright and 316 | Similar Rights licensed here. However, if You fail to comply with 317 | this Public License, then Your rights under this Public License 318 | terminate automatically. 319 | 320 | b. Where Your right to use the Licensed Material has terminated under 321 | Section 6(a), it reinstates: 322 | 323 | 1. automatically as of the date the violation is cured, provided 324 | it is cured within 30 days of Your discovery of the 325 | violation; or 326 | 327 | 2. upon express reinstatement by the Licensor. 328 | 329 | For the avoidance of doubt, this Section 6(b) does not affect any 330 | right the Licensor may have to seek remedies for Your violations 331 | of this Public License. 332 | 333 | c. For the avoidance of doubt, the Licensor may also offer the 334 | Licensed Material under separate terms or conditions or stop 335 | distributing the Licensed Material at any time; however, doing so 336 | will not terminate this Public License. 337 | 338 | d. Sections 1, 5, 6, 7, and 8 survive termination of this Public 339 | License. 340 | 341 | 342 | Section 7 -- Other Terms and Conditions. 343 | 344 | a. The Licensor shall not be bound by any additional or different 345 | terms or conditions communicated by You unless expressly agreed. 346 | 347 | b. Any arrangements, understandings, or agreements regarding the 348 | Licensed Material not stated herein are separate from and 349 | independent of the terms and conditions of this Public License. 350 | 351 | 352 | Section 8 -- Interpretation. 353 | 354 | a. For the avoidance of doubt, this Public License does not, and 355 | shall not be interpreted to, reduce, limit, restrict, or impose 356 | conditions on any use of the Licensed Material that could lawfully 357 | be made without permission under this Public License. 358 | 359 | b. To the extent possible, if any provision of this Public License is 360 | deemed unenforceable, it shall be automatically reformed to the 361 | minimum extent necessary to make it enforceable. If the provision 362 | cannot be reformed, it shall be severed from this Public License 363 | without affecting the enforceability of the remaining terms and 364 | conditions. 365 | 366 | c. No term or condition of this Public License will be waived and no 367 | failure to comply consented to unless expressly agreed to by the 368 | Licensor. 369 | 370 | d. Nothing in this Public License constitutes or may be interpreted 371 | as a limitation upon, or waiver of, any privileges and immunities 372 | that apply to the Licensor or You, including from the legal 373 | processes of any jurisdiction or authority. 374 | 375 | 376 | ======================================================================= 377 | 378 | Creative Commons is not a party to its public 379 | licenses. Notwithstanding, Creative Commons may elect to apply one of 380 | its public licenses to material it publishes and in those instances 381 | will be considered the “Licensor.” The text of the Creative Commons 382 | public licenses is dedicated to the public domain under the CC0 Public 383 | Domain Dedication. Except for the limited purpose of indicating that 384 | material is shared under a Creative Commons public license or as 385 | otherwise permitted by the Creative Commons policies published at 386 | creativecommons.org/policies, Creative Commons does not authorize the 387 | use of the trademark "Creative Commons" or any other trademark or logo 388 | of Creative Commons without its prior written consent including, 389 | without limitation, in connection with any unauthorized modifications 390 | to any of its public licenses or any other arrangements, 391 | understandings, or agreements concerning use of licensed material. For 392 | the avoidance of doubt, this paragraph does not form part of the 393 | public licenses. 394 | 395 | Creative Commons may be contacted at creativecommons.org. 396 | -------------------------------------------------------------------------------- /NAMESPACE: -------------------------------------------------------------------------------- 1 | # Generated by roxygen2: do not edit by hand 2 | 3 | export(exercise) 4 | -------------------------------------------------------------------------------- /R/exercises.R: -------------------------------------------------------------------------------- 1 | #' Get an exercise 2 | #' 3 | #' @param name The name of the exercise 4 | #' @param filename What filename you want to save (defaults to the name of the exercise in the working directory) 5 | #' 6 | #' @return Saves a file to the working directory (or path from filename) 7 | #' @export 8 | #' 9 | #' @examples 10 | #' \dontrun{ 11 | #' exercise("faux") # get exercise for the faux workshop 12 | #' exercise("fixed", "exercises/fixed.Rmd") # save into exercises directory 13 | #' } 14 | exercise <- function(name = c("faux", "fixed", "mixed", "calories"), filename = NULL) { 15 | fname <- sprintf("stubs/%s-stub.Rmd", match.arg(name)) 16 | f <- system.file(fname, package = "dsw") 17 | 18 | if (f == "") stop("Exercise ", name, " doesn't exist") 19 | 20 | if (is.null(filename)) { 21 | filename <- gsub("^stubs/", "", fname) 22 | } 23 | 24 | file.copy(f, filename) 25 | utils::browseURL(filename) 26 | } -------------------------------------------------------------------------------- /README.Rmd: -------------------------------------------------------------------------------- 1 | --- 2 | output: github_document 3 | --- 4 | 5 | ```{r, include = FALSE} 6 | knitr::opts_chunk$set( 7 | collapse = TRUE, 8 | warning = FALSE, 9 | message = FALSE, 10 | comment = "#>", 11 | fig.path = "man/figures/README-", 12 | out.width = "100%" 13 | ) 14 | set.seed(8675309) 15 | ``` 16 | 17 | # Data Simulation Workshop Materials ![](man/figures/logo.png){style="float:right; width:200px;"} 18 | 19 | Being able to simulate data allows you to: 20 | 21 | * prep analysis scripts for pre-registration 22 | * calculate power and sensitivity for analyses that don't have empirical methods 23 | * create reproducible examples when your data are too big or confidential to share 24 | * enhance your understanding of statistical concepts 25 | * create demo data for teaching and tutorials 26 | 27 | ## Installation 28 | 29 | You can install the packages used in these tutorials and get a function that makes it easy to access the workshop .Rmd files by running the following code: 30 | 31 | ```{r, eval = FALSE} 32 | devtools::install_github("debruine/data-sim-workshops") 33 | ``` 34 | 35 | Then you can load exercises with the following code: 36 | 37 | ```{r, eval = FALSE} 38 | dsw::exercise("faux") 39 | dsw::exercise("calories") 40 | dsw::exercise("fixed") 41 | dsw::exercise("mixed") 42 | ``` 43 | 44 | Alternatively, download the stub files and install the specific packages for your workshop. 45 | 46 | * [faux-stub.Rmd](https://raw.githubusercontent.com/debruine/data-sim-workshops/master/inst/stubs/faux-stub.Rmd) 47 | * [calories-stub.Rmd](https://raw.githubusercontent.com/debruine/data-sim-workshops/master/inst/stubs/calories-stub.Rmd) 48 | * [fixed-stub.Rmd](https://raw.githubusercontent.com/debruine/data-sim-workshops/master/inst/stubs/fixed-stub.Rmd) 49 | * [mixed-stub.Rmd](https://raw.githubusercontent.com/debruine/data-sim-workshops/master/inst/stubs/mixed-stub.Rmd) 50 | 51 | ## Upcoming Workshops 52 | 53 | 93 | 94 | When: 2024 February 1-2 95 | Where: [Data Simulation Workshop 2024](https://kogpsy.github.io/datasimulationcourse_24/), Institute of Psychology, Bern, Switzerland 96 | 97 | ### Data Simulation with {faux} 98 | 99 | This session will cover the basics of simulation using {faux}. We will simulate data with factorial designs by specifying the within and between-subjects factor structure, each cell mean and standard deviation, and correlations between cells where appropriate. This can be used to create simulated data sets to be used in preparing the analysis code for pre-registrations or registered reports. We will also create data sets for simulation-based power analyses. Students will need to have very basic knowledge of R and R Markdown, and have installed {faux}, {afex}, {broom} and {tidyverse}. 100 | 101 | #### Prep 102 | 103 | * Install R packages from CRAN: `tidyverse`, `afex`, `faux`, and `broom` 104 | * Download files: [faux-stub.Rmd](https://raw.githubusercontent.com/debruine/data-sim-workshops/master/inst/stubs/faux-stub.Rmd) 105 | 106 | 107 | ### Data simulation for mixed designs 108 | 109 | This session will cover simulating data for a mixed design, where trials are crossed with subjects. We will learn how to analyse this using {lme4}, with a focus on understanding how the simulation parameters correspond to the output. Finally, we will learn how to use simulation to calculate power. Students will need to have basic knowledge of R and R Markdown, some familiarity with mixed designs (even if they don't currently analyse them with mixed models) and have installed {faux}, {afex}, {tidyverse}, and {lme4}. 110 | 111 | #### Prep 112 | 113 | * Install R packages from CRAN: `tidyverse`, `afex`, `lme4`, `broom`, `broom.mixed`, `faux` 114 | * Download files: [mixed-stub.Rmd](https://raw.githubusercontent.com/debruine/data-sim-workshops/master/inst/stubs/mixed-stub.Rmd) 115 | 116 | 117 | 118 | ## Resources 119 | 120 | * [Faux Shiny App](https://rstudio-connect.psy.gla.ac.uk/faux/) 121 | * [Data Skills for Reproducible Research](https://psyteachr.github.io/reprores/) open source textbook introducing tidyverse for psychologists 122 | * [Understanding mixed effects models through data simulation](https://osf.io/3cz2e/) (preprint, code, and shiny apps) 123 | * [Simulate Basic Distributions](https://rstudio-connect.psy.gla.ac.uk/simulate/) 124 | 125 | ## Past Workshops 126 | 127 | * Vrije Universiteit Amsterdam, NL 128 | Fake It Until You Make It: How and why to simulate research data 129 | 2023 September 20 130 | 131 | * Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany 132 | Simulating data with {faux} 133 | 2023 July 27 9:00 - 12:00 (CET) 134 | 135 | * [European Evolutionary Biology Conference](https://www.empseb28.com/workshops), Millport, Scotland 136 | Fake It Until You Make It: How and why to simulate research data 137 | 2023 June 1 14:30 - 16:30 (GMT) 138 | 139 | * University of Glasgow Institute of Neuroscience & Psychology 140 | Data Simulation with {faux} 141 | 2023 January 18 12:00 - 13:00 (GMT) 142 | 143 | * Netherlands Institute for the Study of Crime and Law Enforcement 144 | Data Simulation with {faux} 145 | 2022 December 6 13:00 - 14:00 (CET) 146 | 147 | * Polish Association of Social Psychology Conference, Gdánsk 148 | Data simulation for fixed effects 149 | Data simulation for mixed designs 150 | Practical Session 151 | 2022 September 14 09:00 - 16:00 (CET) 152 | 153 | * [RLadies Glasgow](https://www.meetup.com/rladies-glasgow/events/285942871/) 154 | Data simulation using faux 155 | 2022 May 24 15:00-17:00 (BST) 156 | 157 | * University of York 158 | Data simulation for factorial designs 159 | Data simulation for mixed designs 160 | 2022 April 27 09:00-17:00 (BST) 161 | 162 | * [From Proposal to Publication: Pathways to Open Science](https://www.dropbox.com/s/aydsuk6eahxumzu/OSW-Jul21.pdf?dl=0) 163 | Data simulation for factorial designs 164 | Data simulation for mixed designs 165 | 2022 July 13 13:30-17:00 166 | 167 | * University of Glasgow 168 | Institute of Neuroscience and Psychology 169 | 2020 Jan 28 13:00-15:00 and Feb 5 14:00-16:00 170 | 171 | * University of Grenoble 172 | Understanding Mixed-Effects Models through Data Simulation 173 | 2021 February 5 13:00-15:00 174 | 175 | * [PsyPAG Data Simulation Summer School](https://simsummerschool.github.io/) 176 | Simulation for factorial designs with faux 177 | 2021 June 4 13:00-15:00 178 | 179 | 180 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | 2 | # Data Simulation Workshop Materials 3 | 4 | Being able to simulate data allows you to: 5 | 6 | - prep analysis scripts for pre-registration 7 | - calculate power and sensitivity for analyses that don’t have empirical 8 | methods 9 | - create reproducible examples when your data are too big or 10 | confidential to share 11 | - enhance your understanding of statistical concepts 12 | - create demo data for teaching and tutorials 13 | 14 | ## Installation 15 | 16 | You can install the packages used in these tutorials and get a function 17 | that makes it easy to access the workshop .Rmd files by running the 18 | following code: 19 | 20 | ``` r 21 | devtools::install_github("debruine/data-sim-workshops") 22 | ``` 23 | 24 | Then you can load exercises with the following code: 25 | 26 | ``` r 27 | dsw::exercise("faux") 28 | dsw::exercise("calories") 29 | dsw::exercise("fixed") 30 | dsw::exercise("mixed") 31 | ``` 32 | 33 | Alternatively, download the stub files and install the specific packages 34 | for your workshop. 35 | 36 | - [faux-stub.Rmd](https://raw.githubusercontent.com/debruine/data-sim-workshops/master/inst/stubs/faux-stub.Rmd) 37 | - [calories-stub.Rmd](https://raw.githubusercontent.com/debruine/data-sim-workshops/master/inst/stubs/calories-stub.Rmd) 38 | - [fixed-stub.Rmd](https://raw.githubusercontent.com/debruine/data-sim-workshops/master/inst/stubs/fixed-stub.Rmd) 39 | - [mixed-stub.Rmd](https://raw.githubusercontent.com/debruine/data-sim-workshops/master/inst/stubs/mixed-stub.Rmd) 40 | 41 | ## Upcoming Workshops 42 | 43 | 71 | 72 | When: 2024 February 1-2 Where: [Data Simulation Workshop 73 | 2024](https://kogpsy.github.io/datasimulationcourse_24/), Institute of 74 | Psychology, Bern , Switzerland 75 | 76 | ### Data Simulation with {faux} 77 | 78 | This session will cover the basics of simulation using {faux}. We will 79 | simulate data with factorial designs by specifying the within and 80 | between-subjects factor structure, each cell mean and standard 81 | deviation, and correlations between cells where appropriate. This can be 82 | used to create simulated data sets to be used in preparing the analysis 83 | code for pre-registrations or registered reports. We will also create 84 | data sets for simulation-based power analyses. Students will need to 85 | have very basic knowledge of R and R Markdown, and have installed 86 | {faux}, {afex}, {broom} and {tidyverse}. 87 | 88 | #### Prep 89 | 90 | - Install R packages from CRAN: `tidyverse`, `afex`, `faux`, and `broom` 91 | - Download files: 92 | [faux-stub.Rmd](https://raw.githubusercontent.com/debruine/data-sim-workshops/master/inst/stubs/faux-stub.Rmd) 93 | 94 | ### Data simulation for mixed designs 95 | 96 | This session will cover simulating data for a mixed design, where trials 97 | are crossed with subjects. We will learn how to analyse this using 98 | {lme4}, with a focus on understanding how the simulation parameters 99 | correspond to the output. Finally, we will learn how to use simulation 100 | to calculate power. Students will need to have basic knowledge of R and 101 | R Markdown, some familiarity with mixed designs (even if they don’t 102 | currently analyse them with mixed models) and have installed {faux}, 103 | {afex}, {tidyverse}, and {lme4}. 104 | 105 | #### Prep 106 | 107 | - Install R packages from CRAN: `tidyverse`, `afex`, `lme4`, `broom`, 108 | `broom.mixed`, `faux` 109 | - Download files: 110 | [mixed-stub.Rmd](https://raw.githubusercontent.com/debruine/data-sim-workshops/master/inst/stubs/mixed-stub.Rmd) 111 | 112 | ## Resources 113 | 114 | - [Faux Shiny App](https://rstudio-connect.psy.gla.ac.uk/faux/) 115 | - [Data Skills for Reproducible 116 | Research](https://psyteachr.github.io/reprores/) open source textbook 117 | introducing tidyverse for psychologists 118 | - [Understanding mixed effects models through data 119 | simulation](https://osf.io/3cz2e/) (preprint, code, and shiny apps) 120 | - [Simulate Basic 121 | Distributions](https://rstudio-connect.psy.gla.ac.uk/simulate/) 122 | 123 | ## Past Workshops 124 | 125 | - Vrije Universiteit Amsterdam, NL Fake It Until You Make It: How and 126 | why to simulate research data 2023 September 20 127 | 128 | - Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany 129 | Simulating data with {faux} 130 | 2023 July 27 9:00 - 12:00 (CET) 131 | 132 | - [European Evolutionary Biology 133 | Conference](https://www.empseb28.com/workshops), Millport, Scotland 134 | Fake It Until You Make It: How and why to simulate research data 135 | 2023 June 1 14:30 - 16:30 (GMT) 136 | 137 | - University of Glasgow Institute of Neuroscience & Psychology 138 | Data Simulation with {faux} 139 | 2023 January 18 12:00 - 13:00 (GMT) 140 | 141 | - Netherlands Institute for the Study of Crime and Law Enforcement 142 | Data Simulation with {faux} 143 | 2022 December 6 13:00 - 14:00 (CET) 144 | 145 | - Polish Association of Social Psychology Conference, Gdánsk 146 | Data simulation for fixed effects 147 | Data simulation for mixed designs 148 | Practical Session 149 | 2022 September 14 09:00 - 16:00 (CET) 150 | 151 | - [RLadies 152 | Glasgow](https://www.meetup.com/rladies-glasgow/events/285942871/) 153 | Data simulation using faux 154 | 2022 May 24 15:00-17:00 (BST) 155 | 156 | - University of York Data simulation for factorial designs 157 | Data simulation for mixed designs 158 | 2022 April 27 09:00-17:00 (BST) 159 | 160 | - [From Proposal to Publication: Pathways to Open 161 | Science](https://www.dropbox.com/s/aydsuk6eahxumzu/OSW-Jul21.pdf?dl=0) 162 | Data simulation for factorial designs 163 | Data simulation for mixed designs 164 | 2022 July 13 13:30-17:00 165 | 166 | - University of Glasgow 167 | Institute of Neuroscience and Psychology 168 | 2020 Jan 28 13:00-15:00 and Feb 5 14:00-16:00 169 | 170 | - University of Grenoble 171 | Understanding Mixed-Effects Models through Data Simulation 172 | 2021 February 5 13:00-15:00 173 | 174 | - [PsyPAG Data Simulation Summer 175 | School](https://simsummerschool.github.io/) 176 | Simulation for factorial designs with faux 177 | 2021 June 4 13:00-15:00 178 | -------------------------------------------------------------------------------- /_pkgdown.yml: -------------------------------------------------------------------------------- 1 | url: https://debruine.github.io/data-sim-workshops/ 2 | template: 3 | bootstrap: 5 4 | 5 | -------------------------------------------------------------------------------- /classes/2021_grenoble/.gitignore: -------------------------------------------------------------------------------- 1 | *.html 2 | *.R 3 | -------------------------------------------------------------------------------- /classes/2021_grenoble/data/sim-data-ind-samples.csv: -------------------------------------------------------------------------------- 1 | sub_condition,score 2 | A,6.246779887685582 3 | A,8.012698250843949 4 | A,4.867842392941125 5 | A,9.088501533767825 6 | A,11.191954616879531 7 | A,9.03881103436473 8 | A,13.364185172917518 9 | A,8.573021978401277 10 | A,9.926466129326911 11 | A,12.465760458310296 12 | A,11.006386775570943 13 | A,12.070183895116143 14 | A,10.033619762103571 15 | A,10.995384277495292 16 | A,11.151025482671098 17 | A,11.737529418597154 18 | A,10.107403368529154 19 | A,10.08743955380821 20 | A,11.046650784553353 21 | A,13.324813380199299 22 | A,6.083382763357414 23 | A,7.939830725112717 24 | A,13.491990583570459 25 | A,9.929307036417796 26 | A,13.992512133778424 27 | A,12.974132881190506 28 | A,11.308656447373941 29 | A,4.681767465282951 30 | A,6.83174563061967 31 | A,13.55373559327446 32 | A,9.861460075291635 33 | A,12.482567836867604 34 | A,9.026286027170872 35 | A,8.890472969372913 36 | A,15.303984786835697 37 | A,6.997081253818567 38 | A,8.679393631873973 39 | A,12.022863706260424 40 | A,11.716011616642925 41 | A,8.070539893966673 42 | A,6.96557749307061 43 | A,11.375007340204416 44 | A,3.147117628404277 45 | A,13.330582222085273 46 | A,13.758980277949128 47 | A,10.789485906485362 48 | A,13.519313935692331 49 | A,9.864440483873304 50 | A,13.905192248849703 51 | A,11.397705283579329 52 | B,9.987060833609098 53 | B,12.674322116332078 54 | B,8.650535157300826 55 | B,10.319448936971005 56 | B,9.752828070738742 57 | B,5.1498404825835875 58 | B,10.043538459404154 59 | B,11.65041283679112 60 | B,9.058075464476607 61 | B,15.666221948680269 62 | B,7.023959326665469 63 | B,10.847992748971976 64 | B,8.794097493656123 65 | B,6.858495841865998 66 | B,8.738308394071947 67 | B,12.793683395855911 68 | B,12.135903818770199 69 | B,5.545096598403308 70 | B,10.862029942597166 71 | B,12.108632505565287 72 | B,13.495334631089289 73 | B,9.69541423462096 74 | B,11.943259358441965 75 | B,14.305599422199359 76 | B,11.700269821286627 77 | B,13.284868705790277 78 | B,15.565431929707096 79 | B,10.342210578825032 80 | B,14.111747249830643 81 | B,8.85930297529751 82 | B,5.989949207346347 83 | B,13.572615727642773 84 | B,9.89731802756733 85 | B,15.234728382886065 86 | B,13.22383240275949 87 | B,12.241645039637488 88 | B,13.08037623715171 89 | B,8.708664350861953 90 | B,10.300344672580954 91 | B,10.774925155744045 92 | B,11.136595827526197 93 | B,9.749899247259583 94 | B,7.480331644780475 95 | B,12.411943890011894 96 | B,12.08533760621049 97 | B,12.777919437631445 98 | B,14.986425355932196 99 | B,12.264388568164854 100 | B,7.348951770289929 101 | B,10.713960442867121 102 | -------------------------------------------------------------------------------- /classes/2021_grenoble/faux.Rmd: -------------------------------------------------------------------------------- 1 | --- 2 | title: "Intro to Faux" 3 | author: "Lisa DeBruine" 4 | date: 2021-02-05 5 | output: 6 | html_document: 7 | df_print: kable 8 | --- 9 | 10 | ```{r, include = FALSE} 11 | knitr::opts_chunk$set( 12 | collapse = TRUE, 13 | out.width = "100%", 14 | fig.width = 5, 15 | fig.height = 3, 16 | dpi = 144 17 | ) 18 | set.seed(8675309) # Jenny, I've got your number 19 | ``` 20 | 21 | ```{r libs, message=FALSE} 22 | library(tidyverse) 23 | library(faux) 24 | library(broom) 25 | library(afex) 26 | ``` 27 | 28 | In this tutorial, we'll learn how to simulate data for factorial designs using {faux}. There are more extensive examples at . 29 | 30 | ## Multivariate normal 31 | 32 | You can create sets of correlated normally distributed values using `rnorm_multi()`. 33 | 34 | ```{r} 35 | dat3 <- rnorm_multi( 36 | n = 50, 37 | vars = 3, 38 | mu = c(1, 2, 3), 39 | sd = c(0.5, 1, 1.5), 40 | r = c(0, .25, .5), 41 | varnames = c("A", "B", "C") 42 | ) 43 | ``` 44 | 45 | The function `get_params()` gives you a quick way to see the means, SDs and correlations in the simulated data set to make sure you set the parameters correctly. 46 | 47 | ```{r} 48 | get_params(dat3) 49 | ``` 50 | 51 | If you set `empirical` to `TRUE`, the values you set will be the **sample** parameters, not the **population** parameters. This isn't usually what you want for a simulation, but can be useful to check you set the simulation parameters correctly. 52 | 53 | ```{r} 54 | dat3 <- rnorm_multi( 55 | n = 50, 56 | vars = 3, 57 | mu = c(1, 2, 3), 58 | sd = c(0.5, 1, 1.5), 59 | r = c(0, .25, .5), 60 | varnames = c("A", "B", "C"), 61 | empirical = TRUE 62 | ) 63 | 64 | get_params(dat3) 65 | ``` 66 | 67 | ### Shortcuts 68 | 69 | There are a few shortcuts you can use. Run the following and see if you can guess how they work. 70 | 71 | ```{r} 72 | guess1 <- rnorm_multi(50, mu = c(x = 1, y = 2, z = 3), empirical = TRUE) 73 | 74 | get_params(guess1) 75 | ``` 76 | 77 | ```{r} 78 | guess2 <- rnorm_multi(50, vars = 4, r = 0.5, empirical = TRUE) 79 | 80 | get_params(guess2) 81 | ``` 82 | 83 | ```{r} 84 | iris_r <- cor(iris[, 1:4]) 85 | iris_mu <- summarise_all(iris[, 1:4], mean) %>% t() 86 | iris_sd <- summarise_all(iris[, 1:4], sd) %>% t() 87 | 88 | guess3 <- rnorm_multi(50, 89 | mu = iris_mu, 90 | sd = iris_sd, 91 | r = iris_r) 92 | 93 | get_params(guess3) 94 | ``` 95 | 96 | You can set the r for correlations is a few different ways. 97 | 98 | ```{r} 99 | # all correlations the same value 100 | rho_same <- rnorm_multi(50, 4, r = .5, empirical = TRUE) 101 | get_params(rho_same) 102 | ``` 103 | 104 | ```{r} 105 | # upper right triangle 106 | rho_urt <- rnorm_multi(50, 4, 107 | # X2 X3 X4 108 | r = c(0.5, 0.4, 0.3, # X1 109 | 0.2, 0.1, # X2 110 | 0.0), # X3 111 | empirical = TRUE) 112 | get_params(rho_urt) 113 | ``` 114 | 115 | ```{r} 116 | # full correlation matrix 117 | rho_cormat <- rnorm_multi(50, 4, 118 | # X1 X2 X3 X4 119 | r = c(1.0, 0.5, 0.4, 0.3, # X1 120 | 0.5, 1.0, 0.2, 0.1, # X2 121 | 0.4, 0.2, 1.0, 0.0, # X3 122 | 0.3, 0.1, 0.0, 1.0), # X4 123 | empirical = TRUE) 124 | get_params(rho_cormat) 125 | ``` 126 | 127 | 128 | ```{r} 129 | rnorm_multi(10, 3, r = c(.9, .9, -.9)) 130 | ``` 131 | 132 | 133 | ## Factorial Designs 134 | 135 | You can just use `rnorm_multi()` to simulate data for each between-subjects cell of a factorial design and manually combine the tables, but faux has a function that better maps onto how we usually think and teach about factorial designs. 136 | 137 | The default design is 100 observations of one variable (named `y`) with a mean of 0 and SD of 1. Unless you set `plot = FALSE` or run `faux_options(plot = FALSE)`, this function will show you a plot of your design so you can check that it looks like you expect. 138 | 139 | ```{r} 140 | simdat1 <- sim_design() 141 | ``` 142 | 143 | 144 | ### Factors 145 | 146 | Use lists to set the names and levels of within- and between-subject factors. 147 | 148 | ```{r} 149 | pettime <- sim_design( 150 | within = list(time = c("pre", "post"), 151 | condition = c("A", "B")), 152 | between = list(pet = c("cat", "dog", "ferret")) 153 | ) 154 | ``` 155 | 156 | You can set mu and sd with unnamed vectors, but getting the order right can be tricky. 157 | 158 | ```{r} 159 | pettime <- sim_design( 160 | within = list(time = c("pre", "post")), 161 | between = list(pet = c("cat", "dog", "ferret")), 162 | mu = 1:6 163 | ) 164 | ``` 165 | 166 | You can set values with a named vector for a single type of factor. The values do not have to be in the right order if they're named. 167 | 168 | ```{r} 169 | pettime <- sim_design( 170 | within = list(time = c("pre", "post")), 171 | between = list(pet = c("cat", "dog", "ferret")), 172 | mu = c(cat = 1, ferret = 5, dog = 3), 173 | sd = c(pre = 1, post = 2) 174 | ) 175 | ``` 176 | 177 | Or use a data frame for within- and between-subject factors. 178 | 179 | ```{r} 180 | 181 | mu <- data.frame( 182 | pre_A = c(1, 3, 5), 183 | post_A = c(2, 4, 6), 184 | pre_B = c(10, 30, 50), 185 | post_B = c(20, 40, 60), 186 | row.names = c("cat", "dog", "ferret") 187 | ) 188 | 189 | pettime <- sim_design( 190 | within = list(time = c("pre", "post"), 191 | condition = c("A","B")), 192 | between = list(pet = c("cat", "dog", "ferret")), 193 | mu = mu 194 | ) 195 | ``` 196 | 197 | If you have within-subject factors, set the correlations for each between-subject cell like this. You need to tell `get_params()` if you have any between-subject columns. 198 | 199 | ```{r} 200 | pettime <- sim_design( 201 | within = list(time = c("pre", "post")), 202 | between = list(pet = c("cat", "dog", "ferret")), 203 | r = list(cat = 0.5, 204 | dog = 0.25, 205 | ferret = 0), 206 | empirical = TRUE, 207 | plot = FALSE 208 | ) 209 | 210 | get_params(pettime, between = "pet") 211 | ``` 212 | 213 | You can also change the name of the `dv` and `id` columns and output the data in long format. If you do this, you also need to tell `get_params()` what columns contain the between- and within-subject factors, the dv, and the id. 214 | 215 | ```{r} 216 | dat_long <- sim_design( 217 | within = list(time = c("pre", "post")), 218 | between = list(pet = c("cat", "dog", "ferret")), 219 | id = "subj_id", 220 | dv = "score", 221 | long = TRUE, 222 | plot = FALSE 223 | ) 224 | 225 | get_params(dat_long, 226 | between = "pet", 227 | within = "time", 228 | id = "subj_id", 229 | dv = "score", 230 | digits = 3) 231 | ``` 232 | 233 | ### Anonymous Factors 234 | 235 | If you need to make a quick demo, you can set factors anonymously with integer vectors. 236 | 237 | ```{r} 238 | dat_anon <- sim_design( 239 | n = 50, 240 | between = list(pet = c(dog = "Doggies", cat = "Kittens")), 241 | dv = c(score = "Happiness Score") 242 | ) 243 | 244 | x <- attr(dat_anon, "design") 245 | ``` 246 | 247 | Faux has a quick plotting function for visualising data made with sim_design. 248 | 249 | ```{r} 250 | plot(dat_anon) 251 | ``` 252 | You can change the order of plotting and the types of geoms plotted. This takes a little trial and error, so this function will probably be refined in later versions. 253 | 254 | ```{r} 255 | plot(dat_anon, "B", "A", "C", geoms = c("jitter")) 256 | ``` 257 | 258 | ### Replications 259 | 260 | You often want to simulate data repeatedly to do things like calculate power. The `sim_design()` function has a lot of overhead for checking if a design makes sense and if the correlation matrix is possible, so you can speed up the creation of multiple datasets with the same design using the `rep` argument. This will give you a nested data from with each dataset in the `data` column. 261 | 262 | ```{r} 263 | dat_rep <- sim_design( 264 | within = 2, 265 | n = 20, 266 | mu = c(0, 0.25), 267 | rep = 10, 268 | plot = FALSE 269 | ) 270 | ``` 271 | 272 | You can run analyses on the nested data like this: 273 | 274 | ```{r} 275 | map_df(dat_rep$data, ~{ 276 | t.test(.x$A1, .x$A2, paired = TRUE) %>% broom::tidy() 277 | }) 278 | ``` 279 | 280 | 281 | ## Exercises 282 | 283 | ### Multivariate normal 284 | 285 | Sample 40 values of three variables named `J`, `K` and `L` from a population with means of 10, 20 and 30, and SDs of 5. `J` and `K` are correlated 0.5, `J` and `L` are correlated 0.25, and `K` and `L` are not correlated. 286 | 287 | ```{r, include=FALSE} 288 | ex1 <- rnorm_multi(n = 40, mu = c(J = 10, K = 20, L = 30), 289 | sd = 5, r = c(0.5, 0.25, 0)) 290 | 291 | get_params(ex1) 292 | ``` 293 | 294 | ### From existing data 295 | 296 | Using the data from the built-in dataset `attitude`, simulate a new set of 20 observations drawn from a population with the same means, SDs and correlations for each column as the original data. 297 | 298 | ```{r, include=FALSE} 299 | dat_r <- cor(attitude) 300 | dat_mu <- summarise_all(attitude, mean) %>% t() 301 | dat_sd <- summarise_all(attitude, sd) %>% t() 302 | 303 | ex2 <- rnorm_multi(20, mu = dat_mu, sd = dat_sd,r = dat_r) 304 | 305 | get_params(ex2) 306 | ``` 307 | 308 | 309 | ### 2b 310 | 311 | Create a dataset with a between-subject factor of "pet" having two levels, "cat", and "dog". The DV is "happiness" score. There are 20 cat-owners with a mean happiness score of 10 (SD = 3) and there are 30 dog-owners with a mean happiness score of 11 (SD = 3). 312 | 313 | ```{r, include=FALSE} 314 | dat2b <- sim_design( 315 | between = list(pet = c("cat", "dog")), 316 | dv = "happiness", 317 | n = list(cat = 20, dog = 30), 318 | mu = list(cat = 10, dog = 11), 319 | sd = 3 320 | ) 321 | 322 | get_params(dat2b, between = "pet") 323 | ``` 324 | 325 | ### 3w 326 | 327 | Create a dataset of 20 observations with 1 within-subject variable ("condition") having 3 levels ("A", "B", "C") with means of 10, 20 and 30 and SD of 5. The correlations between each level have r = 0.4. The dataset should look like this: 328 | 329 | | id | condition | score | 330 | |:---|:----------|------:| 331 | |S01 | A | 9.17 | 332 | |... | ... | ... | 333 | |S20 | A | 11.57 | 334 | |S01 | B | 18.44 | 335 | |... | ... | ... | 336 | |S20 | B | 20.04 | 337 | |S01 | C | 35.11 | 338 | |... | ... | ... | 339 | |S20 | C | 29.16 | 340 | 341 | ```{r, include=FALSE} 342 | 343 | dat3w <- sim_design( 344 | within = list(condition = c("A", "B", "C")), 345 | n = 20, 346 | mu = c(10, 20, 30), 347 | sd = 5, 348 | r = .4, 349 | dv = "score", 350 | long = TRUE 351 | ) 352 | 353 | get_params(dat3w) 354 | 355 | ``` 356 | 357 | ### 2w*2w 358 | 359 | Create a dataset with 50 observations of 2 within-subject variables ("A" and "B") each having 2 levels. The mean for all cells is 10 and the SD is 2. The dataset should have 20 subjects. The correlations look like this: 360 | 361 | | | A1_B1 | A1_B2 | A2_B1 | A2_B2 | 362 | |:------|------:|------:|------:|------:| 363 | | A1_B1 | 1.0 | 0.5 | 0.5 | 0.2 | 364 | | A1_B2 | 0.5 | 1.0 | 0.2 | 0.5 | 365 | | A2_B1 | 0.5 | 0.2 | 1.0 | 0.5 | 366 | | A2_B2 | 0.2 | 0.5 | 0.5 | 1.0 | 367 | 368 | 369 | ```{r, include=FALSE} 370 | dat2w2w <- sim_design( 371 | within = c(2,2), 372 | n = 50, 373 | mu = 10, 374 | sd = 2, 375 | r = c(.5, .5, .2, 376 | .2, .5, 377 | .5) 378 | ) 379 | 380 | get_params(dat2w2w) 381 | ``` 382 | 383 | ### 2w*3b 384 | 385 | Create a dataset with a between-subject factor of "pet" having 3 levels ("cat", "dog", and "ferret") and a within-subject factor of "time" having 2 levels ("pre" and "post"). The N in each group should be 10. Means are: 386 | 387 | * cats: pre = 10, post = 12 388 | * dogs: pre = 14, post = 16 389 | * ferrets: pre = 18, post = 20 390 | 391 | SDs are all 5 and within-cell correlations are all 0.25. 392 | 393 | ```{r, include=FALSE} 394 | 395 | mu <- data.frame( 396 | cat = c(10, 12), 397 | dog = c(14, 16), 398 | ferret = c(18, 20) 399 | ) 400 | 401 | dat2w3b <- sim_design( 402 | within = list(time = c("pre", "post")), 403 | between = list(pet = c("cat", "dog", "ferret")), 404 | n = 10, 405 | mu = mu, 406 | sd = 5, 407 | r = 0.25 408 | ) 409 | 410 | get_params(dat2w3b) 411 | 412 | ``` 413 | 414 | ### Replications 415 | 416 | Create 5 datasets with a 2b*2b design, 30 participants in each cell. Each cell's mean should be 0, except A1_B1, which should be 0.5. The SD should be 1. Make the resulting data in long format. 417 | 418 | ```{r, include=FALSE} 419 | dat2b2b <- sim_design( 420 | between = c(2,2), 421 | n = 30, 422 | mu = c(0.5, 0, 0, 0), 423 | rep = 5, 424 | long = TRUE 425 | ) 426 | ``` 427 | 428 | ### Power 429 | 430 | Simulate 100 datasets like the one above and use `lm()` or `afex::aov_ez()` to look at the interaction between A and B. What is the power of this design? 431 | 432 | ```{r, include=FALSE} 433 | dat2b2b_100 <- sim_design( 434 | between = c(2, 2), 435 | n = 30, 436 | mu = c(0.5, 0, 0, 0), 437 | rep = 100, 438 | long = TRUE 439 | ) 440 | 441 | ana_lm <- map_df(dat2b2b_100$data, ~{ 442 | lm(y ~ A*B, data = .x) %>% broom::tidy() 443 | }) 444 | 445 | afex::set_sum_contrasts() # avoids annoying afex message 446 | ana_aov <- map_df(dat2b2b_100$data, ~{ 447 | afex::aov_ez(id = "id", 448 | dv = "y", 449 | between = c("A", "B"), 450 | data = .x, 451 | return = "aov") %>% broom::tidy() 452 | }) 453 | 454 | ana_aov %>% 455 | group_by(term) %>% 456 | summarise(power = mean(p.value < .05), 457 | .groups = "drop") 458 | ``` 459 | 460 | -------------------------------------------------------------------------------- /classes/2021_grenoble/sim1.csv: -------------------------------------------------------------------------------- 1 | effect,group,term,estimate,std.error,statistic,df,p.value 2 | fixed,NA,(Intercept),372.8236962436822,10.08365913366393,36.97305623898212,165.65881763290963,6.07720610062092e-82 3 | fixed,NA,sub_cond.e,68.73580843544242,15.076692205038604,4.559077515190702,222.03595696791052,8.487839790510339e-6 4 | fixed,NA,stim_version.e,50.07083246061155,2.09919721932919,23.852371754099387,154.6466162980542,1.0965014966349348e-53 5 | fixed,NA,sub_cond.e:stim_version.e,4.309073973261419,3.757369398397111,1.1468326683822103,198.16657034852403,0.25283358257310784 6 | ran_pars,sub_id,sd__(Intercept),103.06082058775809,NA,NA,NA,NA 7 | ran_pars,sub_id.1,sd__stim_version.e,21.054200010708943,NA,NA,NA,NA 8 | ran_pars,stim_id,sd__(Intercept),49.26529727552844,NA,NA,NA,NA 9 | ran_pars,stim_id.1,sd__stim_version.e,10.162977497391143,NA,NA,NA,NA 10 | ran_pars,stim_id.2,sd__sub_cond.e,27.158316402506028,NA,NA,NA,NA 11 | ran_pars,stim_id.3,sd__stim_version.e:sub_cond.e,15.417843335648813,NA,NA,NA,NA 12 | ran_pars,Residual,sd__Observation,24.9509561875767,NA,NA,NA,NA 13 | -------------------------------------------------------------------------------- /data-sim-workshops.Rproj: -------------------------------------------------------------------------------- 1 | Version: 1.0 2 | 3 | RestoreWorkspace: Default 4 | SaveWorkspace: Default 5 | AlwaysSaveHistory: Default 6 | 7 | EnableCodeIndexing: Yes 8 | UseSpacesForTab: Yes 9 | NumSpacesForTab: 2 10 | Encoding: UTF-8 11 | 12 | RnwWeave: Sweave 13 | LaTeX: pdfLaTeX 14 | 15 | BuildType: Package 16 | PackageUseDevtools: Yes 17 | PackageInstallArgs: --no-multiarch --with-keep.source 18 | -------------------------------------------------------------------------------- /docs/404.html: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | Page not found (404) • dsw 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 25 | 26 | 27 | Skip to contents 28 | 29 | 30 |
73 |
74 |
75 |

Page not found (404)

76 | 77 |
78 | 79 | Content not found. Please use links in the navbar. 80 | 81 |
82 |
83 | 84 | 85 |
86 |

Developed by Lisa DeBruine.

87 |
88 | 89 |
90 |

Site built with pkgdown 2.0.9.

91 |
92 | 93 |
94 |
95 | 96 | 97 | 98 | 99 | 100 | 101 | 102 | -------------------------------------------------------------------------------- /docs/apple-touch-icon-120x120.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/apple-touch-icon-120x120.png -------------------------------------------------------------------------------- /docs/apple-touch-icon-152x152.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/apple-touch-icon-152x152.png -------------------------------------------------------------------------------- /docs/apple-touch-icon-180x180.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/apple-touch-icon-180x180.png -------------------------------------------------------------------------------- /docs/apple-touch-icon-60x60.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/apple-touch-icon-60x60.png -------------------------------------------------------------------------------- /docs/apple-touch-icon-76x76.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/apple-touch-icon-76x76.png -------------------------------------------------------------------------------- /docs/apple-touch-icon.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/apple-touch-icon.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-13-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-13-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-14-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-14-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-15-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-15-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-16-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-16-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-17-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-17-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-20-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-20-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-21-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-21-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-22-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-22-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-23-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-23-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-24-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-24-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-28-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-28-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-37-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-37-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-4-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-4-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-40-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-40-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-41-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-41-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-42-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-42-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-43-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-43-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-44-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-44-1.png -------------------------------------------------------------------------------- /docs/articles/faux_files/figure-html/unnamed-chunk-45-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/faux_files/figure-html/unnamed-chunk-45-1.png -------------------------------------------------------------------------------- /docs/articles/fixed_files/figure-html/ind-sim-fig-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/fixed_files/figure-html/ind-sim-fig-1.png -------------------------------------------------------------------------------- /docs/articles/fixed_files/figure-html/pair-sim-fig-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/fixed_files/figure-html/pair-sim-fig-1.png -------------------------------------------------------------------------------- /docs/articles/fixed_files/figure-html/rnorm-plot-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/fixed_files/figure-html/rnorm-plot-1.png -------------------------------------------------------------------------------- /docs/articles/fixed_files/figure-html/runif-hist-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/fixed_files/figure-html/runif-hist-1.png -------------------------------------------------------------------------------- /docs/articles/fixed_files/figure-html/sample-prob-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/fixed_files/figure-html/sample-prob-1.png -------------------------------------------------------------------------------- /docs/articles/fixed_files/figure-html/sample-replace-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/fixed_files/figure-html/sample-replace-1.png -------------------------------------------------------------------------------- /docs/articles/fixed_files/figure-html/sim-p-fig-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/fixed_files/figure-html/sim-p-fig-1.png -------------------------------------------------------------------------------- /docs/articles/index.html: -------------------------------------------------------------------------------- 1 | 2 | Articles • dsw 6 | Skip to contents 7 | 8 | 9 |
47 |
48 |
49 |

Articles

50 |
51 | 52 |
53 |

All vignettes

54 |
55 | 56 |
Calorie Placement Re-Simulation
57 |
58 |
Intro to Faux
59 |
60 |
Fixed Effects
61 |
62 |
Mixed Effects
63 |
64 |
65 |
66 | 67 | 68 |
69 |

Developed by Lisa DeBruine.

70 |
71 | 72 |
73 |

Site built with pkgdown 2.0.9.

74 |
75 | 76 |
77 | 78 | 79 | 80 | 81 | 82 | 83 | 84 | -------------------------------------------------------------------------------- /docs/articles/mixed_files/figure-html/ex2-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/mixed_files/figure-html/ex2-1.png -------------------------------------------------------------------------------- /docs/articles/mixed_files/figure-html/plot-dv-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/mixed_files/figure-html/plot-dv-1.png -------------------------------------------------------------------------------- /docs/articles/mixed_files/figure-html/plot-ixn-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/mixed_files/figure-html/plot-ixn-1.png -------------------------------------------------------------------------------- /docs/articles/mixed_files/figure-html/plot-stim-ranef-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/mixed_files/figure-html/plot-stim-ranef-1.png -------------------------------------------------------------------------------- /docs/articles/mixed_files/figure-html/plot-sub-ranef-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/mixed_files/figure-html/plot-sub-ranef-1.png -------------------------------------------------------------------------------- /docs/articles/mixed_files/figure-html/rslope-plot-dv-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/mixed_files/figure-html/rslope-plot-dv-1.png -------------------------------------------------------------------------------- /docs/articles/mixed_files/figure-html/unnamed-chunk-11-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/mixed_files/figure-html/unnamed-chunk-11-1.png -------------------------------------------------------------------------------- /docs/articles/mixed_files/figure-html/unnamed-chunk-7-1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/articles/mixed_files/figure-html/unnamed-chunk-7-1.png -------------------------------------------------------------------------------- /docs/authors.html: -------------------------------------------------------------------------------- 1 | 2 | Authors and Citation • dsw 6 | Skip to contents 7 | 8 | 9 |
47 |
48 |
49 |

Authors and Citation

50 |
51 | 52 |
53 |

Authors

54 | 55 |
  • 56 |

    Lisa DeBruine. Author, maintainer. 57 |

    58 |
    • 59 |
60 | 61 |
62 |

Citation

63 |

Source: DESCRIPTION

64 | 65 |

DeBruine L (2024). 66 | dsw: Data Simulation Workshops. 67 | R package version 0.0.0.9007, 68 | https://debruine.github.io/data-sim-workshops/, https://github.com/debruine/data-sim-workshops. 69 |

70 | 
@Manual{,
71 |   title = {dsw: Data Simulation Workshops},
72 |   author = {Lisa DeBruine},
73 |   year = {2024},
74 |   note = {R package version 0.0.0.9007, 
75 | https://debruine.github.io/data-sim-workshops/},
76 |   url = {https://github.com/debruine/data-sim-workshops},
77 | }
78 |
79 |
81 | 82 | 83 |
84 |

Developed by Lisa DeBruine.

85 |
86 | 87 |
88 |

Site built with pkgdown 2.0.9.

89 |
90 | 91 |
92 | 93 | 94 | 95 | 96 | 97 | 98 | 99 | -------------------------------------------------------------------------------- /docs/deps/data-deps.txt: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | -------------------------------------------------------------------------------- /docs/favicon-16x16.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/favicon-16x16.png -------------------------------------------------------------------------------- /docs/favicon-32x32.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/favicon-32x32.png -------------------------------------------------------------------------------- /docs/favicon.ico: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/favicon.ico -------------------------------------------------------------------------------- /docs/index.html: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | Data Simulation Workshops • dsw 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 27 | 28 | 29 | Skip to contents 30 | 31 | 32 |
75 |
76 |
77 |
78 |

Data Simulation Workshop Materials 79 |

80 |
81 |

Being able to simulate data allows you to:

82 |
    83 |
  • prep analysis scripts for pre-registration
    • 84 |
  • calculate power and sensitivity for analyses that don’t have empirical methods
    • 85 |
  • create reproducible examples when your data are too big or confidential to share
    • 86 |
  • enhance your understanding of statistical concepts
    • 87 |
  • create demo data for teaching and tutorials
    • 88 |
89 |
90 |

Installation 91 |

92 |

You can install the packages used in these tutorials and get a function that makes it easy to access the workshop .Rmd files by running the following code:

93 | 
 94 | devtools::install_github("debruine/data-sim-workshops")
95 |

Then you can load exercises with the following code:

96 | 
 97 | dsw::exercise("faux")
 98 | dsw::exercise("calories")
 99 | dsw::exercise("fixed")
100 | dsw::exercise("mixed")
101 |

Alternatively, download the stub files and install the specific packages for your workshop.

102 | 108 |
109 |
110 |

Upcoming Workshops 111 |

112 | 140 |

When: 2024 February 1-2 Where: Data Simulation Workshop 2024, Institute of Psychology, Bern , Switzerland

141 |
142 |

Data Simulation with {faux} 143 |

144 |

This session will cover the basics of simulation using {faux}. We will simulate data with factorial designs by specifying the within and between-subjects factor structure, each cell mean and standard deviation, and correlations between cells where appropriate. This can be used to create simulated data sets to be used in preparing the analysis code for pre-registrations or registered reports. We will also create data sets for simulation-based power analyses. Students will need to have very basic knowledge of R and R Markdown, and have installed {faux}, {afex}, {broom} and {tidyverse}.

145 |
146 |

Prep 147 |

148 |
    149 |
  • Install R packages from CRAN: tidyverse, afex, faux, and broom 150 |
    • 151 |
  • Download files: faux-stub.Rmd 152 |
    • 153 |
154 |
155 |
156 |
157 |

Data simulation for mixed designs 158 |

159 |

This session will cover simulating data for a mixed design, where trials are crossed with subjects. We will learn how to analyse this using {lme4}, with a focus on understanding how the simulation parameters correspond to the output. Finally, we will learn how to use simulation to calculate power. Students will need to have basic knowledge of R and R Markdown, some familiarity with mixed designs (even if they don’t currently analyse them with mixed models) and have installed {faux}, {afex}, {tidyverse}, and {lme4}.

160 |
161 |

Prep 162 |

163 |
    164 |
  • Install R packages from CRAN: tidyverse, afex, lme4, broom, broom.mixed, faux 165 |
    • 166 |
  • Download files: mixed-stub.Rmd 167 |
    • 168 |
169 |
170 |
171 |
172 |
173 |

Resources 174 |

175 | 183 |
184 |
185 |

Past Workshops 186 |

187 |
    188 |

  • Vrije Universiteit Amsterdam, NL Fake It Until You Make It: How and why to simulate research data 2023 September 20

    • 189 |

  • Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany
    190 | Simulating data with {faux}
    191 | 2023 July 27 9:00 - 12:00 (CET)

    • 192 |

  • European Evolutionary Biology Conference, Millport, Scotland
    193 | Fake It Until You Make It: How and why to simulate research data
    194 | 2023 June 1 14:30 - 16:30 (GMT)

    • 195 |

  • University of Glasgow Institute of Neuroscience & Psychology
    196 | Data Simulation with {faux}
    197 | 2023 January 18 12:00 - 13:00 (GMT)

    • 198 |

  • Netherlands Institute for the Study of Crime and Law Enforcement
    199 | Data Simulation with {faux}
    200 | 2022 December 6 13:00 - 14:00 (CET)

    • 201 |

  • Polish Association of Social Psychology Conference, Gdánsk
    202 | Data simulation for fixed effects
    203 | Data simulation for mixed designs
    204 | Practical Session
    205 | 2022 September 14 09:00 - 16:00 (CET)

    • 206 |

  • RLadies Glasgow
    207 | Data simulation using faux
    208 | 2022 May 24 15:00-17:00 (BST)

    • 209 |

  • University of York Data simulation for factorial designs
    210 | Data simulation for mixed designs
    211 | 2022 April 27 09:00-17:00 (BST)

    • 212 |

  • From Proposal to Publication: Pathways to Open Science
    213 | Data simulation for factorial designs
    214 | Data simulation for mixed designs
    215 | 2022 July 13 13:30-17:00

    • 216 |

  • University of Glasgow
    217 | Institute of Neuroscience and Psychology
    218 | 2020 Jan 28 13:00-15:00 and Feb 5 14:00-16:00

    • 219 |

  • University of Grenoble
    220 | Understanding Mixed-Effects Models through Data Simulation
    221 | 2021 February 5 13:00-15:00

    • 222 |

  • PsyPAG Data Simulation Summer School
    223 | Simulation for factorial designs with faux
    224 | 2021 June 4 13:00-15:00

    • 225 |
226 |
227 |
228 |
262 |
263 | 264 | 265 |
266 |

Developed by Lisa DeBruine.

267 |
268 | 269 |
270 |

Site built with pkgdown 2.0.9.

271 |
272 | 273 |
274 |
275 | 276 | 277 | 278 | 279 | 280 | 281 | 282 | -------------------------------------------------------------------------------- /docs/link.svg: -------------------------------------------------------------------------------- 1 | 2 | 3 | 5 | 8 | 12 | 13 | -------------------------------------------------------------------------------- /docs/logo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/logo.png -------------------------------------------------------------------------------- /docs/pkgdown.js: -------------------------------------------------------------------------------- 1 | /* http://gregfranko.com/blog/jquery-best-practices/ */ 2 | (function($) { 3 | $(function() { 4 | 5 | $('nav.navbar').headroom(); 6 | 7 | Toc.init({ 8 | $nav: $("#toc"), 9 | $scope: $("main h2, main h3, main h4, main h5, main h6") 10 | }); 11 | 12 | if ($('#toc').length) { 13 | $('body').scrollspy({ 14 | target: '#toc', 15 | offset: $("nav.navbar").outerHeight() + 1 16 | }); 17 | } 18 | 19 | // Activate popovers 20 | $('[data-bs-toggle="popover"]').popover({ 21 | container: 'body', 22 | html: true, 23 | trigger: 'focus', 24 | placement: "top", 25 | sanitize: false, 26 | }); 27 | 28 | $('[data-bs-toggle="tooltip"]').tooltip(); 29 | 30 | /* Clipboard --------------------------*/ 31 | 32 | function changeTooltipMessage(element, msg) { 33 | var tooltipOriginalTitle=element.getAttribute('data-bs-original-title'); 34 | element.setAttribute('data-bs-original-title', msg); 35 | $(element).tooltip('show'); 36 | element.setAttribute('data-bs-original-title', tooltipOriginalTitle); 37 | } 38 | 39 | if(ClipboardJS.isSupported()) { 40 | $(document).ready(function() { 41 | var copyButton = ""; 42 | 43 | $("div.sourceCode").addClass("hasCopyButton"); 44 | 45 | // Insert copy buttons: 46 | $(copyButton).prependTo(".hasCopyButton"); 47 | 48 | // Initialize tooltips: 49 | $('.btn-copy-ex').tooltip({container: 'body'}); 50 | 51 | // Initialize clipboard: 52 | var clipboard = new ClipboardJS('[data-clipboard-copy]', { 53 | text: function(trigger) { 54 | return trigger.parentNode.textContent.replace(/\n#>[^\n]*/g, ""); 55 | } 56 | }); 57 | 58 | clipboard.on('success', function(e) { 59 | changeTooltipMessage(e.trigger, 'Copied!'); 60 | e.clearSelection(); 61 | }); 62 | 63 | clipboard.on('error', function(e) { 64 | changeTooltipMessage(e.trigger,'Press Ctrl+C or Command+C to copy'); 65 | }); 66 | 67 | }); 68 | } 69 | 70 | /* Search marking --------------------------*/ 71 | var url = new URL(window.location.href); 72 | var toMark = url.searchParams.get("q"); 73 | var mark = new Mark("main#main"); 74 | if (toMark) { 75 | mark.mark(toMark, { 76 | accuracy: { 77 | value: "complementary", 78 | limiters: [",", ".", ":", "/"], 79 | } 80 | }); 81 | } 82 | 83 | /* Search --------------------------*/ 84 | /* Adapted from https://github.com/rstudio/bookdown/blob/2d692ba4b61f1e466c92e78fd712b0ab08c11d31/inst/resources/bs4_book/bs4_book.js#L25 */ 85 | // Initialise search index on focus 86 | var fuse; 87 | $("#search-input").focus(async function(e) { 88 | if (fuse) { 89 | return; 90 | } 91 | 92 | $(e.target).addClass("loading"); 93 | var response = await fetch($("#search-input").data("search-index")); 94 | var data = await response.json(); 95 | 96 | var options = { 97 | keys: ["what", "text", "code"], 98 | ignoreLocation: true, 99 | threshold: 0.1, 100 | includeMatches: true, 101 | includeScore: true, 102 | }; 103 | fuse = new Fuse(data, options); 104 | 105 | $(e.target).removeClass("loading"); 106 | }); 107 | 108 | // Use algolia autocomplete 109 | var options = { 110 | autoselect: true, 111 | debug: true, 112 | hint: false, 113 | minLength: 2, 114 | }; 115 | var q; 116 | async function searchFuse(query, callback) { 117 | await fuse; 118 | 119 | var items; 120 | if (!fuse) { 121 | items = []; 122 | } else { 123 | q = query; 124 | var results = fuse.search(query, { limit: 20 }); 125 | items = results 126 | .filter((x) => x.score <= 0.75) 127 | .map((x) => x.item); 128 | if (items.length === 0) { 129 | items = [{dir:"Sorry 😿",previous_headings:"",title:"No results found.",what:"No results found.",path:window.location.href}]; 130 | } 131 | } 132 | callback(items); 133 | } 134 | $("#search-input").autocomplete(options, [ 135 | { 136 | name: "content", 137 | source: searchFuse, 138 | templates: { 139 | suggestion: (s) => { 140 | if (s.title == s.what) { 141 | return `${s.dir} >
${s.title}
`; 142 | } else if (s.previous_headings == "") { 143 | return `${s.dir} >
${s.title}
> ${s.what}`; 144 | } else { 145 | return `${s.dir} >
${s.title}
> ${s.previous_headings} > ${s.what}`; 146 | } 147 | }, 148 | }, 149 | }, 150 | ]).on('autocomplete:selected', function(event, s) { 151 | window.location.href = s.path + "?q=" + q + "#" + s.id; 152 | }); 153 | }); 154 | })(window.jQuery || window.$) 155 | 156 | 157 | -------------------------------------------------------------------------------- /docs/pkgdown.yml: -------------------------------------------------------------------------------- 1 | pandoc: 3.1.1 2 | pkgdown: 2.0.9 3 | pkgdown_sha: ~ 4 | articles: 5 | calories: calories.html 6 | faux: faux.html 7 | fixed: fixed.html 8 | mixed: mixed.html 9 | last_built: 2024-04-24T07:57Z 10 | urls: 11 | reference: https://debruine.github.io/data-sim-workshops/reference 12 | article: https://debruine.github.io/data-sim-workshops/articles 13 | 14 | -------------------------------------------------------------------------------- /docs/reference/Rplot001.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/reference/Rplot001.png -------------------------------------------------------------------------------- /docs/reference/exercise.html: -------------------------------------------------------------------------------- 1 | 2 | Get an exercise — exercise • dsw 6 | Skip to contents 7 | 8 | 9 |
47 |
48 |
49 |

Get an exercise

50 | Source: R/exercises.R 51 |
exercise.Rd
52 |
53 | 54 |
55 |

Get an exercise

56 |
57 | 58 |
59 |

Usage

60 | 
exercise(name = c("faux", "fixed", "mixed", "calories"), filename = NULL)
61 |
62 | 63 |
64 |

Arguments

65 |
name
66 |

The name of the exercise

67 | 68 | 69 |
filename
70 |

What filename you want to save (defaults to the name of the exercise in the working directory)

71 | 72 |
73 |
74 |

Value

75 | 76 | 77 |

Saves a file to the working directory (or path from filename)

78 |
79 | 80 |
81 |

Examples

82 | 
if (FALSE) {
 83 | exercise("faux") # get exercise for the faux workshop
 84 | exercise("fixed", "exercises/fixed.Rmd") # save into exercises directory
 85 | }
 86 |
87 |
88 |
90 | 91 | 92 |
93 |

Developed by Lisa DeBruine.

94 |
95 | 96 |
97 |

Site built with pkgdown 2.0.9.

98 |
99 | 100 |
101 | 102 | 103 | 104 | 105 | 106 | 107 | 108 | -------------------------------------------------------------------------------- /docs/reference/figures/logo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/docs/reference/figures/logo.png -------------------------------------------------------------------------------- /docs/reference/index.html: -------------------------------------------------------------------------------- 1 | 2 | Function reference • dsw 6 | Skip to contents 7 | 8 | 9 |
47 |
48 |
49 |

Function reference

50 |
51 | 52 |
53 |

All functions

54 | 55 | 56 | 57 | 58 |
59 | 60 | 61 | 62 | 63 |
64 | 65 | exercise() 66 |
67 |
Get an exercise
68 |
69 |
70 | 71 | 72 |
73 |

Developed by Lisa DeBruine.

74 |
75 | 76 |
77 |

Site built with pkgdown 2.0.9.

78 |
79 | 80 |
81 | 82 | 83 | 84 | 85 | 86 | 87 | 88 | -------------------------------------------------------------------------------- /docs/sitemap.xml: -------------------------------------------------------------------------------- 1 | 2 | 3 | 4 | https://debruine.github.io/data-sim-workshops/404.html 5 | 6 | 7 | https://debruine.github.io/data-sim-workshops/LICENSE.html 8 | 9 | 10 | https://debruine.github.io/data-sim-workshops/articles/calories.html 11 | 12 | 13 | https://debruine.github.io/data-sim-workshops/articles/faux.html 14 | 15 | 16 | https://debruine.github.io/data-sim-workshops/articles/fixed.html 17 | 18 | 19 | https://debruine.github.io/data-sim-workshops/articles/index.html 20 | 21 | 22 | https://debruine.github.io/data-sim-workshops/articles/mixed.html 23 | 24 | 25 | https://debruine.github.io/data-sim-workshops/authors.html 26 | 27 | 28 | https://debruine.github.io/data-sim-workshops/index.html 29 | 30 | 31 | https://debruine.github.io/data-sim-workshops/reference/exercise.html 32 | 33 | 34 | https://debruine.github.io/data-sim-workshops/reference/index.html 35 | 36 | 37 | -------------------------------------------------------------------------------- /inst/stubs/calories-stub.Rmd: -------------------------------------------------------------------------------- 1 | --- 2 | title: "Calorie Placement Re-Simulation" 3 | output: 4 | html_document: 5 | df_print: kable 6 | --- 7 | 8 | ```{r setup, include=FALSE} 9 | knitr::opts_chunk$set(error = TRUE) 10 | library(tidyverse) 11 | library(faux) 12 | library(afex) 13 | library(emmeans) 14 | faux_options(plot = FALSE) 15 | 16 | set.seed(8675309) 17 | ``` 18 | 19 | ## Data Source 20 | 21 | We will be replicating some of the re-analyses in Francis & Thunell's (2020) Meta-Psychology paper: Excess success in "Don't count calorie labeling out: Calorie counts on the left side of menu items lead to lower calorie food choices". 22 | 23 | They ran power analyses for all 6 studies in Dallas, Liu, and Ubel's (2019) study showing that people order food with significantly fewer calories when the calorie count was placed to the left of the item than to the right (or having no calorie label). They then used these power estimates to calculate the probability of all 6 out of 6 studies being significant, given the observed power of each study. 24 | 25 | * [Re-analysis](https://doi.org/10.15626/MP.2019.2266) 26 | * [Re-analysis code](https://osf.io/xrdhj/) 27 | * [Original paper](https://doi.org/10.1002/jcpy.1053) 28 | 29 | Table 1 of the re-analysis paper provides all of the parameters we will need. 30 | 31 | ## Reanalyses 32 | 33 | ### Study 2 34 | 35 | We'll start with S2 because the analysis is very straightforward. It's a between-subjects design, where 143 subjects saw calorie placement on the left and their mean calories ordered were 1249.83 (SD = 449.07), while 132 subjects saw calorie placement on the right and their mean calories ordered were 1362.31 (SD = 447.35). 36 | 37 | Let's first simulate a single data table with these parameters and set up our analysis. 38 | 39 | ```{r} 40 | data <- NULL 41 | ``` 42 | 43 | Wrap the analysis in a function using the `tidy()` function from {broom} to get the results in a tidy table. Check that it works by running it on the single data set above. 44 | 45 | ```{r} 46 | s2_analyse <- function(data) { 47 | } 48 | 49 | s2_analyse(data) 50 | ``` 51 | 52 | 53 | Now, simulate the data 500 times. 54 | 55 | ```{r} 56 | s2 <- NULL 57 | ``` 58 | 59 | Run the analysis on each data set. 60 | 61 | ```{r} 62 | s2_sim <- NULL 63 | 64 | head(s2_sim) 65 | ``` 66 | 67 | Summarise the `p.value` column to get power. 68 | 69 | ```{r} 70 | s2_power <- NULL 71 | ``` 72 | 73 | Compare this value (`r s2_power`) with the value in the paper (0.5426). 74 | 75 | ### Study 1 76 | 77 | Study 1 is a little more complicated because the design includes a "no label" condition, so the decision rule for supporting the hypothesis is more complicated. 78 | 79 | The data simulation is relatively straightforward, though. 80 | 81 | ```{r} 82 | mu = c(left = 654.53, right = 865.41, none = 914.34) 83 | sd = c(left = 390.45, right = 517.26, none = 560.94) 84 | n = c(left = 45, right = 54, none = 50) 85 | 86 | data <- NULL 87 | ``` 88 | 89 | Set up the analysis. Here, we really just care about three p-values, so we'll just return those. We can use a function from the {emmeans} package to check the two pairwise comparisons. 90 | 91 | ```{r} 92 | afex::set_sum_contrasts() # avoids annoying afex message on each run 93 | afex_options(include_aov = TRUE) # we need aov for emmeans 94 | 95 | s1_analyse <- function(data) { 96 | # main effect of placement 97 | a <- afex::aov_ez( 98 | id = "id", 99 | dv = "calories", 100 | between = "placement", 101 | data = data 102 | ) 103 | 104 | # contrasts 105 | e <- emmeans(a, "placement") 106 | c1 <- list(lr = c(-0.5, 0.5, 0), 107 | ln = c(-0.5, 0, 0.5)) 108 | b <- contrast(e, c1, adjust = "holm") |> 109 | broom::tidy() 110 | 111 | data.frame( 112 | p_all = a$anova_table$`Pr(>F)`[[1]], 113 | p_1 = b$adj.p.value[[1]], 114 | p_2 = b$adj.p.value[[2]] 115 | ) 116 | } 117 | 118 | s1_analyse(data) 119 | ``` 120 | 121 | Let's just replicate this 100 times so the simulation doesn't take too long to run at first. We can always increase it later after we've run some sense checks. 122 | 123 | ```{r} 124 | s1 <- NULL 125 | ``` 126 | 127 | Run the analysis on each data set. 128 | 129 | ```{r} 130 | s1_sim <- NULL 131 | ``` 132 | 133 | Calculating power is a little trickier here, as all three p-values need to be significant here to support the hypothesis. 134 | 135 | ```{r} 136 | s1_power <- NULL 137 | ``` 138 | 139 | Compare this value (`r s1_power`) with the value in the paper (0.4582). 140 | 141 | ### Study 3 142 | 143 | Now you can use the pattern from Study 1 to analyse the data for Study 3. We'll start with the repeated data set. 144 | 145 | ```{r} 146 | mu = c(left = 1428.24, right = 1308.66, none = 1436.79) 147 | sd = c(left = 377.02, right = 420.14, none = 378.47) 148 | n = c(left = 85, right = 86, none = 81) 149 | 150 | s3 <- NULL 151 | ``` 152 | 153 | These data were collected in the Hebrew language, which reads right to left, so the paired contrasts will be different. 154 | 155 | ```{r} 156 | s3_analyse <- function(data) { 157 | 158 | } 159 | ``` 160 | 161 | Run the analysis on each data set. 162 | 163 | ```{r} 164 | s3_sim <- NULL 165 | ``` 166 | 167 | ```{r} 168 | s3_power <- NULL 169 | ``` 170 | 171 | Compare this value (`r s3_power`) with the value in the paper (0.3626). 172 | 173 | 174 | ### Study S1 175 | 176 | Now you can use the pattern from Study 2 to analyse the data for Study S1. You can even reuse the analysis function `s2_analyse()`! 177 | 178 | ```{r} 179 | mu = c(left = 185.94, right = 215.73) 180 | sd = c(left = 93.92, right = 95.33) 181 | n = c(left = 99, right = 77) 182 | 183 | ss1 <- NULL 184 | ``` 185 | 186 | ```{r} 187 | ss1_sim <- NULL 188 | ``` 189 | 190 | 191 | ```{r} 192 | ss1_power <- NULL 193 | ``` 194 | 195 | 196 | ### Study S2 197 | 198 | Now you can use the pattern from Study 1 to analyse the data for Study S2. You can even reuse the analysis function `s1_analyse()`! 199 | 200 | ```{r} 201 | mu = c(left = 1182.15, right = 1302.23, none = 1373.74) 202 | sd = c(left = 477.60, right = 434.41, none = 475.77) 203 | n = c(left = 139, right = 141, none = 151) 204 | 205 | ss2 <- NULL 206 | ``` 207 | 208 | ```{r} 209 | ss2_sim <- NULL 210 | ``` 211 | 212 | ```{r} 213 | ss2_power <- NULL 214 | ``` 215 | 216 | ### Study S3 217 | 218 | Now you can use the pattern from Study 1 to analyse the data for Study S3. 219 | 220 | ```{r} 221 | mu = c(left = 1302.03, right = 1373.15, none = 1404.35) 222 | sd = c(left = 480.02, right = 442.49, none = 422.03) 223 | n = c(left = 336, right = 337, none = 333) 224 | 225 | ss3 <- NULL 226 | ``` 227 | 228 | ```{r} 229 | ss3_sim <- NULL 230 | ``` 231 | 232 | ```{r} 233 | ss3_power <- NULL 234 | ``` 235 | 236 | ## Conclusion 237 | 238 | Now that you've calculated power for each of the 6 studies, just multiply the 6 power values together to get the probability that all 6 studies will be significant. 239 | 240 | 241 | ```{r} 242 | power_table <- tribble( 243 | ~study, ~power_ft, ~ power_my, 244 | "1", 0.4582, s1_power, 245 | "2", 0.5426, s2_power, 246 | "3", 0.3626, s3_power, 247 | "S1", 0.5358, ss1_power, 248 | "S2", 0.5667, ss2_power, 249 | "S3", 0.4953, ss3_power 250 | ) 251 | 252 | power_table 253 | ``` 254 | 255 | The `reduce()` function from {purrr} applies a function sequentially over a vector, so can give up the product of all the values in the power columns. 256 | 257 | ```{r} 258 | prob_ft <- purrr::reduce(power_table$power_ft, `*`) 259 | prob_my <- purrr::reduce(power_table$power_my, `*`) 260 | ``` 261 | 262 | The Francis & Thunell paper showed a `r prob_ft` probability of getting 6 of 6 studies significant. Our re-simulation showed a `r prob_my` probability. 263 | 264 | -------------------------------------------------------------------------------- /inst/stubs/faux-stub.Rmd: -------------------------------------------------------------------------------- 1 | --- 2 | title: "Intro to Faux" 3 | output: 4 | html_document: 5 | df_print: paged 6 | toc: true 7 | toc_float: true 8 | --- 9 | 10 | 11 | ```{r, include = FALSE} 12 | # control the appearance of the knitted result 13 | knitr::opts_chunk$set( 14 | collapse = TRUE, 15 | out.width = "100%", 16 | fig.width = 5, 17 | fig.height = 3, 18 | dpi = 144 19 | ) 20 | ``` 21 | 22 | 23 | In this tutorial, we'll learn how to simulate data for factorial designs using {faux}. There are more extensive examples at . 24 | 25 | ## Setup 26 | 27 | We'll be using 4 packages in this tutorial. 28 | 29 | ```{r libs, message=FALSE} 30 | library(tidyverse) # for data wrangling 31 | library(faux) # for simulation 32 | library(broom) # for tidy analysis results 33 | library(afex) # for ANOVA 34 | 35 | set.seed(8675309) # Jenny, I've got your number 36 | ``` 37 | 38 | A seed makes randomness reproducible. Run the following code several times. Change the seed to your favourite integer. If the seed is the same, the random numbers after it will be the same, as long as the code is always executed in the same order. 39 | 40 | ```{r} 41 | set.seed(0) 42 | rnorm(1) 43 | ``` 44 | 45 | ## Normal 46 | 47 | Let's start with a normal distribution using the base R function `rnorm()`, which returns `n` values from a normal distribution with a mean of 0 and a standard deviation of 1. 48 | 49 | ```{r} 50 | rnorm(n = 10) 51 | ``` 52 | 53 | You can change the mean and SD. Simulate a lot of values (1e5 == 100,000), save them in a variable, and visualise them with `hist()`. 54 | 55 | ```{r} 56 | x <- rnorm(1e5, mean = 30, sd = 5) 57 | 58 | hist(x) 59 | ``` 60 | 61 | ## Multivariate normal 62 | 63 | But how do you create correlated values? You can do this with `MASS::mvrnorm()`, but you need to construct the `Sigma` argument yourself from the correlation matrix and the standard deviations of the populations, and then you need to turn the resulting matrix into a data frame for many use cases. This isn't very difficult, but can be tedious with larger numbers of variables. 64 | 65 | ```{r} 66 | n = 1e5 # this is a large number to demonstrate that the result is as expected 67 | mu = c(A = 1, B = 2, C = 3) 68 | sd = c(0.5, 1, 1.5) 69 | r = c(0, .25, .5) 70 | 71 | cor_mat <- matrix(c(1, r[1], r[2], 72 | r[1], 1, r[3], 73 | r[2], r[3], 1), 74 | nrow = 3) 75 | Sigma <- (sd %*% t(sd)) * cor_mat 76 | vars <- MASS::mvrnorm(n, mu, Sigma) |> as.data.frame() 77 | 78 | cor(vars) |> round(2) 79 | ``` 80 | 81 | ### rnorm_multi 82 | 83 | In faux, you can create sets of correlated normally distributed values using `rnorm_multi()`. 84 | 85 | ```{r} 86 | dat3 <- rnorm_multi( 87 | n = 50, 88 | mu = c(A = 1, B = 2, C = 3), 89 | sd = c(0.5, 1, 1.5), 90 | r = c(0, .25, .5) 91 | ) 92 | ``` 93 | 94 | The function `get_params()` gives you a quick way to see the means, SDs and correlations in the simulated data set to make sure you set the parameters correctly. 95 | 96 | ```{r} 97 | get_params(dat3) 98 | ``` 99 | 100 | If you set `empirical` to `TRUE`, the values you set will be the **sample** parameters, not the **population** parameters. This isn't usually what you want for a simulation, but can be useful to check you set the parameters correctly. 101 | 102 | ```{r} 103 | dat3 <- rnorm_multi( 104 | n = 50, 105 | mu = c(A = 1, B = 2, C = 3), 106 | sd = c(0.5, 1, 1.5), 107 | r = c(0, .25, .5), 108 | empirical = TRUE 109 | ) 110 | 111 | get_params(dat3) 112 | ``` 113 | 114 | 115 | ### Setting r 116 | 117 | You can set the `r` argument for correlations in a few different ways. 118 | 119 | If all correlations have the same value, just set r equal to a single number. 120 | 121 | ```{r} 122 | # all correlations the same value 123 | rho_same <- rnorm_multi(50, 4, r = .5, empirical = TRUE) 124 | get_params(rho_same) 125 | ``` 126 | 127 | You can set `r` to a vector or matrix of the full correlation matrix. This is convenient when you're getting the values from an existing dataset, where you can just use the output of the `cor()` function. 128 | 129 | ```{r} 130 | rho <- cor(iris[1:4]) 131 | round(rho, 2) 132 | ``` 133 | 134 | Notice how, since we didn't specify the names of the 4 variables anywhere else, `rnorm_multi()` will take them from the named correlation matrix. 135 | 136 | ```{r} 137 | rho_cormat <- rnorm_multi(50, 4, r = rho, empirical = TRUE) 138 | get_params(rho_cormat) 139 | ``` 140 | 141 | Alternatively, you can just specify the values from the upper right triangle of a correlation matrix. This might be easier if you're reading the values out of a paper. 142 | 143 | ```{r} 144 | # upper right triangle 145 | # X2 X3 X4 146 | rho <- c(0.5, 0.4, 0.3, # X1 147 | 0.2, 0.1, # X2 148 | 0.0) # X3 149 | 150 | rho_urt <- rnorm_multi(50, 4, r = rho, empirical = TRUE) 151 | get_params(rho_urt) 152 | ``` 153 | 154 | 155 | ## Factorial Designs 156 | 157 | You can use `rnorm_multi()` to simulate data for each between-subjects cell of a factorial design and manually combine the tables, but faux has a function that better maps onto how we usually think and teach about factorial designs. 158 | 159 | The default design is 100 observations of one variable (named `y`) with a mean of 0 and SD of 1. Unless you set `plot = FALSE` or run `faux_options(plot = FALSE)`, this function will show you a plot of your design so you can check that it looks like you expect. 160 | 161 | ```{r} 162 | simdat1 <- sim_design() 163 | ``` 164 | 165 | 166 | ### Factors 167 | 168 | Use named lists to set the names and levels of `within` and `between` subject factors. 169 | 170 | ```{r} 171 | pettime <- sim_design( 172 | between = list(pet = c("cat", "dog", "ferret")), 173 | within = list(time = c("pre", "post")) 174 | ) 175 | ``` 176 | 177 | You can set `mu` and `sd` with unnamed vectors, but getting the order right can take some trial and error. 178 | 179 | ```{r} 180 | pettime <- sim_design( 181 | between = list(pet = c("cat", "dog", "ferret")), 182 | within = list(time = c("pre", "post")), 183 | mu = 1:6 184 | ) 185 | ``` 186 | 187 | You can set values with a named vector for a single type of factor. The values do not have to be in the right order if they're named. 188 | 189 | ```{r} 190 | pettime <- sim_design( 191 | between = list(pet = c("cat", "dog", "ferret")), 192 | within = list(time = c("pre", "post")), 193 | mu = c(cat = 1, ferret = 5, dog = 3), 194 | sd = c(pre = 1, post = 2) 195 | ) 196 | ``` 197 | 198 | Or use a data frame for within- and between-subject factors. 199 | 200 | ```{r} 201 | pettime <- sim_design( 202 | between = list(pet = c("cat", "dog", "ferret")), 203 | within = list(time = c("pre", "post")), 204 | mu = data.frame( 205 | pre = c(1, 3, 5), 206 | post = c(2, 4, 6), 207 | row.names = c("cat", "dog", "ferret") 208 | ) 209 | ) 210 | ``` 211 | 212 | If you have within-subject factors, set the correlations for each between-subject cell like this. 213 | 214 | ```{r} 215 | pettime <- sim_design( 216 | between = list(pet = c("cat", "dog", "ferret")), 217 | within = list(time = c("pre", "post")), 218 | r = list(cat = 0.5, 219 | dog = 0.25, 220 | ferret = 0), 221 | empirical = TRUE, 222 | plot = FALSE 223 | ) 224 | 225 | get_params(pettime) 226 | ``` 227 | 228 | You can also change the name of the `dv` and `id` columns and output the data in long format. If you do this, you also need to tell `get_params()` what columns contain the between- and within-subject factors, the dv, and the id. 229 | 230 | ```{r} 231 | dat_long <- sim_design( 232 | between = list(pet = c("cat", "dog", "ferret")), 233 | within = list(time = c("pre", "post")), 234 | id = "subj_id", 235 | dv = "score", 236 | long = TRUE, 237 | plot = FALSE 238 | ) 239 | 240 | get_params(dat_long, digits = 3) 241 | ``` 242 | 243 | ### Multiple Factors 244 | 245 | Set more than one within-or between-subject factor like this: 246 | 247 | ```{r} 248 | dat_multi <- sim_design( 249 | between = list(pet = c("cat", "dog", "ferret"), 250 | country = c("UK", "NL")), 251 | within = list(time = c("pre", "post"), 252 | condition = c("ctl", "exp")), 253 | mu = data.frame( 254 | cat_UK = 1:4, 255 | cat_NL = 5:8, 256 | dog_UK = 9:12, 257 | dog_NL = 13:16, 258 | ferret_UK = 17:20, 259 | ferret_NL = 21:24, 260 | row.names = c("pre_ctl", "pre_exp", "post_ctl", "post_exp") 261 | ) 262 | ) 263 | ``` 264 | 265 | 266 | Because faux uses an underscore for the separator, you have to set the `sep` argument to something different if you want to use underscores in your variable names (or set the separator globally with `faux_options`). 267 | 268 | ```{r} 269 | # faux_options(sep = ".") 270 | 271 | dat_multi <- sim_design( 272 | between = list(pet = c("cat", "dog", "ferret"), 273 | country = c("Glasgow_UK", "Rotterdam_NL")), 274 | within = list(time = c("pre", "post"), 275 | condition = c("ctl", "exp")), 276 | mu = data.frame( 277 | cat.Glasgow_UK = 1:4, 278 | cat.Rotterdam_NL = 5:8, 279 | dog.Glasgow_UK = 9:12, 280 | dog.Rotterdam_NL = 13:16, 281 | ferret.Glasgow_UK = 17:20, 282 | ferret.Rotterdam_NL = 21:24, 283 | row.names = c("pre.ctl", "pre.exp", "post.ctl", "post.exp") 284 | ), 285 | sep = "." 286 | ) 287 | ``` 288 | 289 | ### Anonymous Factors 290 | 291 | If you need to make a quick demo, you can set factors anonymously with integer vectors. For example, the following code makes 3B\*2B\*2W mixed design. 292 | 293 | ```{r} 294 | dat_anon <- sim_design( 295 | n = 50, 296 | between = c(3, 2), 297 | within = 2, 298 | mu = 1:12 299 | ) 300 | ``` 301 | 302 | Faux has a quick plotting function for visualising data made with faux. The plot created by `sim_design()` shows the *design*, while this function shows the simulated *data*. 303 | 304 | ```{r} 305 | plot(dat_anon) 306 | ``` 307 | 308 | You can change the order of plotting and the types of geoms plotted. This takes a little trial and error, so this function will probably be refined in later versions. 309 | 310 | ```{r} 311 | plot(dat_anon, "B1", "B2", "W1", geoms = c("violin", "pointrangeSD")) 312 | ``` 313 | 314 | 315 | 316 | ## Replications 317 | 318 | You often want to simulate data repeatedly to do things like calculate power. The `sim_design()` function has a lot of overhead for checking if a design makes sense and if the correlation matrix is possible, so you can speed up the creation of multiple datasets with the same design using the `rep` argument. This will give you a nested data frame with each dataset in the `data` column. 319 | 320 | ```{r} 321 | dat_rep <- sim_design( 322 | within = 2, 323 | n = 20, 324 | mu = c(0, 0.25), 325 | rep = 5, 326 | plot = FALSE 327 | ) 328 | ``` 329 | 330 | ### Analyse each replicate 331 | 332 | You can run analyses on the nested data by wrapping your analysis code in a function then using `map()` to run the analysis on each data set and `unnest()` to expand the results into a data table. 333 | 334 | ```{r} 335 | # define function 336 | analyse <- function(data) { 337 | t.test(data$W1a, data$W1b, paired = TRUE) %>% broom::tidy() 338 | } 339 | 340 | # get one test data set 341 | data <- dat_rep$data[[1]] 342 | 343 | # check function returns what you want 344 | analyse(data) 345 | ``` 346 | 347 | 348 | ```{r} 349 | # run the function on each data set 350 | dat_rep |> 351 | mutate(analysis = map(data, analyse)) |> 352 | select(-data) |> 353 | unnest(analysis) 354 | ``` 355 | 356 | ### ANOVA 357 | 358 | Use the same pattern to run an ANOVA on a version of the `pettime` dataset. 359 | 360 | First, simulate 100 datasets in long format. These data will have small main effects of pet and time, but no interaction. 361 | 362 | ```{r} 363 | pettime100 <- sim_design( 364 | between = list(pet = c("cat", "dog")), 365 | within = list(time = c("pre", "post")), 366 | n = c(cat = 50, dog = 40), 367 | mu = data.frame( 368 | pre = c(1, 1.2), 369 | post = c(1.2, 1.4), 370 | row.names = c("cat", "dog") 371 | ), 372 | sd = 1, 373 | id = "pet_id", 374 | dv = "score", 375 | r = 0.5, 376 | long = TRUE, 377 | rep = 100 378 | ) 379 | ``` 380 | 381 | Then set up your analysis. We'll use the `aov_ez()` function from the {afex} package because its arguments match those of `sim_design()`. There's a little setup to run first to get rid of annoying messages and make this run faster by omitting calculations we won't need. 382 | 383 | ```{r} 384 | afex::set_sum_contrasts() # avoids annoying afex message 385 | afex_options(include_aov = FALSE) # runs faster 386 | afex_options(es_aov = "pes") # changes effect size measure to partial eta squared 387 | ``` 388 | 389 | This custom function takes the data frame as input and runs our ANOVA on it. The code at the end just cleans up the resulting table a bit. 390 | 391 | ```{r} 392 | analyse <- function(data) { 393 | a <- afex::aov_ez( 394 | id = "pet_id", 395 | dv = "score", 396 | between = "pet", 397 | within = "time", 398 | data = data 399 | ) 400 | # return anova_table for GG-corrected DF 401 | as_tibble(a$anova_table, rownames = "term") |> 402 | mutate(term = factor(term, levels = term)) |> # keeps terms in order 403 | rename(p.value = `Pr(>F)`) # fixes annoying p.value name 404 | } 405 | ``` 406 | 407 | Test the analysis code on the first simulated data frame. 408 | 409 | ```{r} 410 | analyse( pettime100$data[[1]] ) 411 | ``` 412 | 413 | 414 | Use the same code we used in the first example to make a table of the results of each analysis: 415 | 416 | ```{r} 417 | pettime_sim <- pettime100 |> 418 | mutate(analysis = map(data, analyse)) |> 419 | select(-data) |> 420 | unnest(analysis) 421 | ``` 422 | 423 | ```{r, echo = FALSE} 424 | # show the first 6 rows 425 | head(pettime_sim) |> 426 | mutate(across(5:8, \(x) round(x, 3))) 427 | ``` 428 | 429 | Then you can summarise the data to calculate things like power for each effect or mean effect size. 430 | 431 | ```{r} 432 | pettime_sim |> 433 | group_by(term) |> 434 | summarise(power = mean(p.value < 0.05), 435 | mean_pes = mean(pes) |> round(3), 436 | .groups = "drop") 437 | ``` 438 | 439 | The power for the between-subjects effect of pet is smaller than for the within-subjects effect of time. What happens if you reduce the correlation between pre and post? 440 | 441 | ## Non-normal Distributions 442 | 443 | The newest version of faux has a new function for simulating non-normal distributions using the NORTA method (NORmal To Anything). The `dist` argument lists the variables with their distribution names (e.g., "norm", "pois", unif", "truncnorm", or anything that has an "rdist" function). The `params` argument lists the distribution function argument values for each variable (e.g., arguments to `rnorm`, `rpois`, `runif`, `rtruncnorm`). 444 | 445 | This function simulates multivariate non-normal distributions by using simulation to work out the correlations for a multivariate normal distribution that will produce the desired correlations after the normal distributions are converted to the desired distributions. This simulation can take a while if you have several variables and should warn you if you're requesting an impossible combination (but is still an experimental function, so let Lisa know if you have any problems). 446 | 447 | ```{r} 448 | dat_norta <- rmulti( 449 | n = 1000, 450 | dist = c(U = "unif", 451 | T = "truncnorm", 452 | L = "likert"), 453 | params = list( 454 | U = list(min = 0, max = 10), 455 | T = list(a = 1, b = 7, mean = 3.5, sd = 2.1), 456 | L = list(prob = c(`much less` = .10, 457 | `less` = .20, 458 | `equal` = .35, 459 | `more` = .25, 460 | `much more` = .10)) 461 | ), 462 | r = c(-0.5, 0, 0.5) 463 | ) 464 | ``` 465 | 466 | The "likert" type is a set of distribution functions provided by faux to make creating Likert scale variables easier (see `?rlikert`). You may need to convert Likert-scale variables to numbers before analysis or calculating descriptives. 467 | 468 | ```{r} 469 | # convert likert-scale variable to integer 470 | dat_norta$L <- as.integer(dat_norta$L) 471 | 472 | get_params(dat_norta) 473 | ``` 474 | 475 | 476 | 477 | ## Exercises 478 | 479 | ### Multivariate normal 480 | 481 | Sample 40 values of three variables named `J`, `K` and `L` from a population with means of 10, 20 and 30, and SDs of 5. `J` and `K` are correlated 0.5, `J` and `L` are correlated 0.25, and `K` and `L` are not correlated. 482 | 483 | ```{r} 484 | 485 | ``` 486 | 487 | ### From existing data 488 | 489 | Using the data from the built-in dataset `attitude`, simulate a new set of 20 observations drawn from a population with the same means, SDs and correlations for each column as the original data. 490 | 491 | ```{r} 492 | 493 | ``` 494 | 495 | 496 | ### 2b 497 | 498 | Create a dataset with a between-subject factor of "pet" having two levels, "cat", and "dog". The DV is "happiness" score. There are 20 cat-owners with a mean happiness score of 10 (SD = 3) and there are 30 dog-owners with a mean happiness score of 11 (SD = 3). 499 | 500 | ```{r} 501 | 502 | ``` 503 | 504 | ### 3w 505 | 506 | Create a dataset of 20 observations with 1 within-subject variable ("condition") having 3 levels ("A", "B", "C") with means of 10, 20 and 30 and SD of 5. The correlations between each level have r = 0.4. The dataset should look like this: 507 | 508 | | id | condition | score | 509 | |:---|:----------|------:| 510 | |S01 | A | 9.17 | 511 | |... | ... | ... | 512 | |S20 | A | 11.57 | 513 | |S01 | B | 18.44 | 514 | |... | ... | ... | 515 | |S20 | B | 20.04 | 516 | |S01 | C | 35.11 | 517 | |... | ... | ... | 518 | |S20 | C | 29.16 | 519 | 520 | ```{r} 521 | 522 | ``` 523 | 524 | ### 2w*2w 525 | 526 | Create a dataset with 50 subjects and 2 within-subject variables ("W1" and "W2") each having 2 levels. The mean for all cells is 10 and the SD is 2. The correlations look like this: 527 | 528 | | | W1a_W2a | W1a_W2b | W1b_W2a | W1b_W2b | 529 | |:--------|------:|------:|------:|------:| 530 | | W1a_W2a | 1.0 | 0.5 | 0.5 | 0.2 | 531 | | W1a_W2b | 0.5 | 1.0 | 0.2 | 0.5 | 532 | | W1b_W2a | 0.5 | 0.2 | 1.0 | 0.5 | 533 | | W1b_W2b | 0.2 | 0.5 | 0.5 | 1.0 | 534 | 535 | 536 | ```{r} 537 | 538 | ``` 539 | 540 | ### 2w*3b 541 | 542 | Create a dataset with a between-subject factor of "pet" having 3 levels ("cat", "dog", and "ferret") and a within-subject factor of "time" having 2 levels ("pre" and "post"). The N in each group should be 10. Means are: 543 | 544 | * cats: pre = 10, post = 12 545 | * dogs: pre = 14, post = 16 546 | * ferrets: pre = 18, post = 20 547 | 548 | SDs are all 5 and within-cell correlations are all 0.25. 549 | 550 | ```{r} 551 | 552 | ``` 553 | 554 | ### Replications 555 | 556 | Create 5 datasets with a 2b*2b design, 30 participants in each cell. Each cell's mean should be 0, except B1a:B2a, which should be 0.5. The SD should be 1. Make the resulting data in long format. 557 | 558 | ```{r} 559 | 560 | ``` 561 | 562 | ### Power 563 | 564 | Simulate 100 datasets like the one above and use `lm()` or `afex::aov_ez()` to look at the interaction between B1 and B2. What is the power of this design? 565 | 566 | ```{r} 567 | 568 | ``` 569 | 570 | -------------------------------------------------------------------------------- /inst/stubs/fixed-stub.Rmd: -------------------------------------------------------------------------------- 1 | --- 2 | title: "Fixed Effects" 3 | output: 4 | html_document: 5 | df_print: kable 6 | --- 7 | 8 | ```{r, include = FALSE} 9 | knitr::opts_chunk$set( 10 | collapse = TRUE, 11 | out.width = "100%", 12 | fig.width = 5, 13 | fig.height = 3, 14 | dpi = 144 15 | ) 16 | ``` 17 | 18 | ```{r libs, message=FALSE} 19 | library(tidyverse) 20 | library(faux) 21 | library(afex) # for anova and lmer 22 | library(broom) 23 | library(broom.mixed) # to make tidy tables of lmer output 24 | 25 | theme_set(theme_minimal(base_size = 14)) 26 | ``` 27 | 28 | 29 | ## Simulation functions 30 | 31 | The functions below are commonly used when you're setting up a simulated dataset. 32 | 33 | ### Repeating 34 | 35 | The function `rep()` lets you repeat the first argument a number of times. 36 | 37 | Use `rep()` to create a vector of alternating `"A"` and `"B"` values of length 24. 38 | 39 | ```{r rep1-times} 40 | rep(c("A", "B"), times = 12) 41 | ``` 42 | 43 | If the second argument is a vector that is the same length as the first argument, each element in the first vector is repeated that many times. Use `rep()` to create a vector of 11 `"A"` values followed by 3 `"B"` values. 44 | 45 | ```{r rep-vector} 46 | rep(c("A", "B"), c(11, 3)) 47 | ``` 48 | 49 | You can repeat each element of the vector a specified number of times using the `each` argument, Use `rep()` to create a vector of 12 `"A"` values followed by 12 `"B"` values. 50 | 51 | ```{r rep-each} 52 | rep(c("A", "B"), each = 12) 53 | ``` 54 | 55 | What do you think will happen if you set `times` to 3 and `each` to 2? 56 | 57 | ```{r rep-times-each} 58 | rep(c("A", "B"), times = 3, each = 2) 59 | ``` 60 | 61 | 62 | ### Sequences 63 | 64 | The function `seq()` is useful for generating a sequence of numbers with some pattern. 65 | 66 | Use `seq()` to create a vector of the integers 0 to 10. 67 | 68 | ```{r seq1-10} 69 | seq(0, 10) 70 | ``` 71 | 72 | You can set the `by` argument to count by numbers other than 1 (the default). Use `seq()` to create a vector of the numbers 0 to 100 by 10s. 73 | 74 | ```{r seq-by} 75 | seq(0, 100, by = 10) 76 | ``` 77 | 78 | The argument `length.out` is useful if you know how many steps you want to divide something into. Use `seq()` to create a vector that starts with 0, ends with 100, and has 12 equally spaced steps (hint: how many numbers would be in a vector with 2 *steps*?). 79 | 80 | ```{r seq-length-out} 81 | seq(0, 100, length.out = 13) 82 | ``` 83 | 84 | ### Uniform Distribution 85 | 86 | The uniform distribution is the simplest distribution. All numbers in the range have an equal probability of being sampled. Use `runif()` to sample from a continuous uniform distribution. 87 | 88 | ```{r runif} 89 | runif(n = 10, min = 0, max = 1) 90 | ``` 91 | 92 | 93 | Pipe the result to `hist()` to make a quick histogram of your simulated data. 94 | 95 | ```{r runif-hist} 96 | runif(100000, min = 0, max = 1) %>% hist() 97 | ``` 98 | 99 | ### Discrete Distribution 100 | 101 | You can use `sample()` to simulate events like rolling dice or choosing from a deck of cards. The code below simulates rolling a 6-sided die 10000 times. We set `replace` to `TRUE` so that each event is independent. See what happens if you set `replace` to `FALSE`. 102 | 103 | ```{r sample-replace, fig.cap = "Distribution of dice rolls."} 104 | rolls <- sample(1:6, 10000, replace = TRUE) 105 | 106 | # plot the results 107 | as.factor(rolls) %>% plot() 108 | ``` 109 | 110 | You can also use sample to sample from a list of named outcomes. 111 | 112 | ```{r sample-list} 113 | pet_types <- c("cat", "dog", "ferret", "bird", "fish") 114 | sample(pet_types, 10, replace = TRUE) 115 | ``` 116 | 117 | Ferrets, while the best pet, are a much less common pet than cats and dogs, so our sample isn't very realistic. You can set the probabilities of each item in the list with the `prob` argument. 118 | 119 | ```{r sample-prob} 120 | pet_types <- c("cat", "dog", "ferret", "bird", "fish") 121 | pet_prob <- c(0.3, 0.4, 0.1, 0.1, 0.1) 122 | pet_data <- sample(pet_types, 100, replace = TRUE, prob = pet_prob) 123 | 124 | as.factor(pet_data) %>% plot() 125 | ``` 126 | 127 | 128 | ### Binomial Distribution 129 | 130 | The `rbinom` function will generate a random binomial distribution. 131 | 132 | * `n` = number of observations 133 | * `size` = number of trials 134 | * `prob` = probability of success on each trial 135 | 136 | Coin flips are a typical example of a binomial distribution, where we can assign heads to 1 and tails to 0. 137 | 138 | ```{r rbinom-fair} 139 | # 20 individual coin flips of a fair coin 140 | rbinom(20, 1, 0.5) 141 | ``` 142 | 143 | 144 | ```{r rbinom-bias} 145 | # 20 individual coin flips of a baised (0.75) coin 146 | rbinom(20, 1, 0.75) 147 | ``` 148 | 149 | You can generate the total number of heads in 1 set of 20 coin flips by setting `size` to 20 and `n` to 1. 150 | 151 | ```{r rbinom-size} 152 | # 1 set of 20 fair coin flips 153 | rbinom(1, 20, 0.75) 154 | ``` 155 | 156 | You can generate more sets of 20 coin flips by increasing the `n`. 157 | 158 | ```{r rbinom-n} 159 | # 10 sets of 20 fair coin flips 160 | rbinom(10, 20, 0.5) 161 | ``` 162 | 163 | ### Normal Distribution 164 | 165 | We can simulate a normal distribution of size `n` if we know the `mean` and standard deviation (`sd`). 166 | 167 | ```{r rnorm} 168 | # 10 samples from a normal distribution with a mean of 0 and SD of 1 169 | rnorm(10, 0, 1) 170 | ``` 171 | 172 | A density plot is usually the best way to visualise this type of data. 173 | 174 | ```{r rnorm-plot} 175 | # 100 samples from a normal distribution with a mean of 10 and SD of 2 176 | dv <- rnorm(100, 10, 2) 177 | 178 | # use sample to get a random colour 179 | fill_colour <- sample(colours(), 1) 180 | 181 | ggplot() + 182 | geom_density(aes(dv), fill = fill_colour) + 183 | scale_x_continuous( 184 | limits = c(0,20), 185 | breaks = seq(0,20) 186 | ) 187 | ``` 188 | 189 | Run the simulation above several times, noting how the density plot changes. Try changing the values of `n`, `mean`, and `sd`. 190 | 191 | ## Independent samples 192 | 193 | Now we're ready to start simulating some data. Let's start with a simple independent-samples design where the variables are from a normal distribution. Each subject produces one score (in condition A or B). What we need to know about these scores is: 194 | 195 | * How many subjects are in each condition? 196 | * What are the score means? 197 | * What are the score variances (or SDs)? 198 | 199 | ### Parameters 200 | 201 | First, set parameters for these values. This way, you can use these variables wherever you need them in the rest of the code and you can easily change them. 202 | 203 | ```{r ind-vars} 204 | 205 | A_sub_n <- 50 206 | B_sub_n <- 50 207 | A_mean <- 10 208 | B_mean <- 11 209 | A_sd <- 2.5 210 | B_sd <- 2.5 211 | 212 | ``` 213 | 214 | ### Scores 215 | 216 | We can the generate the scores using the `rnorm()` function. 217 | 218 | ```{r ind-dat} 219 | A_scores <- rnorm(A_sub_n, A_mean, A_sd) 220 | B_scores <- rnorm(B_sub_n, B_mean, B_sd) 221 | ``` 222 | 223 | You can stop here and just analyse your simulated data with `t.test(A_scores, B_scores)`, but usually you want to get your simulated data into a data table that looks like what you might eventually import from a CSV file with your actual experimental data. 224 | 225 | ```{r ind-tibble} 226 | dat <- tibble( 227 | sub_condition = rep( c("A", "B"), c(A_sub_n, B_sub_n) ), 228 | score = c(A_scores, B_scores) 229 | ) 230 | ``` 231 | 232 | If you're simulating data for a script where you will eventually import data from a csv file, you can save these data to a csv file and then re-read them in, so when you get your real data, all you need to do is comment out the simulation steps. 233 | 234 | ```{r} 235 | # make a data directory if there isn't one already 236 | if (!dir.exists("data")) dir.create("data") 237 | 238 | # save your simulated data 239 | write_csv(dat, "data/sim-data-ind-samples.csv") 240 | 241 | # start your analysis here 242 | dat <- read_csv("data/sim-data-ind-samples.csv") 243 | 244 | ``` 245 | 246 | 247 | ### Check your data 248 | 249 | Always examine your simulated data after you generate it to make sure it looks like you want. 250 | 251 | ```{r ind-check} 252 | dat %>% 253 | group_by(sub_condition) %>% 254 | summarise(n = n() , 255 | mean = mean(score), 256 | sd = sd(score), 257 | .groups = "drop") 258 | ``` 259 | 260 | 261 | ### Analysis 262 | 263 | ```{r ind-test} 264 | t.test(score~sub_condition, dat) 265 | ``` 266 | 267 | ### Function 268 | 269 | You can wrap all this in a function so you can run it many times to do a power calculation. Put all your parameters as arguments to the function. 270 | 271 | ```{r ind-func} 272 | 273 | ind_sim <- function(A_sub_n, B_sub_n, 274 | A_mean, B_mean, 275 | A_sd, B_sd) { 276 | # simulate data for groups A and B 277 | A_scores <- rnorm(A_sub_n, A_mean, A_sd) 278 | B_scores <- rnorm(B_sub_n, B_mean, B_sd) 279 | 280 | # put the data into a table 281 | dat <- tibble( 282 | sub_condition = rep( c("A", "B"), c(A_sub_n, B_sub_n) ), 283 | score = c(A_scores, B_scores) 284 | ) 285 | 286 | # analyse the data 287 | t <- t.test(score~sub_condition, dat) 288 | 289 | # return a list of the values you care about 290 | # the double brackets ([[]]) get rid of the name of named numbers 291 | list( 292 | t = t$statistic[[1]], 293 | ci_lower = t$conf.int[[1]], 294 | ci_upper = t$conf.int[[2]], 295 | p = t$p.value[[1]], 296 | estimate = t$estimate[[1]] - t$estimate[[2]] 297 | ) 298 | } 299 | 300 | ``` 301 | 302 | Now run your new function with the values you used above. 303 | 304 | ```{r} 305 | # str() prints the resulting list in a shorter format 306 | ind_sim(50, 50, 10, 11, 2.5, 2.5) %>% str() 307 | ``` 308 | 309 | Now you can use this function to run many simulations. The function `map_df` from the `purrr` package (loaded with `tidyverse`) is one of many ways to run a function many times and organise the results into a table. 310 | 311 | ```{r} 312 | mysim <- map_df(1:1000, ~ind_sim(50, 50, 10, 11, 2.5, 2.5)) 313 | ``` 314 | 315 | Now you can graph the data from your simulations. 316 | 317 | ```{r sim-p-fig} 318 | # set boundary = 0 when plotting p-values 319 | ggplot(mysim, aes(p)) + 320 | geom_histogram(binwidth = 0.05, boundary = 0, 321 | fill = "white", colour = "black") 322 | ``` 323 | 324 | 325 | ```{r ind-sim-fig, fig.cap = "Distribution of results from simulated independent samples data"} 326 | mysim %>% 327 | gather(stat, value, t:estimate) %>% 328 | ggplot() + 329 | geom_density(aes(value, color = stat), show.legend = FALSE) + 330 | facet_wrap(~stat, scales = "free") 331 | ``` 332 | 333 | You can calculate power as the proportion of simulations on which the p-value was less than your alpha. 334 | 335 | ```{r} 336 | alpha <- 0.05 337 | power <- mean(mysim$p < alpha) 338 | power 339 | ``` 340 | 341 | 342 | 343 | ## Paired samples 344 | 345 | Now let's try a paired-samples design where the variables are from a normal distribution. Each subject produces two scores (in conditions A and B). What we need to know about these two scores is: 346 | 347 | * How many subjects? 348 | * What are the score means? 349 | * What are the score variances (or SDs)? 350 | * What is the correlation between the scores? 351 | 352 | ### Parameters {#paired-params} 353 | 354 | ```{r paired-vars} 355 | 356 | sub_n <- 100 357 | A_mean <- 10 358 | B_mean <- 11 359 | A_sd <- 2.5 360 | B_sd <- 2.5 361 | AB_r <- 0.5 362 | 363 | ``` 364 | 365 | 366 | ### Correlated Scores 367 | 368 | You can then use `rnorm_multi()` to generate a data table with simulated values for correlated scores: 369 | 370 | ```{r sim-design} 371 | dat <- faux::rnorm_multi( 372 | n = sub_n, 373 | vars = 2, 374 | r = AB_r, 375 | mu = c(A_mean, B_mean), 376 | sd = c(A_sd, B_sd), 377 | varnames = c("A", "B") 378 | ) 379 | ``` 380 | 381 | You can also do this using the `MASS::mvrnorm` function, but `faux::rnorm_multi` is easier when you have more variables to simulate. 382 | 383 | ```{r} 384 | # make the correlation matrix 385 | cormat <- matrix(c( 1, AB_r, 386 | AB_r, 1), 387 | nrow = 2, byrow = TRUE) 388 | 389 | # make a corresponding matrix of the variance 390 | # (multiply the SDs for each cell) 391 | varmat <- matrix(c(A_sd * A_sd, A_sd * B_sd, 392 | A_sd * B_sd, B_sd * B_sd), 393 | nrow = 2, byrow = TRUE) 394 | 395 | # create correlated variables with the specified parameters 396 | S <- MASS::mvrnorm(n = sub_n, 397 | mu = c(A_mean, B_mean), 398 | Sigma = cormat * varmat) 399 | dat <- data.frame( 400 | A = S[, 1], 401 | B = S[, 2] 402 | ) 403 | 404 | ``` 405 | 406 | 407 | ### Check your data 408 | 409 | Now check your data; `faux` has a function `get_params()` that gives you the correlation table, means, and SDs for each numeric column in a data table. 410 | 411 | ```{r paired-check} 412 | faux::get_params(dat) 413 | ``` 414 | 415 | ### Analysis 416 | 417 | ```{r paired-test} 418 | # paired-samples t-test 419 | t.test(dat$A, dat$B, paired = TRUE) 420 | ``` 421 | 422 | ### Function 423 | 424 | ```{r paired-func} 425 | 426 | paired_sim <- function(sub_n, A_mean, B_mean, A_sd, B_sd, AB_r) { 427 | 428 | dat <- faux::rnorm_multi( 429 | n = sub_n, 430 | vars = 2, 431 | r = AB_r, 432 | mu = c(A_mean, B_mean), 433 | sd = c(A_sd, B_sd), 434 | varnames = c("A", "B") 435 | ) 436 | t <- t.test(dat$A, dat$B, paired = TRUE) 437 | 438 | # return just the values you care about 439 | list( 440 | t = t$statistic[[1]], 441 | ci_lower = t$conf.int[[1]], 442 | ci_upper = t$conf.int[[2]], 443 | p = t$p.value[[1]], 444 | estimate = t$estimate[[1]] 445 | ) 446 | } 447 | 448 | ``` 449 | 450 | Run 1000 simulations and graph the results. 451 | 452 | ```{r} 453 | mysim_p <- map_df(1:1000, ~paired_sim(100, 10, 11, 2.5, 2.5, .5)) 454 | ``` 455 | 456 | ```{r pair-sim-fig, fig.cap = "Distribution of results from simulated paired samples data"} 457 | mysim_p %>% 458 | gather(stat, value, t:estimate) %>% 459 | ggplot() + 460 | geom_density(aes(value, color = stat), show.legend = FALSE) + 461 | facet_wrap(~stat, scales = "free") 462 | ``` 463 | 464 | ```{r} 465 | alpha <- 0.05 466 | power <- mean(mysim_p$p < alpha) 467 | power 468 | ``` 469 | 470 | 471 | ## Intercept model 472 | 473 | Now I'm going to show you a different way to simulate the same design. This might seem excessively complicated, but you will need this pattern when you start simulating data for mixed effects models. 474 | 475 | ### Parameters 476 | 477 | Remember, we used the following parameters to set up our simulation above: 478 | 479 | ```{r paired-vars2} 480 | sub_n <- 100 481 | A_mean <- 10 482 | B_mean <- 11 483 | A_sd <- 2.5 484 | B_sd <- 2.5 485 | AB_r <- 0.5 486 | ``` 487 | 488 | From these, we can calculate the grand intercept (the overall mean regardless of condition), and the effect of condition (the mean of B minus A). 489 | 490 | ```{r} 491 | grand_i <- (A_mean + B_mean)/2 492 | AB_effect <- B_mean - A_mean 493 | ``` 494 | 495 | We also need to think about variance a little differently. First, calculate the pooled variance as the mean of the variances for A and B (remember, variance is SD squared). 496 | 497 | ```{r} 498 | pooled_var <- (A_sd^2 + B_sd^2)/2 499 | ``` 500 | 501 | The variance of the subject intercepts is `r` times this pooled variance and the error variance is what is left over. We take the square root (`sqrt()`) to set the subject intercept and error SDs for simulation later. 502 | 503 | ```{r} 504 | sub_sd <- sqrt(pooled_var * AB_r) 505 | error_sd <- sqrt(pooled_var * (1-AB_r)) 506 | ``` 507 | 508 | 509 | ### Subject intercepts 510 | 511 | Now we use these variables to create a data table for our subjects. Each subject gets an ID and a **random intercept** (`sub_i`). The intercept is simulated from a random normal distribution with a mean of 0 and an SD of `sub_sd`. This represents how much higher or lower than the average score each subject tends to be (regardless of condition). 512 | 513 | ```{r} 514 | sub <- tibble( 515 | sub_id = 1:sub_n, 516 | sub_i = rnorm(sub_n, 0, sub_sd) 517 | ) 518 | ``` 519 | 520 | ### Observations 521 | 522 | Next, set up a table where each row represents one observation. We'll use one of my favourite functions for simulation: `crossing()`. This creates every possible combination of the listed factors (it works the same as `expand.grid()`, but the results are in a more intuitive order). Here, we're using it to create a row for each subject in each condition, since this is a fully within-subjects design. 523 | 524 | ```{r} 525 | obs <- crossing( 526 | sub_id = 1:sub_n, 527 | condition = c("A", "B") 528 | ) 529 | ``` 530 | 531 | ### Calculate the score 532 | 533 | Next, we join the subject table so each row has the information about the subject's random intercept and then calculate the score. I've done it in a few steps below for clarity. The score is just the sum of: 534 | 535 | * the overall mean (`grand_i`) 536 | * the subject-specific intercept (`sub_i`) 537 | * the effect (`effect`): the numeric code for condition (`condition.e`) multiplied by the effect of condition (`AB_effect`) 538 | * the error term (simulated from a normal distribution with mean of 0 and SD of `error_sd`) 539 | 540 | ```{r im-data} 541 | dat <- obs %>% 542 | left_join(sub, by = "sub_id") %>% 543 | mutate( 544 | condition.e = recode(condition, "A" = -0.5, "B" = 0.5), 545 | effect = AB_effect * condition.e, 546 | error = rnorm(nrow(.), 0, error_sd), 547 | score = grand_i + sub_i + effect + error 548 | ) 549 | ``` 550 | 551 | Use `get_params` to check the data. With data in long format, you need to specify the columns that contain the id, dv, and within-id variables. 552 | 553 | ```{r im-get-params} 554 | # check the data 555 | faux::get_params(dat, 556 | id = "sub_id", 557 | dv = "score", 558 | within = "condition") 559 | ``` 560 | 561 | You can use the following code to put the data table into a more familiar "wide" format. 562 | 563 | ```{r im-wide} 564 | dat_wide <- dat %>% 565 | select(sub_id, condition, score) %>% 566 | spread(condition, score) 567 | ``` 568 | 569 | ### Analyses 570 | 571 | You can analyse the data with a paired-samples t-test from the wide format: 572 | 573 | ```{r im-wide-t} 574 | # paired-samples t-test from dat_wide 575 | t.test(dat_wide$A, dat_wide$B, paired = TRUE) 576 | ``` 577 | 578 | Or in the long format: 579 | 580 | ```{r im-long-t} 581 | # paired-samples t-test from dat (long) 582 | t.test(score ~ condition, dat, paired = TRUE) 583 | ``` 584 | 585 | You can analyse the data with ANOVA using the `aov_4()` function from `afex`. (Notice how the F-value is the square of the t-value above.) 586 | 587 | ```{r im-afex} 588 | # anova using afex::aov_4 589 | aov <- afex::aov_4(score ~ (condition | sub_id), data = dat) 590 | 591 | aov$anova_table 592 | ``` 593 | 594 | 595 | You can even analyse the data with a mixed effects model using the `lmer` function (the `afex` version gives you p-values, but the `lme4` version does not). 596 | 597 | ```{r im-lmer} 598 | # mixed effect model using afex::lmer 599 | lmem <- afex::lmer(score ~ condition.e + (1 | sub_id), data = dat) 600 | 601 | # displays a tidy table of the fixed effects 602 | broom.mixed::tidy(lmem, effects = "fixed") 603 | ``` 604 | 605 | ## Simulate a dataset from an analysis 606 | 607 | Simulate a dataset from the parameters of an analysis. We'll use the built-in dataset `mtcars` to predict miles per gallon (`mpg`) from transmission type (`am`) and engine type (`vs`). 608 | 609 | ```{r} 610 | model <- lm(mpg ~ am * vs, data = mtcars) 611 | broom::tidy(model) 612 | ``` 613 | 614 | ### Simulate 615 | 616 | We can now simulate a dataset with 50 observations from each transmission type (`am`) and engine type (`vs`) combination, then use the model parameters to generate predicted values for `mpg`. 617 | 618 | ```{r} 619 | err_sd <- sigma(model) # SD of the error term from the model 620 | fx <- coefficients(model) # fixed effect coefficients 621 | 622 | sim_mtcars <- tibble( 623 | am = rep(c(0, 0, 1, 1), each = 50), 624 | vs = rep(c(0, 1, 0, 1), each = 50) 625 | ) %>% 626 | mutate(err = rnorm(200, 0, err_sd), 627 | mpg = fx[1] + 628 | fx["am"]*am + 629 | fx["vs"]*vs + 630 | fx["am:vs"]*am*vs + err) 631 | 632 | ``` 633 | 634 | Analyse the simulated data with `lm()` and output the results as a table using `broom::tidy()` 635 | 636 | ```{r} 637 | sim_model <- lm(mpg ~ am * vs, data = sim_mtcars) 638 | broom::tidy(sim_model) 639 | ``` 640 | 641 | ### Function 642 | 643 | ```{r} 644 | carsim <- function(n, b0, b_am, b_vs, b_am_vs, err_sd) { 645 | sim_mtcars <- tibble( 646 | am = rep(c(0, 0, 1, 1), each = n), 647 | vs = rep(c(0, 1, 0, 1), each = n) 648 | ) %>% 649 | mutate(err = rnorm(n*4, 0, err_sd), 650 | mpg = b0 + b_am*am + b_vs*vs + b_am_vs*am*vs + err) 651 | 652 | sim_model <- lm(mpg ~ am * vs, data = sim_mtcars) 653 | broom::tidy(sim_model) 654 | } 655 | ``` 656 | 657 | 658 | Run the function with the values from the original model, but cut the fixed effect sizes in half. 659 | 660 | ```{r} 661 | err_sd <- sigma(model) 662 | fx2 <- coefficients(model)/2 663 | 664 | carsim(50, fx2[1], fx2[2], fx2[3], fx2[4], err_sd) 665 | ``` 666 | 667 | Repeat this 100 time and calculate power for each effect. 668 | 669 | ```{r} 670 | simstats <- map_df(1:100, ~carsim(50, fx2[1], fx2[2], fx2[3], fx2[4], err_sd)) 671 | 672 | simstats %>% 673 | group_by(term) %>% 674 | summarise(power = mean(p.value < .05), .groups = "drop") 675 | ``` 676 | 677 | ## Exercises 678 | 679 | Using the dataset below, predict `moral` disgust from the interaction between `pathogen` and `sexual` disgust using `lm()`. 680 | 681 | ```{r} 682 | disgust <- read_csv("https://psyteachr.github.io/msc-data-skills/data/disgust_scores.csv") 683 | ``` 684 | 685 | 686 | ```{r} 687 | 688 | ``` 689 | 690 | Simulate a new dataset of 100 people with a similar pathogen and sexual disgust distribution to the original dataset. Remember that these are likely to be correlated and that scores can only range from 0 to 6. (Hint: look at the help for `norm2trunc`) 691 | 692 | ```{r} 693 | 694 | ``` 695 | 696 | Write a function to simulate data, analyse it, and return a table of results. Make sure you can vary the important parameters using arguments. 697 | 698 | ```{r} 699 | 700 | ``` 701 | 702 | 703 | Calculate power for the same fixed effects as in the original analysis. Adjust the N until the dsign has around .80 power to detect a main effect of pathogen disgust. 704 | 705 | ```{r} 706 | 707 | ``` 708 | 709 | -------------------------------------------------------------------------------- /man/exercise.Rd: -------------------------------------------------------------------------------- 1 | % Generated by roxygen2: do not edit by hand 2 | % Please edit documentation in R/exercises.R 3 | \name{exercise} 4 | \alias{exercise} 5 | \title{Get an exercise} 6 | \usage{ 7 | exercise(name = c("faux", "fixed", "mixed", "calories"), filename = NULL) 8 | } 9 | \arguments{ 10 | \item{name}{The name of the exercise} 11 | 12 | \item{filename}{What filename you want to save (defaults to the name of the exercise in the working directory)} 13 | } 14 | \value{ 15 | Saves a file to the working directory (or path from filename) 16 | } 17 | \description{ 18 | Get an exercise 19 | } 20 | \examples{ 21 | \dontrun{ 22 | exercise("faux") # get exercise for the faux workshop 23 | exercise("fixed", "exercises/fixed.Rmd") # save into exercises directory 24 | } 25 | } 26 | -------------------------------------------------------------------------------- /man/figures/logo.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/man/figures/logo.png -------------------------------------------------------------------------------- /pkgdown/favicon/apple-touch-icon-120x120.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/pkgdown/favicon/apple-touch-icon-120x120.png -------------------------------------------------------------------------------- /pkgdown/favicon/apple-touch-icon-152x152.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/pkgdown/favicon/apple-touch-icon-152x152.png -------------------------------------------------------------------------------- /pkgdown/favicon/apple-touch-icon-180x180.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/pkgdown/favicon/apple-touch-icon-180x180.png -------------------------------------------------------------------------------- /pkgdown/favicon/apple-touch-icon-60x60.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/pkgdown/favicon/apple-touch-icon-60x60.png -------------------------------------------------------------------------------- /pkgdown/favicon/apple-touch-icon-76x76.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/pkgdown/favicon/apple-touch-icon-76x76.png -------------------------------------------------------------------------------- /pkgdown/favicon/apple-touch-icon.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/pkgdown/favicon/apple-touch-icon.png -------------------------------------------------------------------------------- /pkgdown/favicon/favicon-16x16.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/pkgdown/favicon/favicon-16x16.png -------------------------------------------------------------------------------- /pkgdown/favicon/favicon-32x32.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/pkgdown/favicon/favicon-32x32.png -------------------------------------------------------------------------------- /pkgdown/favicon/favicon.ico: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/debruine/data-sim-workshops/9bd222cd3e1c988e30d005a2d80cf57f50aa8994/pkgdown/favicon/favicon.ico -------------------------------------------------------------------------------- /vignettes/.gitignore: -------------------------------------------------------------------------------- 1 | *.html 2 | *.R 3 | -------------------------------------------------------------------------------- /vignettes/calories.Rmd: -------------------------------------------------------------------------------- 1 | --- 2 | title: "Calorie Placement Re-Simulation" 3 | author: "Lisa DeBruine" 4 | output: 5 | rmarkdown::html_document: 6 | df_print: paged 7 | vignette: > 8 | %\VignetteIndexEntry{Calorie Placement Re-Simulation} 9 | %\VignetteEngine{knitr::rmarkdown} 10 | %\VignetteEncoding{UTF-8} 11 | --- 12 | 13 | ```{r setup, include=FALSE} 14 | knitr::opts_chunk$set(echo = TRUE) 15 | 16 | library(tidyverse) 17 | library(faux) 18 | library(afex) 19 | library(emmeans) 20 | faux_options(plot = FALSE) 21 | 22 | set.seed(8675309) 23 | ``` 24 | 25 | ## Data Source 26 | 27 | We will be replicating some of the re-analyses in Francis & Thunell's (2020) Meta-Psychology paper: Excess success in "Don't count calorie labeling out: Calorie counts on the left side of menu items lead to lower calorie food choices". 28 | 29 | They ran power analyses for all 6 studies in Dallas, Liu, and Ubel's (2019) study showing that people order food with significantly fewer calories when the calorie count was placed to the left of the item than to the right (or having no calorie label). They then used these power estimates to calculate the probability of all 6 out of 6 studies being significant, given the observed power of each study. 30 | 31 | * [Re-analysis](https://doi.org/10.15626/MP.2019.2266) 32 | * [Re-analysis code](https://osf.io/xrdhj/) 33 | * [Original paper](https://doi.org/10.1002/jcpy.1053) 34 | 35 | Table 1 of the re-analysis paper provides all of the parameters we will need. 36 | 37 | ## Reanalyses 38 | 39 | ### Study 2 40 | 41 | We'll start with S2 because the analysis is very straightforward. It's a between-subjects design, where 143 subjects saw calorie placement on the left and their mean calories ordered were 1249.83 (SD = 449.07), while 132 subjects saw calorie placement on the right and their mean calories ordered were 1362.31 (SD = 447.35). 42 | 43 | Let's first simulate a single data table with these parameters and set up our analysis. 44 | 45 | ```{r} 46 | data <- sim_design( 47 | between = list(placement = c("left", "right")), 48 | mu = c(left = 1249.83, right = 1362.31), 49 | sd = c(left = 449.07, right = 447.35), 50 | n = c(left = 143, right = 132), 51 | dv = "calories" 52 | ) 53 | ``` 54 | 55 | Wrap the analysis in a function using the `tidy()` function from {broom} to get the results in a tidy table. Check that it works by running it on the single data set above. 56 | 57 | ```{r} 58 | s2_analyse <- function(data) { 59 | t.test(calories ~ placement, data = data) |> 60 | broom::tidy() 61 | } 62 | 63 | s2_analyse(data) 64 | ``` 65 | 66 | 67 | Now, simulate the data 500 times. 68 | 69 | ```{r} 70 | s2 <- sim_design( 71 | between = list(placement = c("left", "right")), 72 | mu = c(left = 1249.83, right = 1362.31), 73 | sd = c(left = 449.07, right = 447.35), 74 | n = c(left = 143, right = 132), 75 | dv = "calories", 76 | rep = 500 77 | ) 78 | ``` 79 | 80 | Run the analysis on each data set. 81 | 82 | ```{r} 83 | s2_sim <- s2 |> 84 | mutate(analysis = map(data, s2_analyse)) |> 85 | select(-data) |> 86 | unnest(analysis) 87 | 88 | head(s2_sim) 89 | ``` 90 | 91 | Summarise the `p.value` column to get power. 92 | 93 | ```{r} 94 | s2_power <- s2_sim |> 95 | mutate(sig = p.value < .05) |> 96 | summarise(power = mean(sig)) |> 97 | pull(power) 98 | ``` 99 | 100 | Compare this value (`r s2_power`) with the value in the paper (0.5426). 101 | 102 | ### Study 1 103 | 104 | Study 1 is a little more complicated because the design includes a "no label" condition, so the decision rule for supporting the hypothesis is more complicated. 105 | 106 | The data simulation is relatively straightforward, though. 107 | 108 | ```{r} 109 | data <- sim_design( 110 | between = list(placement = c("left", "right", "none")), 111 | mu = c(654.53, 865.41, 914.34), 112 | sd = c(390.45, 517.26, 560.94), 113 | n = c(45, 54, 50), 114 | dv = "calories" 115 | ) 116 | ``` 117 | 118 | Set up the analysis. Here, we really just care about three p-values, so we'll just return those. We can use a function from the {emmeans} package to check the two pairwise comparisons. 119 | 120 | ```{r} 121 | afex::set_sum_contrasts() # avoids annoying afex message on each run 122 | afex_options(include_aov = TRUE) # we need aov for lsmeans 123 | 124 | s1_analyse <- function(data) { 125 | # main effect of placement 126 | a <- afex::aov_ez( 127 | id = "id", 128 | dv = "calories", 129 | between = "placement", 130 | data = data 131 | ) 132 | 133 | # contrasts 134 | e <- emmeans(a, "placement") 135 | c1 <- list(lr = c(-0.5, 0.5, 0), 136 | ln = c(-0.5, 0, 0.5)) 137 | b <- contrast(e, c1, adjust = "holm") |> 138 | broom::tidy() 139 | 140 | data.frame( 141 | p_all = a$anova_table$`Pr(>F)`[[1]], 142 | p_1 = b$adj.p.value[[1]], 143 | p_2 = b$adj.p.value[[2]] 144 | ) 145 | } 146 | 147 | s1_analyse(data) 148 | ``` 149 | 150 | Let's just replicate this 100 times so the simulation doesn't take too long to run at first. We can always increase it later after we've run some sense checks. 151 | 152 | ```{r} 153 | s1 <- sim_design( 154 | between = list(placement = c("left", "right", "none")), 155 | mu = c(654.53, 865.41, 914.34), 156 | sd = c(390.45, 517.26, 560.94), 157 | n = c(45, 54, 50), 158 | dv = "calories", 159 | rep = 100 160 | ) 161 | ``` 162 | 163 | Run the analysis on each data set. 164 | 165 | ```{r} 166 | s1_sim <- s1 |> 167 | mutate(analysis = map(data, s1_analyse)) |> 168 | select(-data) |> 169 | unnest(analysis) 170 | 171 | head(s1_sim) 172 | ``` 173 | 174 | Calculating power is a little trickier here, as all three p-values need to be significant here to support the hypothesis. 175 | 176 | ```{r} 177 | s1_power <- s1_sim |> 178 | mutate(sig = (p_all < .05) & 179 | (p_1 < .05) & 180 | (p_2 < .05) ) |> 181 | summarise(power = mean(sig)) |> 182 | pull(power) 183 | ``` 184 | 185 | Compare this value (`r s1_power`) with the value in the paper (0.4582). 186 | 187 | ### Study 3 188 | 189 | Now you can use the pattern from Study 1 to analyse the data for Study 3. We'll start with the repeated data set. 190 | 191 | ```{r} 192 | s3 <- sim_design( 193 | between = list(placement = c("left", "right", "none")), 194 | mu = c(1428.24, 1308.66, 1436.79), 195 | sd = c(377.02, 420.14, 378.47), 196 | n = c(85, 86, 81), 197 | dv = "calories", 198 | rep = 100 199 | ) 200 | ``` 201 | 202 | These data were collected in the Hebrew language, which reads right to left, so the paired contrasts will be different. 203 | 204 | ```{r} 205 | s3_analyse <- function(data) { 206 | # main effect of placement 207 | a <- afex::aov_ez( 208 | id = "id", 209 | dv = "calories", 210 | between = "placement", 211 | data = data 212 | ) 213 | 214 | # contrasts (reversed) 215 | e <- emmeans(a, "placement") 216 | c1 <- list(rl = c(0.5, -0.5, 0), 217 | ln = c(0, -0.5, 0.5)) 218 | b <- contrast(e, c1, adjust = "holm") |> 219 | broom::tidy() 220 | 221 | data.frame( 222 | p_all = a$anova_table$`Pr(>F)`[[1]], 223 | p_1 = b$adj.p.value[[1]], 224 | p_2 = b$adj.p.value[[2]] 225 | ) 226 | } 227 | ``` 228 | 229 | 230 | Run the analysis on each data set. 231 | 232 | ```{r} 233 | s3_sim <- s3 |> 234 | mutate(analysis = map(data, s3_analyse)) |> 235 | select(-data) |> 236 | unnest(analysis) 237 | 238 | head(s3_sim) 239 | ``` 240 | 241 | ```{r} 242 | s3_power <- s3_sim |> 243 | mutate(sig = (p_all < .05) & 244 | (p_1 < .05) & 245 | (p_2 < .05) ) |> 246 | summarise(power = mean(sig)) |> 247 | pull(power) 248 | ``` 249 | 250 | Compare this value (`r s3_power`) with the value in the paper (0.3626). 251 | 252 | 253 | ### Study S1 254 | 255 | Now you can use the pattern from Study 2 to analyse the data for Study S1. You can even reuse the analysis function `s2_analyse()`! 256 | 257 | ```{r} 258 | ss1 <- sim_design( 259 | between = list(placement = c("left", "right")), 260 | mu = c(left = 185.94, right = 215.73), 261 | sd = c(left = 93.92, right = 95.33), 262 | n = c(left = 99, right = 77), 263 | dv = "calories", 264 | rep = 1000 265 | ) 266 | ``` 267 | 268 | ```{r} 269 | ss1_sim <- ss1 |> 270 | mutate(analysis = map(data, s2_analyse)) |> 271 | select(-data) |> 272 | unnest(analysis) 273 | ``` 274 | 275 | 276 | ```{r} 277 | ss1_power <- ss1_sim |> 278 | mutate(sig = p.value < .05) |> 279 | summarise(power = mean(sig)) |> 280 | pull(power) 281 | ``` 282 | 283 | 284 | ### Study S2 285 | 286 | Now you can use the pattern from Study 1 to analyse the data for Study S2. You can even reuse the analysis function `s1_analyse()`! 287 | 288 | ```{r} 289 | ss2 <- sim_design( 290 | between = list(placement = c("left", "right", "none")), 291 | mu = c(1182.15, 1302.23, 1373.74), 292 | sd = c(477.60, 434.41, 475.77), 293 | n = c(139, 141, 151), 294 | dv = "calories", 295 | rep = 100 296 | ) 297 | ``` 298 | 299 | ```{r} 300 | ss2_sim <- ss2 |> 301 | mutate(analysis = map(data, s1_analyse)) |> 302 | select(-data) |> 303 | unnest(analysis) 304 | ``` 305 | 306 | ```{r} 307 | ss2_power <- ss2_sim |> 308 | mutate(sig = (p_all < .05) & 309 | (p_1 < .05) & 310 | (p_2 < .05) ) |> 311 | summarise(power = mean(sig)) |> 312 | pull(power) 313 | ``` 314 | 315 | ### Study S3 316 | 317 | Now you can use the pattern from Study 1 to analyse the data for Study S3. 318 | 319 | ```{r} 320 | ss3 <- sim_design( 321 | between = list(placement = c("left", "right", "none")), 322 | mu = c(1302.03, 1373.15, 1404.35), 323 | sd = c(480.02, 442.49, 422.03), 324 | n = c(336, 337, 333), 325 | dv = "calories", 326 | rep = 100 327 | ) 328 | ``` 329 | 330 | ```{r} 331 | ss3_sim <- ss3 |> 332 | mutate(analysis = map(data, s1_analyse)) |> 333 | select(-data) |> 334 | unnest(analysis) 335 | ``` 336 | 337 | ```{r} 338 | ss3_power <- ss3_sim |> 339 | mutate(sig = (p_all < .05) & 340 | (p_1 < .05) & 341 | (p_2 < .05) ) |> 342 | summarise(power = mean(sig)) |> 343 | pull(power) 344 | ``` 345 | 346 | ## Conclusion 347 | 348 | Now that you've calculated power for each of the 6 studies, just multiply the 6 power values together to get the probability that all 6 studies will be significant. 349 | 350 | 351 | ```{r} 352 | power_table <- tribble( 353 | ~study, ~power_ft, ~ power_my, 354 | "1", 0.4582, s1_power, 355 | "2", 0.5426, s2_power, 356 | "3", 0.3626, s3_power, 357 | "S1", 0.5358, ss1_power, 358 | "S2", 0.5667, ss2_power, 359 | "S3", 0.4953, ss3_power 360 | ) 361 | 362 | power_table 363 | ``` 364 | 365 | The `reduce()` function from {purrr} applies a function sequentially over a vector, so can give up the product of all the values in the power columns. 366 | 367 | ```{r} 368 | prob_ft <- purrr::reduce(power_table$power_ft, `*`) 369 | prob_my <- purrr::reduce(power_table$power_my, `*`) 370 | ``` 371 | 372 | The Francis & Thunell paper showed a `r prob_ft` probability of getting 6 of 6 studies significant. Our re-simulation showed a `r prob_my` probability. 373 | 374 | -------------------------------------------------------------------------------- /vignettes/data/sim-data-ind-samples.csv: -------------------------------------------------------------------------------- 1 | sub_condition,score 2 | A,8.659143827311953 3 | A,14.306681811819699 4 | A,10.713040492669082 5 | A,10.414204142174725 6 | A,10.613585812486479 7 | A,7.953210054952468 8 | A,6.52177813719714 9 | A,8.365117743095091 10 | A,11.092587459730957 11 | A,14.062281356126192 12 | A,12.051544481443418 13 | A,14.500260485544738 14 | A,11.327941790503406 15 | A,4.849147729673105 16 | A,7.994833894625758 17 | A,11.496559868710417 18 | A,11.12338443739307 19 | A,9.41681818225328 20 | A,8.378648100425668 21 | A,12.25368069014773 22 | A,12.208135307589908 23 | A,12.833950353709977 24 | A,11.822052862134656 25 | A,10.061984283518894 26 | A,10.119329581809868 27 | A,8.034391626028608 28 | A,10.238392796236885 29 | A,6.914536464377451 30 | A,7.886804263927875 31 | A,12.683818897306685 32 | A,13.50246310263745 33 | A,12.48201122505167 34 | A,9.835467681159713 35 | A,14.427085525719825 36 | A,6.965452117476058 37 | A,7.558862658433231 38 | A,8.068466141423594 39 | A,10.806290949776782 40 | A,10.658575820446046 41 | A,12.830161666637538 42 | A,7.544107890339511 43 | A,10.15804909878836 44 | A,7.574301063484666 45 | A,7.157640215571564 46 | A,10.763093438116673 47 | A,6.8443775356767595 48 | A,11.406947180834656 49 | A,8.565681068680025 50 | A,16.47173345179801 51 | A,13.280822099741172 52 | B,7.290556337684006 53 | B,7.560913283694237 54 | B,9.601957010873326 55 | B,14.672506746437755 56 | B,14.160166038360277 57 | B,11.784889131576811 58 | B,14.34143310885541 59 | B,12.55246678752024 60 | B,10.840201103843262 61 | B,14.977746875421019 62 | B,10.11201265244846 63 | B,8.116960727589733 64 | B,8.345164636310528 65 | B,11.580494569521154 66 | B,8.991330584941077 67 | B,8.570957515146812 68 | B,11.035994512923782 69 | B,13.05100239532186 70 | B,8.163754096912456 71 | B,12.920561183356474 72 | B,10.499891656247017 73 | B,7.965685472608644 74 | B,13.936347873516183 75 | B,8.949386887496775 76 | B,11.217647958524797 77 | B,10.781408041597441 78 | B,8.632934410323612 79 | B,14.889816106792939 80 | B,9.764900268001272 81 | B,12.65100663200029 82 | B,12.603346488280355 83 | B,9.333126007262221 84 | B,10.874738393461369 85 | B,11.540604067658668 86 | B,13.589746212707297 87 | B,11.552525767214489 88 | B,12.144489376016997 89 | B,10.939591720847423 90 | B,9.496188379394045 91 | B,7.380340991356158 92 | B,8.388628883391284 93 | B,16.916674965222786 94 | B,10.204337529817664 95 | B,9.368852181073988 96 | B,6.157268812910302 97 | B,8.561531899970976 98 | B,11.802891862870466 99 | B,11.58933949228355 100 | B,10.56113629759848 101 | B,9.240614581122362 102 | --------------------------------------------------------------------------------