-
Notifications
You must be signed in to change notification settings - Fork 1
/
Chapter_35.html
775 lines (733 loc) · 76.1 KB
/
Chapter_35.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
<!DOCTYPE html>
<html lang="" xml:lang="">
<head>
<meta charset="utf-8" />
<meta http-equiv="X-UA-Compatible" content="IE=edge" />
<title>Chapter 35 Randomisation | Fundamental statistical concepts and techniques in the biological and environmental sciences: With jamovi</title>
<meta name="description" content="This is an introductory statistics textbook for students in the biological and environmental sciences with examples using jamovi statistical software." />
<meta name="generator" content="bookdown 0.39 and GitBook 2.6.7" />
<meta property="og:title" content="Chapter 35 Randomisation | Fundamental statistical concepts and techniques in the biological and environmental sciences: With jamovi" />
<meta property="og:type" content="book" />
<meta property="og:image" content="/img/cover.png" />
<meta property="og:description" content="This is an introductory statistics textbook for students in the biological and environmental sciences with examples using jamovi statistical software." />
<meta name="github-repo" content="bradduthie/stats" />
<meta name="twitter:card" content="summary" />
<meta name="twitter:title" content="Chapter 35 Randomisation | Fundamental statistical concepts and techniques in the biological and environmental sciences: With jamovi" />
<meta name="twitter:description" content="This is an introductory statistics textbook for students in the biological and environmental sciences with examples using jamovi statistical software." />
<meta name="twitter:image" content="/img/cover.png" />
<meta name="author" content="A. Bradley Duthie" />
<meta name="date" content="2024-08-06" />
<meta name="viewport" content="width=device-width, initial-scale=1" />
<meta name="apple-mobile-web-app-capable" content="yes" />
<meta name="apple-mobile-web-app-status-bar-style" content="black" />
<link rel="prev" href="Chapter_34.html"/>
<link rel="next" href="appendexA.html"/>
<script src="libs/jquery-3.6.0/jquery-3.6.0.min.js"></script>
<script src="https://cdn.jsdelivr.net/npm/[email protected]/dist/fuse.min.js"></script>
<link href="libs/gitbook-2.6.7/css/style.css" rel="stylesheet" />
<link href="libs/gitbook-2.6.7/css/plugin-table.css" rel="stylesheet" />
<link href="libs/gitbook-2.6.7/css/plugin-bookdown.css" rel="stylesheet" />
<link href="libs/gitbook-2.6.7/css/plugin-highlight.css" rel="stylesheet" />
<link href="libs/gitbook-2.6.7/css/plugin-search.css" rel="stylesheet" />
<link href="libs/gitbook-2.6.7/css/plugin-fontsettings.css" rel="stylesheet" />
<link href="libs/gitbook-2.6.7/css/plugin-clipboard.css" rel="stylesheet" />
<link href="libs/anchor-sections-1.1.0/anchor-sections.css" rel="stylesheet" />
<link href="libs/anchor-sections-1.1.0/anchor-sections-hash.css" rel="stylesheet" />
<script src="libs/anchor-sections-1.1.0/anchor-sections.js"></script>
<style type="text/css">
pre > code.sourceCode { white-space: pre; position: relative; }
pre > code.sourceCode > span { display: inline-block; line-height: 1.25; }
pre > code.sourceCode > span:empty { height: 1.2em; }
.sourceCode { overflow: visible; }
code.sourceCode > span { color: inherit; text-decoration: inherit; }
pre.sourceCode { margin: 0; }
@media screen {
div.sourceCode { overflow: auto; }
}
@media print {
pre > code.sourceCode { white-space: pre-wrap; }
pre > code.sourceCode > span { text-indent: -5em; padding-left: 5em; }
}
pre.numberSource code
{ counter-reset: source-line 0; }
pre.numberSource code > span
{ position: relative; left: -4em; counter-increment: source-line; }
pre.numberSource code > span > a:first-child::before
{ content: counter(source-line);
position: relative; left: -1em; text-align: right; vertical-align: baseline;
border: none; display: inline-block;
-webkit-touch-callout: none; -webkit-user-select: none;
-khtml-user-select: none; -moz-user-select: none;
-ms-user-select: none; user-select: none;
padding: 0 4px; width: 4em;
}
pre.numberSource { margin-left: 3em; padding-left: 4px; }
div.sourceCode
{ }
@media screen {
pre > code.sourceCode > span > a:first-child::before { text-decoration: underline; }
}
code span.al { font-weight: bold; } /* Alert */
code span.an { font-style: italic; } /* Annotation */
code span.cf { font-weight: bold; } /* ControlFlow */
code span.co { font-style: italic; } /* Comment */
code span.cv { font-style: italic; } /* CommentVar */
code span.do { font-style: italic; } /* Documentation */
code span.dt { text-decoration: underline; } /* DataType */
code span.er { font-weight: bold; } /* Error */
code span.in { font-style: italic; } /* Information */
code span.kw { font-weight: bold; } /* Keyword */
code span.pp { font-weight: bold; } /* Preprocessor */
code span.wa { font-style: italic; } /* Warning */
</style>
<style type="text/css">
div.hanging-indent{margin-left: 1.5em; text-indent: -1.5em;}
</style>
<style type="text/css">
/* Used with Pandoc 2.11+ new --citeproc when CSL is used */
div.csl-bib-body { }
div.csl-entry {
clear: both;
}
.hanging div.csl-entry {
margin-left:2em;
text-indent:-2em;
}
div.csl-left-margin {
min-width:2em;
float:left;
}
div.csl-right-inline {
margin-left:2em;
padding-left:1em;
}
div.csl-indent {
margin-left: 2em;
}
</style>
<link rel="stylesheet" href="style.css" type="text/css" />
</head>
<body>
<div class="book without-animation with-summary font-size-2 font-family-1" data-basepath=".">
<div class="book-summary">
<nav role="navigation">
<ul class="summary">
<li><a href="./">Statistics with jamovi</a></li>
<li class="divider"></li>
<li class="chapter" data-level="" data-path="index.html"><a href="index.html"><i class="fa fa-check"></i>Preface</a>
<ul>
<li class="chapter" data-level="" data-path="index.html"><a href="index.html#structure"><i class="fa fa-check"></i>How this book is structured</a></li>
<li class="chapter" data-level="" data-path="index.html"><a href="index.html#datasets"><i class="fa fa-check"></i>Datasets used in this book</a></li>
<li class="chapter" data-level="" data-path="index.html"><a href="index.html#acknowledgements"><i class="fa fa-check"></i>Acknowledgements</a></li>
<li class="chapter" data-level="" data-path="index.html"><a href="index.html#author"><i class="fa fa-check"></i>About the author</a></li>
</ul></li>
<li class="chapter" data-level="1" data-path="Chapter_1.html"><a href="Chapter_1.html"><i class="fa fa-check"></i><b>1</b> Background mathematics</a>
<ul>
<li class="chapter" data-level="1.1" data-path="Chapter_1.html"><a href="Chapter_1.html#numbers-and-operations"><i class="fa fa-check"></i><b>1.1</b> Numbers and operations</a></li>
<li class="chapter" data-level="1.2" data-path="Chapter_1.html"><a href="Chapter_1.html#logarithms"><i class="fa fa-check"></i><b>1.2</b> Logarithms</a></li>
<li class="chapter" data-level="1.3" data-path="Chapter_1.html"><a href="Chapter_1.html#order-of-operations"><i class="fa fa-check"></i><b>1.3</b> Order of operations</a></li>
</ul></li>
<li class="chapter" data-level="2" data-path="Chapter_2.html"><a href="Chapter_2.html"><i class="fa fa-check"></i><b>2</b> Data organisation</a>
<ul>
<li class="chapter" data-level="2.1" data-path="Chapter_2.html"><a href="Chapter_2.html#tidy-data"><i class="fa fa-check"></i><b>2.1</b> Tidy data</a></li>
<li class="chapter" data-level="2.2" data-path="Chapter_2.html"><a href="Chapter_2.html#data-files"><i class="fa fa-check"></i><b>2.2</b> Data files</a></li>
<li class="chapter" data-level="2.3" data-path="Chapter_2.html"><a href="Chapter_2.html#managing-data-files"><i class="fa fa-check"></i><b>2.3</b> Managing data files</a></li>
</ul></li>
<li class="chapter" data-level="3" data-path="Chapter_3.html"><a href="Chapter_3.html"><i class="fa fa-check"></i><b>3</b> <em>Practical</em>. Preparing data</a>
<ul>
<li class="chapter" data-level="3.1" data-path="Chapter_3.html"><a href="Chapter_3.html#transferring-data-to-a-spreadsheet"><i class="fa fa-check"></i><b>3.1</b> Transferring data to a spreadsheet</a></li>
<li class="chapter" data-level="3.2" data-path="Chapter_3.html"><a href="Chapter_3.html#making-spreadsheet-data-tidy"><i class="fa fa-check"></i><b>3.2</b> Making spreadsheet data tidy</a></li>
<li class="chapter" data-level="3.3" data-path="Chapter_3.html"><a href="Chapter_3.html#making-data-tidy-again"><i class="fa fa-check"></i><b>3.3</b> Making data tidy again</a></li>
<li class="chapter" data-level="3.4" data-path="Chapter_3.html"><a href="Chapter_3.html#tidy-data-and-spreadsheet-calculations"><i class="fa fa-check"></i><b>3.4</b> Tidy data and spreadsheet calculations</a></li>
<li class="chapter" data-level="3.5" data-path="Chapter_3.html"><a href="Chapter_3.html#summary"><i class="fa fa-check"></i><b>3.5</b> Summary</a></li>
</ul></li>
<li class="chapter" data-level="4" data-path="Chapter_4.html"><a href="Chapter_4.html"><i class="fa fa-check"></i><b>4</b> Populations and samples</a></li>
<li class="chapter" data-level="5" data-path="Chapter_5.html"><a href="Chapter_5.html"><i class="fa fa-check"></i><b>5</b> Types of variables</a></li>
<li class="chapter" data-level="6" data-path="Chapter_6.html"><a href="Chapter_6.html"><i class="fa fa-check"></i><b>6</b> Accuracy, precision, and units</a>
<ul>
<li class="chapter" data-level="6.1" data-path="Chapter_6.html"><a href="Chapter_6.html#accuracy"><i class="fa fa-check"></i><b>6.1</b> Accuracy</a></li>
<li class="chapter" data-level="6.2" data-path="Chapter_6.html"><a href="Chapter_6.html#precision"><i class="fa fa-check"></i><b>6.2</b> Precision</a></li>
<li class="chapter" data-level="6.3" data-path="Chapter_6.html"><a href="Chapter_6.html#systems-of-units"><i class="fa fa-check"></i><b>6.3</b> Systems of units</a></li>
</ul></li>
<li class="chapter" data-level="7" data-path="Chapter_7.html"><a href="Chapter_7.html"><i class="fa fa-check"></i><b>7</b> Uncertainty propagation</a>
<ul>
<li class="chapter" data-level="7.1" data-path="Chapter_7.html"><a href="Chapter_7.html#adding-or-subtracting-errors"><i class="fa fa-check"></i><b>7.1</b> Adding or subtracting errors</a></li>
<li class="chapter" data-level="7.2" data-path="Chapter_7.html"><a href="Chapter_7.html#multiplying-or-dividing-errors"><i class="fa fa-check"></i><b>7.2</b> Multiplying or dividing errors</a></li>
</ul></li>
<li class="chapter" data-level="8" data-path="Chapter_8.html"><a href="Chapter_8.html"><i class="fa fa-check"></i><b>8</b> <em>Practical</em>. Introduction to jamovi</a>
<ul>
<li class="chapter" data-level="8.1" data-path="Chapter_8.html"><a href="Chapter_8.html#summary_statistics_02"><i class="fa fa-check"></i><b>8.1</b> Summary statistics</a></li>
<li class="chapter" data-level="8.2" data-path="Chapter_8.html"><a href="Chapter_8.html#transforming_variables_02"><i class="fa fa-check"></i><b>8.2</b> Transforming variables</a></li>
<li class="chapter" data-level="8.3" data-path="Chapter_8.html"><a href="Chapter_8.html#computing_variables_02"><i class="fa fa-check"></i><b>8.3</b> Computing variables</a></li>
<li class="chapter" data-level="8.4" data-path="Chapter_8.html"><a href="Chapter_8.html#summary-1"><i class="fa fa-check"></i><b>8.4</b> Summary</a></li>
</ul></li>
<li class="chapter" data-level="9" data-path="Chapter_9.html"><a href="Chapter_9.html"><i class="fa fa-check"></i><b>9</b> Decimal places, significant figures, and rounding</a>
<ul>
<li class="chapter" data-level="9.1" data-path="Chapter_9.html"><a href="Chapter_9.html#decimal-places-and-significant-figures"><i class="fa fa-check"></i><b>9.1</b> Decimal places and significant figures</a></li>
<li class="chapter" data-level="9.2" data-path="Chapter_9.html"><a href="Chapter_9.html#rounding"><i class="fa fa-check"></i><b>9.2</b> Rounding</a></li>
</ul></li>
<li class="chapter" data-level="10" data-path="Chapter_10.html"><a href="Chapter_10.html"><i class="fa fa-check"></i><b>10</b> Graphs</a>
<ul>
<li class="chapter" data-level="10.1" data-path="Chapter_10.html"><a href="Chapter_10.html#histograms"><i class="fa fa-check"></i><b>10.1</b> Histograms</a></li>
<li class="chapter" data-level="10.2" data-path="Chapter_10.html"><a href="Chapter_10.html#barplots-and-pie-charts"><i class="fa fa-check"></i><b>10.2</b> Barplots and pie charts</a></li>
<li class="chapter" data-level="10.3" data-path="Chapter_10.html"><a href="Chapter_10.html#box-whisker-plots"><i class="fa fa-check"></i><b>10.3</b> Box-whisker plots</a></li>
</ul></li>
<li class="chapter" data-level="11" data-path="Chapter_11.html"><a href="Chapter_11.html"><i class="fa fa-check"></i><b>11</b> Measures of central tendency</a>
<ul>
<li class="chapter" data-level="11.1" data-path="Chapter_11.html"><a href="Chapter_11.html#the-mean"><i class="fa fa-check"></i><b>11.1</b> The mean</a></li>
<li class="chapter" data-level="11.2" data-path="Chapter_11.html"><a href="Chapter_11.html#the-mode"><i class="fa fa-check"></i><b>11.2</b> The mode</a></li>
<li class="chapter" data-level="11.3" data-path="Chapter_11.html"><a href="Chapter_11.html#the-median-and-quantiles"><i class="fa fa-check"></i><b>11.3</b> The median and quantiles</a></li>
</ul></li>
<li class="chapter" data-level="12" data-path="Chapter_12.html"><a href="Chapter_12.html"><i class="fa fa-check"></i><b>12</b> Measures of spread</a>
<ul>
<li class="chapter" data-level="12.1" data-path="Chapter_12.html"><a href="Chapter_12.html#the-range"><i class="fa fa-check"></i><b>12.1</b> The range</a></li>
<li class="chapter" data-level="12.2" data-path="Chapter_12.html"><a href="Chapter_12.html#the-inter-quartile-range"><i class="fa fa-check"></i><b>12.2</b> The inter-quartile range</a></li>
<li class="chapter" data-level="12.3" data-path="Chapter_12.html"><a href="Chapter_12.html#the-variance"><i class="fa fa-check"></i><b>12.3</b> The variance</a></li>
<li class="chapter" data-level="12.4" data-path="Chapter_12.html"><a href="Chapter_12.html#the-standard-deviation"><i class="fa fa-check"></i><b>12.4</b> The standard deviation</a></li>
<li class="chapter" data-level="12.5" data-path="Chapter_12.html"><a href="Chapter_12.html#the-coefficient-of-variation"><i class="fa fa-check"></i><b>12.5</b> The coefficient of variation</a></li>
<li class="chapter" data-level="12.6" data-path="Chapter_12.html"><a href="Chapter_12.html#the-standard-error"><i class="fa fa-check"></i><b>12.6</b> The standard error</a></li>
</ul></li>
<li class="chapter" data-level="13" data-path="Chapter_13.html"><a href="Chapter_13.html"><i class="fa fa-check"></i><b>13</b> Skew and kurtosis</a>
<ul>
<li class="chapter" data-level="13.1" data-path="Chapter_13.html"><a href="Chapter_13.html#skew"><i class="fa fa-check"></i><b>13.1</b> Skew</a></li>
<li class="chapter" data-level="13.2" data-path="Chapter_13.html"><a href="Chapter_13.html#kurtosis"><i class="fa fa-check"></i><b>13.2</b> Kurtosis</a></li>
<li class="chapter" data-level="13.3" data-path="Chapter_13.html"><a href="Chapter_13.html#moments"><i class="fa fa-check"></i><b>13.3</b> Moments</a></li>
</ul></li>
<li class="chapter" data-level="14" data-path="Chapter_14.html"><a href="Chapter_14.html"><i class="fa fa-check"></i><b>14</b> <em>Practical</em>. Plotting and statistical summaries in jamovi</a>
<ul>
<li class="chapter" data-level="14.1" data-path="Chapter_14.html"><a href="Chapter_14.html#reorganise-the-dataset-into-a-tidy-format"><i class="fa fa-check"></i><b>14.1</b> Reorganise the dataset into a tidy format</a></li>
<li class="chapter" data-level="14.2" data-path="Chapter_14.html"><a href="Chapter_14.html#histograms-and-box-whisker-plots"><i class="fa fa-check"></i><b>14.2</b> Histograms and box-whisker plots</a></li>
<li class="chapter" data-level="14.3" data-path="Chapter_14.html"><a href="Chapter_14.html#calculate-summary-statistics"><i class="fa fa-check"></i><b>14.3</b> Calculate summary statistics</a></li>
<li class="chapter" data-level="14.4" data-path="Chapter_14.html"><a href="Chapter_14.html#reporting-decimals-and-significant-figures"><i class="fa fa-check"></i><b>14.4</b> Reporting decimals and significant figures</a></li>
<li class="chapter" data-level="14.5" data-path="Chapter_14.html"><a href="Chapter_14.html#comparing-across-sites"><i class="fa fa-check"></i><b>14.5</b> Comparing across sites</a></li>
</ul></li>
<li class="chapter" data-level="15" data-path="Chapter_15.html"><a href="Chapter_15.html"><i class="fa fa-check"></i><b>15</b> Introduction to probability models</a>
<ul>
<li class="chapter" data-level="15.1" data-path="Chapter_15.html"><a href="Chapter_15.html#instructive-example"><i class="fa fa-check"></i><b>15.1</b> Instructive example</a></li>
<li class="chapter" data-level="15.2" data-path="Chapter_15.html"><a href="Chapter_15.html#biological-applications"><i class="fa fa-check"></i><b>15.2</b> Biological applications</a></li>
<li class="chapter" data-level="15.3" data-path="Chapter_15.html"><a href="Chapter_15.html#sampling-with-and-without-replacement"><i class="fa fa-check"></i><b>15.3</b> Sampling with and without replacement</a></li>
<li class="chapter" data-level="15.4" data-path="Chapter_15.html"><a href="Chapter_15.html#probability-distributions"><i class="fa fa-check"></i><b>15.4</b> Probability distributions</a>
<ul>
<li class="chapter" data-level="15.4.1" data-path="Chapter_15.html"><a href="Chapter_15.html#binomial-distribution"><i class="fa fa-check"></i><b>15.4.1</b> Binomial distribution</a></li>
<li class="chapter" data-level="15.4.2" data-path="Chapter_15.html"><a href="Chapter_15.html#poisson-distribution"><i class="fa fa-check"></i><b>15.4.2</b> Poisson distribution</a></li>
<li class="chapter" data-level="15.4.3" data-path="Chapter_15.html"><a href="Chapter_15.html#uniform-distribution"><i class="fa fa-check"></i><b>15.4.3</b> Uniform distribution</a></li>
<li class="chapter" data-level="15.4.4" data-path="Chapter_15.html"><a href="Chapter_15.html#normal-distribution"><i class="fa fa-check"></i><b>15.4.4</b> Normal distribution</a></li>
</ul></li>
<li class="chapter" data-level="15.5" data-path="Chapter_15.html"><a href="Chapter_15.html#summary-2"><i class="fa fa-check"></i><b>15.5</b> Summary</a></li>
</ul></li>
<li class="chapter" data-level="16" data-path="Chapter_16.html"><a href="Chapter_16.html"><i class="fa fa-check"></i><b>16</b> Central Limit Theorem</a>
<ul>
<li class="chapter" data-level="16.1" data-path="Chapter_16.html"><a href="Chapter_16.html#the-distribution-of-means-is-normal"><i class="fa fa-check"></i><b>16.1</b> The distribution of means is normal</a></li>
<li class="chapter" data-level="16.2" data-path="Chapter_16.html"><a href="Chapter_16.html#probability-and-z-scores"><i class="fa fa-check"></i><b>16.2</b> Probability and z-scores</a></li>
</ul></li>
<li class="chapter" data-level="17" data-path="Chapter_17.html"><a href="Chapter_17.html"><i class="fa fa-check"></i><b>17</b> <em>Practical</em>. Probability and simulation</a>
<ul>
<li class="chapter" data-level="17.1" data-path="Chapter_17.html"><a href="Chapter_17.html#probabilities-from-a-dataset"><i class="fa fa-check"></i><b>17.1</b> Probabilities from a dataset</a></li>
<li class="chapter" data-level="17.2" data-path="Chapter_17.html"><a href="Chapter_17.html#probabilities-from-a-normal-distribution"><i class="fa fa-check"></i><b>17.2</b> Probabilities from a normal distribution</a></li>
<li class="chapter" data-level="17.3" data-path="Chapter_17.html"><a href="Chapter_17.html#central-limit-theorem"><i class="fa fa-check"></i><b>17.3</b> Central limit theorem</a></li>
</ul></li>
<li class="chapter" data-level="18" data-path="Chapter_18.html"><a href="Chapter_18.html"><i class="fa fa-check"></i><b>18</b> Confidence intervals</a>
<ul>
<li class="chapter" data-level="18.1" data-path="Chapter_18.html"><a href="Chapter_18.html#normal-distribution-cis"><i class="fa fa-check"></i><b>18.1</b> Normal distribution CIs</a></li>
<li class="chapter" data-level="18.2" data-path="Chapter_18.html"><a href="Chapter_18.html#binomial-distribution-cis"><i class="fa fa-check"></i><b>18.2</b> Binomial distribution CIs</a></li>
</ul></li>
<li class="chapter" data-level="19" data-path="Chapter_19.html"><a href="Chapter_19.html"><i class="fa fa-check"></i><b>19</b> The t-interval</a></li>
<li class="chapter" data-level="20" data-path="Chapter_20.html"><a href="Chapter_20.html"><i class="fa fa-check"></i><b>20</b> <em>Practical</em>. z- and t-intervals</a>
<ul>
<li class="chapter" data-level="20.1" data-path="Chapter_20.html"><a href="Chapter_20.html#confidence-intervals-with-distraction"><i class="fa fa-check"></i><b>20.1</b> Confidence intervals with distrACTION</a></li>
<li class="chapter" data-level="20.2" data-path="Chapter_20.html"><a href="Chapter_20.html#confidence-intervals-from-z--and-t-scores"><i class="fa fa-check"></i><b>20.2</b> Confidence intervals from z- and t-scores</a></li>
<li class="chapter" data-level="20.3" data-path="Chapter_20.html"><a href="Chapter_20.html#confidence-intervals-for-different-sample-sizes"><i class="fa fa-check"></i><b>20.3</b> Confidence intervals for different sample sizes</a></li>
<li class="chapter" data-level="20.4" data-path="Chapter_20.html"><a href="Chapter_20.html#proportion-confidence-intervals"><i class="fa fa-check"></i><b>20.4</b> Proportion confidence intervals</a></li>
<li class="chapter" data-level="20.5" data-path="Chapter_20.html"><a href="Chapter_20.html#another-proportion-confidence-interval"><i class="fa fa-check"></i><b>20.5</b> Another proportion confidence interval</a></li>
</ul></li>
<li class="chapter" data-level="21" data-path="Chapter_21.html"><a href="Chapter_21.html"><i class="fa fa-check"></i><b>21</b> What is hypothesis testing?</a>
<ul>
<li class="chapter" data-level="21.1" data-path="Chapter_21.html"><a href="Chapter_21.html#how-ridiculous-is-our-hypothesis"><i class="fa fa-check"></i><b>21.1</b> How ridiculous is our hypothesis?</a></li>
<li class="chapter" data-level="21.2" data-path="Chapter_21.html"><a href="Chapter_21.html#statistical-hypothesis-testing"><i class="fa fa-check"></i><b>21.2</b> Statistical hypothesis testing</a></li>
<li class="chapter" data-level="21.3" data-path="Chapter_21.html"><a href="Chapter_21.html#p-values-false-positives-and-power"><i class="fa fa-check"></i><b>21.3</b> P-values, false positives, and power</a></li>
</ul></li>
<li class="chapter" data-level="22" data-path="Chapter_22.html"><a href="Chapter_22.html"><i class="fa fa-check"></i><b>22</b> The t-test</a>
<ul>
<li class="chapter" data-level="22.1" data-path="Chapter_22.html"><a href="Chapter_22.html#one-sample-t-test"><i class="fa fa-check"></i><b>22.1</b> One sample t-test</a></li>
<li class="chapter" data-level="22.2" data-path="Chapter_22.html"><a href="Chapter_22.html#independent-samples-t-test"><i class="fa fa-check"></i><b>22.2</b> Independent samples t-test</a></li>
<li class="chapter" data-level="22.3" data-path="Chapter_22.html"><a href="Chapter_22.html#paired-samples-t-test"><i class="fa fa-check"></i><b>22.3</b> Paired samples t-test</a></li>
<li class="chapter" data-level="22.4" data-path="Chapter_22.html"><a href="Chapter_22.html#assumptions-of-t-tests"><i class="fa fa-check"></i><b>22.4</b> Assumptions of t-tests</a></li>
<li class="chapter" data-level="22.5" data-path="Chapter_22.html"><a href="Chapter_22.html#non-parametric-alternatives"><i class="fa fa-check"></i><b>22.5</b> Non-parametric alternatives</a>
<ul>
<li class="chapter" data-level="22.5.1" data-path="Chapter_22.html"><a href="Chapter_22.html#wilcoxon-test"><i class="fa fa-check"></i><b>22.5.1</b> Wilcoxon test</a></li>
<li class="chapter" data-level="22.5.2" data-path="Chapter_22.html"><a href="Chapter_22.html#mann-whitney-u-test"><i class="fa fa-check"></i><b>22.5.2</b> Mann-Whitney U test</a></li>
</ul></li>
<li class="chapter" data-level="22.6" data-path="Chapter_22.html"><a href="Chapter_22.html#summary-3"><i class="fa fa-check"></i><b>22.6</b> Summary</a></li>
</ul></li>
<li class="chapter" data-level="23" data-path="Chapter_23.html"><a href="Chapter_23.html"><i class="fa fa-check"></i><b>23</b> <em>Practical</em>. Hypothesis testing and t-tests</a>
<ul>
<li class="chapter" data-level="23.1" data-path="Chapter_23.html"><a href="Chapter_23.html#one-sample-t-test-1"><i class="fa fa-check"></i><b>23.1</b> One sample t-test</a></li>
<li class="chapter" data-level="23.2" data-path="Chapter_23.html"><a href="Chapter_23.html#paired-t-test"><i class="fa fa-check"></i><b>23.2</b> Paired t-test</a></li>
<li class="chapter" data-level="23.3" data-path="Chapter_23.html"><a href="Chapter_23.html#wilcoxon-test-1"><i class="fa fa-check"></i><b>23.3</b> Wilcoxon test</a></li>
<li class="chapter" data-level="23.4" data-path="Chapter_23.html"><a href="Chapter_23.html#independent-samples-t-test-1"><i class="fa fa-check"></i><b>23.4</b> Independent samples t-test</a></li>
<li class="chapter" data-level="23.5" data-path="Chapter_23.html"><a href="Chapter_23.html#mann-whitney-u-test-1"><i class="fa fa-check"></i><b>23.5</b> Mann-Whitney U Test</a></li>
</ul></li>
<li class="chapter" data-level="24" data-path="Chapter_24.html"><a href="Chapter_24.html"><i class="fa fa-check"></i><b>24</b> Analysis of variance</a>
<ul>
<li class="chapter" data-level="24.1" data-path="Chapter_24.html"><a href="Chapter_24.html#f-distribution"><i class="fa fa-check"></i><b>24.1</b> F-distribution</a></li>
<li class="chapter" data-level="24.2" data-path="Chapter_24.html"><a href="Chapter_24.html#one-way-anova"><i class="fa fa-check"></i><b>24.2</b> One-way ANOVA</a>
<ul>
<li class="chapter" data-level="24.2.1" data-path="Chapter_24.html"><a href="Chapter_24.html#anova-mean-variance-among-groups"><i class="fa fa-check"></i><b>24.2.1</b> ANOVA mean variance among groups</a></li>
<li class="chapter" data-level="24.2.2" data-path="Chapter_24.html"><a href="Chapter_24.html#anova-mean-variance-within-groups"><i class="fa fa-check"></i><b>24.2.2</b> ANOVA mean variance within groups</a></li>
<li class="chapter" data-level="24.2.3" data-path="Chapter_24.html"><a href="Chapter_24.html#anova-f-statistic-calculation"><i class="fa fa-check"></i><b>24.2.3</b> ANOVA F-statistic calculation</a></li>
</ul></li>
<li class="chapter" data-level="24.3" data-path="Chapter_24.html"><a href="Chapter_24.html#assumptions-of-anova"><i class="fa fa-check"></i><b>24.3</b> Assumptions of ANOVA</a></li>
</ul></li>
<li class="chapter" data-level="25" data-path="Chapter_25.html"><a href="Chapter_25.html"><i class="fa fa-check"></i><b>25</b> Multiple comparisons</a></li>
<li class="chapter" data-level="26" data-path="Chapter_26.html"><a href="Chapter_26.html"><i class="fa fa-check"></i><b>26</b> Kruskal-Wallis H test</a></li>
<li class="chapter" data-level="27" data-path="Chapter_27.html"><a href="Chapter_27.html"><i class="fa fa-check"></i><b>27</b> Two-way ANOVA</a></li>
<li class="chapter" data-level="28" data-path="Chapter_28.html"><a href="Chapter_28.html"><i class="fa fa-check"></i><b>28</b> <em>Practical</em>. ANOVA and associated tests</a>
<ul>
<li class="chapter" data-level="28.1" data-path="Chapter_28.html"><a href="Chapter_28.html#one-way-anova-site"><i class="fa fa-check"></i><b>28.1</b> One-way ANOVA (site)</a></li>
<li class="chapter" data-level="28.2" data-path="Chapter_28.html"><a href="Chapter_28.html#one-way-anova-profile"><i class="fa fa-check"></i><b>28.2</b> One-way ANOVA (profile)</a></li>
<li class="chapter" data-level="28.3" data-path="Chapter_28.html"><a href="Chapter_28.html#multiple-comparisons"><i class="fa fa-check"></i><b>28.3</b> Multiple comparisons</a></li>
<li class="chapter" data-level="28.4" data-path="Chapter_28.html"><a href="Chapter_28.html#kruskal-wallis-h-test"><i class="fa fa-check"></i><b>28.4</b> Kruskal-Wallis H test</a></li>
<li class="chapter" data-level="28.5" data-path="Chapter_28.html"><a href="Chapter_28.html#two-way-anova"><i class="fa fa-check"></i><b>28.5</b> Two-way ANOVA</a></li>
</ul></li>
<li class="chapter" data-level="29" data-path="Chapter_29.html"><a href="Chapter_29.html"><i class="fa fa-check"></i><b>29</b> Frequency and count data</a>
<ul>
<li class="chapter" data-level="29.1" data-path="Chapter_29.html"><a href="Chapter_29.html#chi-square-distribution"><i class="fa fa-check"></i><b>29.1</b> Chi-square distribution</a></li>
<li class="chapter" data-level="29.2" data-path="Chapter_29.html"><a href="Chapter_29.html#chi-square-goodness-of-fit"><i class="fa fa-check"></i><b>29.2</b> Chi-square goodness of fit</a></li>
<li class="chapter" data-level="29.3" data-path="Chapter_29.html"><a href="Chapter_29.html#chi-square-test-of-association"><i class="fa fa-check"></i><b>29.3</b> Chi-square test of association</a></li>
</ul></li>
<li class="chapter" data-level="30" data-path="Chapter_30.html"><a href="Chapter_30.html"><i class="fa fa-check"></i><b>30</b> Correlation</a>
<ul>
<li class="chapter" data-level="30.1" data-path="Chapter_30.html"><a href="Chapter_30.html#scatterplots"><i class="fa fa-check"></i><b>30.1</b> Scatterplots</a></li>
<li class="chapter" data-level="30.2" data-path="Chapter_30.html"><a href="Chapter_30.html#correlation-coefficient"><i class="fa fa-check"></i><b>30.2</b> Correlation coefficient</a>
<ul>
<li class="chapter" data-level="30.2.1" data-path="Chapter_30.html"><a href="Chapter_30.html#pearson-product-moment-correlation-coefficient"><i class="fa fa-check"></i><b>30.2.1</b> Pearson product moment correlation coefficient</a></li>
<li class="chapter" data-level="30.2.2" data-path="Chapter_30.html"><a href="Chapter_30.html#spearmans-rank-correlation-coefficient"><i class="fa fa-check"></i><b>30.2.2</b> Spearman’s rank correlation coefficient</a></li>
</ul></li>
<li class="chapter" data-level="30.3" data-path="Chapter_30.html"><a href="Chapter_30.html#correlation-hypothesis-testing"><i class="fa fa-check"></i><b>30.3</b> Correlation hypothesis testing</a></li>
</ul></li>
<li class="chapter" data-level="31" data-path="Chapter_31.html"><a href="Chapter_31.html"><i class="fa fa-check"></i><b>31</b> <em>Practical</em>. Analysis of counts and correlations</a>
<ul>
<li class="chapter" data-level="31.1" data-path="Chapter_31.html"><a href="Chapter_31.html#survival-goodness-of-fit"><i class="fa fa-check"></i><b>31.1</b> Survival goodness of fit</a></li>
<li class="chapter" data-level="31.2" data-path="Chapter_31.html"><a href="Chapter_31.html#colony-goodness-of-fit"><i class="fa fa-check"></i><b>31.2</b> Colony goodness of fit</a></li>
<li class="chapter" data-level="31.3" data-path="Chapter_31.html"><a href="Chapter_31.html#chi-square-test-of-association-1"><i class="fa fa-check"></i><b>31.3</b> Chi-Square test of association</a></li>
<li class="chapter" data-level="31.4" data-path="Chapter_31.html"><a href="Chapter_31.html#pearson-product-moment-correlation-test"><i class="fa fa-check"></i><b>31.4</b> Pearson product moment correlation test</a></li>
<li class="chapter" data-level="31.5" data-path="Chapter_31.html"><a href="Chapter_31.html#spearmans-rank-correlation-test"><i class="fa fa-check"></i><b>31.5</b> Spearman’s rank correlation test</a></li>
<li class="chapter" data-level="31.6" data-path="Chapter_31.html"><a href="Chapter_31.html#untidy-goodness-of-fit"><i class="fa fa-check"></i><b>31.6</b> Untidy goodness of fit</a></li>
</ul></li>
<li class="chapter" data-level="32" data-path="Chapter_32.html"><a href="Chapter_32.html"><i class="fa fa-check"></i><b>32</b> Simple linear regression</a>
<ul>
<li class="chapter" data-level="32.1" data-path="Chapter_32.html"><a href="Chapter_32.html#visual-interpretation-of-regression"><i class="fa fa-check"></i><b>32.1</b> Visual interpretation of regression</a></li>
<li class="chapter" data-level="32.2" data-path="Chapter_32.html"><a href="Chapter_32.html#intercepts-slopes-and-residuals"><i class="fa fa-check"></i><b>32.2</b> Intercepts, slopes, and residuals</a></li>
<li class="chapter" data-level="32.3" data-path="Chapter_32.html"><a href="Chapter_32.html#regression-coefficients"><i class="fa fa-check"></i><b>32.3</b> Regression coefficients</a></li>
<li class="chapter" data-level="32.4" data-path="Chapter_32.html"><a href="Chapter_32.html#regression-line-calculation"><i class="fa fa-check"></i><b>32.4</b> Regression line calculation</a></li>
<li class="chapter" data-level="32.5" data-path="Chapter_32.html"><a href="Chapter_32.html#coefficient-of-determination"><i class="fa fa-check"></i><b>32.5</b> Coefficient of determination</a></li>
<li class="chapter" data-level="32.6" data-path="Chapter_32.html"><a href="Chapter_32.html#regression-assumptions"><i class="fa fa-check"></i><b>32.6</b> Regression assumptions</a></li>
<li class="chapter" data-level="32.7" data-path="Chapter_32.html"><a href="Chapter_32.html#regression-hypothesis-testing"><i class="fa fa-check"></i><b>32.7</b> Regression hypothesis testing</a>
<ul>
<li class="chapter" data-level="32.7.1" data-path="Chapter_32.html"><a href="Chapter_32.html#overall-model-significance"><i class="fa fa-check"></i><b>32.7.1</b> Overall model significance</a></li>
<li class="chapter" data-level="32.7.2" data-path="Chapter_32.html"><a href="Chapter_32.html#significance-of-the-intercept"><i class="fa fa-check"></i><b>32.7.2</b> Significance of the intercept</a></li>
<li class="chapter" data-level="32.7.3" data-path="Chapter_32.html"><a href="Chapter_32.html#significance-of-the-slope"><i class="fa fa-check"></i><b>32.7.3</b> Significance of the slope</a></li>
<li class="chapter" data-level="32.7.4" data-path="Chapter_32.html"><a href="Chapter_32.html#simple-regression-output"><i class="fa fa-check"></i><b>32.7.4</b> Simple regression output</a></li>
</ul></li>
<li class="chapter" data-level="32.8" data-path="Chapter_32.html"><a href="Chapter_32.html#prediction-with-linear-models"><i class="fa fa-check"></i><b>32.8</b> Prediction with linear models</a></li>
<li class="chapter" data-level="32.9" data-path="Chapter_32.html"><a href="Chapter_32.html#conclusion"><i class="fa fa-check"></i><b>32.9</b> Conclusion</a></li>
</ul></li>
<li class="chapter" data-level="33" data-path="Chapter_33.html"><a href="Chapter_33.html"><i class="fa fa-check"></i><b>33</b> Multiple regression</a>
<ul>
<li class="chapter" data-level="33.1" data-path="Chapter_33.html"><a href="Chapter_33.html#adjusted-coefficient-of-determination"><i class="fa fa-check"></i><b>33.1</b> Adjusted coefficient of determination</a></li>
</ul></li>
<li class="chapter" data-level="34" data-path="Chapter_34.html"><a href="Chapter_34.html"><i class="fa fa-check"></i><b>34</b> <em>Practical</em>. Using regression</a>
<ul>
<li class="chapter" data-level="34.1" data-path="Chapter_34.html"><a href="Chapter_34.html#predicting-pyrogenic-carbon-from-soil-depth"><i class="fa fa-check"></i><b>34.1</b> Predicting pyrogenic carbon from soil depth</a></li>
<li class="chapter" data-level="34.2" data-path="Chapter_34.html"><a href="Chapter_34.html#predicting-pyrogenic-carbon-from-fire-frequency"><i class="fa fa-check"></i><b>34.2</b> Predicting pyrogenic carbon from fire frequency</a></li>
<li class="chapter" data-level="34.3" data-path="Chapter_34.html"><a href="Chapter_34.html#multiple-regression-depth-and-fire-frequency"><i class="fa fa-check"></i><b>34.3</b> Multiple regression depth and fire frequency</a></li>
<li class="chapter" data-level="34.4" data-path="Chapter_34.html"><a href="Chapter_34.html#large-multiple-regression"><i class="fa fa-check"></i><b>34.4</b> Large multiple regression</a></li>
<li class="chapter" data-level="34.5" data-path="Chapter_34.html"><a href="Chapter_34.html#predicting-temperature-from-fire-frequency"><i class="fa fa-check"></i><b>34.5</b> Predicting temperature from fire frequency</a></li>
</ul></li>
<li class="chapter" data-level="35" data-path="Chapter_35.html"><a href="Chapter_35.html"><i class="fa fa-check"></i><b>35</b> Randomisation</a>
<ul>
<li class="chapter" data-level="35.1" data-path="Chapter_35.html"><a href="Chapter_35.html#summary-of-parametric-hypothesis-testing"><i class="fa fa-check"></i><b>35.1</b> Summary of parametric hypothesis testing</a></li>
<li class="chapter" data-level="35.2" data-path="Chapter_35.html"><a href="Chapter_35.html#randomisation-approach"><i class="fa fa-check"></i><b>35.2</b> Randomisation approach</a></li>
<li class="chapter" data-level="35.3" data-path="Chapter_35.html"><a href="Chapter_35.html#randomisation-for-hypothesis-testing"><i class="fa fa-check"></i><b>35.3</b> Randomisation for hypothesis testing</a></li>
<li class="chapter" data-level="35.4" data-path="Chapter_35.html"><a href="Chapter_35.html#randomisation-assumptions"><i class="fa fa-check"></i><b>35.4</b> Randomisation assumptions</a></li>
<li class="chapter" data-level="35.5" data-path="Chapter_35.html"><a href="Chapter_35.html#bootstrapping"><i class="fa fa-check"></i><b>35.5</b> Bootstrapping</a></li>
<li class="chapter" data-level="35.6" data-path="Chapter_35.html"><a href="Chapter_35.html#randomisation-conclusions"><i class="fa fa-check"></i><b>35.6</b> Randomisation conclusions</a></li>
</ul></li>
<li class="appendix"><span><b>Appendix</b></span></li>
<li class="chapter" data-level="A" data-path="appendexA.html"><a href="appendexA.html"><i class="fa fa-check"></i><b>A</b> Answers to chapter exercises</a>
<ul>
<li class="chapter" data-level="A.1" data-path="appendexA.html"><a href="appendexA.html#chapter-3"><i class="fa fa-check"></i><b>A.1</b> Chapter 3</a>
<ul>
<li class="chapter" data-level="A.1.1" data-path="appendexA.html"><a href="appendexA.html#exercise-3.1"><i class="fa fa-check"></i><b>A.1.1</b> Exercise 3.1:</a></li>
<li class="chapter" data-level="A.1.2" data-path="appendexA.html"><a href="appendexA.html#exercise-3.2"><i class="fa fa-check"></i><b>A.1.2</b> Exercise 3.2</a></li>
<li class="chapter" data-level="A.1.3" data-path="appendexA.html"><a href="appendexA.html#exercise-3.3"><i class="fa fa-check"></i><b>A.1.3</b> Exercise 3.3</a></li>
<li class="chapter" data-level="A.1.4" data-path="appendexA.html"><a href="appendexA.html#exercise-3.4"><i class="fa fa-check"></i><b>A.1.4</b> Exercise 3.4</a></li>
</ul></li>
<li class="chapter" data-level="A.2" data-path="appendexA.html"><a href="appendexA.html#chapter-8"><i class="fa fa-check"></i><b>A.2</b> Chapter 8</a>
<ul>
<li class="chapter" data-level="A.2.1" data-path="appendexA.html"><a href="appendexA.html#exercise-8.1"><i class="fa fa-check"></i><b>A.2.1</b> Exercise 8.1</a></li>
<li class="chapter" data-level="A.2.2" data-path="appendexA.html"><a href="appendexA.html#exercise-8.2"><i class="fa fa-check"></i><b>A.2.2</b> Exercise 8.2</a></li>
<li class="chapter" data-level="A.2.3" data-path="appendexA.html"><a href="appendexA.html#exercise-8.3"><i class="fa fa-check"></i><b>A.2.3</b> Exercise 8.3</a></li>
</ul></li>
<li class="chapter" data-level="A.3" data-path="appendexA.html"><a href="appendexA.html#chapter-14"><i class="fa fa-check"></i><b>A.3</b> Chapter 14</a>
<ul>
<li class="chapter" data-level="A.3.1" data-path="appendexA.html"><a href="appendexA.html#exercise-14.1"><i class="fa fa-check"></i><b>A.3.1</b> Exercise 14.1</a></li>
<li class="chapter" data-level="A.3.2" data-path="appendexA.html"><a href="appendexA.html#exercise-14.2"><i class="fa fa-check"></i><b>A.3.2</b> Exercise 14.2</a></li>
<li class="chapter" data-level="A.3.3" data-path="appendexA.html"><a href="appendexA.html#exercise-14.3"><i class="fa fa-check"></i><b>A.3.3</b> Exercise 14.3</a></li>
<li class="chapter" data-level="A.3.4" data-path="appendexA.html"><a href="appendexA.html#exercise-14.4"><i class="fa fa-check"></i><b>A.3.4</b> Exercise 14.4</a></li>
<li class="chapter" data-level="A.3.5" data-path="appendexA.html"><a href="appendexA.html#exercise-14.5"><i class="fa fa-check"></i><b>A.3.5</b> Exercise 14.5</a></li>
</ul></li>
<li class="chapter" data-level="A.4" data-path="appendexA.html"><a href="appendexA.html#chapter-17"><i class="fa fa-check"></i><b>A.4</b> Chapter 17</a>
<ul>
<li class="chapter" data-level="A.4.1" data-path="appendexA.html"><a href="appendexA.html#exercise-17.1"><i class="fa fa-check"></i><b>A.4.1</b> Exercise 17.1</a></li>
<li class="chapter" data-level="A.4.2" data-path="appendexA.html"><a href="appendexA.html#exercise-17.2"><i class="fa fa-check"></i><b>A.4.2</b> Exercise 17.2</a></li>
<li class="chapter" data-level="A.4.3" data-path="appendexA.html"><a href="appendexA.html#exercise-17.3"><i class="fa fa-check"></i><b>A.4.3</b> Exercise 17.3</a></li>
</ul></li>
<li class="chapter" data-level="A.5" data-path="appendexA.html"><a href="appendexA.html#chapter-20"><i class="fa fa-check"></i><b>A.5</b> Chapter 20</a>
<ul>
<li class="chapter" data-level="A.5.1" data-path="appendexA.html"><a href="appendexA.html#exercise-20.1"><i class="fa fa-check"></i><b>A.5.1</b> Exercise 20.1</a></li>
<li class="chapter" data-level="A.5.2" data-path="appendexA.html"><a href="appendexA.html#exercise-20.2"><i class="fa fa-check"></i><b>A.5.2</b> Exercise 20.2</a></li>
<li class="chapter" data-level="A.5.3" data-path="appendexA.html"><a href="appendexA.html#exercise-20.3"><i class="fa fa-check"></i><b>A.5.3</b> Exercise 20.3</a></li>
<li class="chapter" data-level="A.5.4" data-path="appendexA.html"><a href="appendexA.html#exercise-20.4"><i class="fa fa-check"></i><b>A.5.4</b> Exercise 20.4</a></li>
<li class="chapter" data-level="A.5.5" data-path="appendexA.html"><a href="appendexA.html#exercise-20.5"><i class="fa fa-check"></i><b>A.5.5</b> Exercise 20.5</a></li>
</ul></li>
<li class="chapter" data-level="A.6" data-path="appendexA.html"><a href="appendexA.html#chapter-23"><i class="fa fa-check"></i><b>A.6</b> Chapter 23</a>
<ul>
<li class="chapter" data-level="A.6.1" data-path="appendexA.html"><a href="appendexA.html#exercise-23.1"><i class="fa fa-check"></i><b>A.6.1</b> Exercise 23.1</a></li>
<li class="chapter" data-level="A.6.2" data-path="appendexA.html"><a href="appendexA.html#exercise-23.2"><i class="fa fa-check"></i><b>A.6.2</b> Exercise 23.2</a></li>
<li class="chapter" data-level="A.6.3" data-path="appendexA.html"><a href="appendexA.html#exercise-23.3"><i class="fa fa-check"></i><b>A.6.3</b> Exercise 23.3</a></li>
<li class="chapter" data-level="A.6.4" data-path="appendexA.html"><a href="appendexA.html#exercise-23.4"><i class="fa fa-check"></i><b>A.6.4</b> Exercise 23.4</a></li>
<li class="chapter" data-level="A.6.5" data-path="appendexA.html"><a href="appendexA.html#exercise-23.5"><i class="fa fa-check"></i><b>A.6.5</b> Exercise 23.5</a></li>
</ul></li>
<li class="chapter" data-level="A.7" data-path="appendexA.html"><a href="appendexA.html#chapter-28"><i class="fa fa-check"></i><b>A.7</b> Chapter 28</a>
<ul>
<li class="chapter" data-level="A.7.1" data-path="appendexA.html"><a href="appendexA.html#exercise-28.1"><i class="fa fa-check"></i><b>A.7.1</b> Exercise 28.1</a></li>
<li class="chapter" data-level="A.7.2" data-path="appendexA.html"><a href="appendexA.html#exercise-28.2"><i class="fa fa-check"></i><b>A.7.2</b> Exercise 28.2</a></li>
<li class="chapter" data-level="A.7.3" data-path="appendexA.html"><a href="appendexA.html#exercise-28.3"><i class="fa fa-check"></i><b>A.7.3</b> Exercise 28.3</a></li>
<li class="chapter" data-level="A.7.4" data-path="appendexA.html"><a href="appendexA.html#exercise-28.4"><i class="fa fa-check"></i><b>A.7.4</b> Exercise 28.4</a></li>
</ul></li>
<li class="chapter" data-level="A.8" data-path="appendexA.html"><a href="appendexA.html#chapter-31"><i class="fa fa-check"></i><b>A.8</b> Chapter 31</a>
<ul>
<li class="chapter" data-level="A.8.1" data-path="appendexA.html"><a href="appendexA.html#exercise-31.1"><i class="fa fa-check"></i><b>A.8.1</b> Exercise 31.1</a></li>
<li class="chapter" data-level="A.8.2" data-path="appendexA.html"><a href="appendexA.html#exercise-31.2"><i class="fa fa-check"></i><b>A.8.2</b> Exercise 31.2</a></li>
<li class="chapter" data-level="A.8.3" data-path="appendexA.html"><a href="appendexA.html#exercise-31.3"><i class="fa fa-check"></i><b>A.8.3</b> Exercise 31.3</a></li>
<li class="chapter" data-level="A.8.4" data-path="appendexA.html"><a href="appendexA.html#exercise-31.4"><i class="fa fa-check"></i><b>A.8.4</b> Exercise 31.4</a></li>
<li class="chapter" data-level="A.8.5" data-path="appendexA.html"><a href="appendexA.html#exercise-31.5"><i class="fa fa-check"></i><b>A.8.5</b> Exercise 31.5</a></li>
</ul></li>
<li class="chapter" data-level="A.9" data-path="appendexA.html"><a href="appendexA.html#chapter-34"><i class="fa fa-check"></i><b>A.9</b> Chapter 34</a>
<ul>
<li class="chapter" data-level="A.9.1" data-path="appendexA.html"><a href="appendexA.html#exercise-34.1"><i class="fa fa-check"></i><b>A.9.1</b> Exercise 34.1</a></li>
<li class="chapter" data-level="A.9.2" data-path="appendexA.html"><a href="appendexA.html#exercise-34.2"><i class="fa fa-check"></i><b>A.9.2</b> Exercise 34.2</a></li>
<li class="chapter" data-level="A.9.3" data-path="appendexA.html"><a href="appendexA.html#exercise-34.3"><i class="fa fa-check"></i><b>A.9.3</b> Exercise 34.3</a></li>
<li class="chapter" data-level="A.9.4" data-path="appendexA.html"><a href="appendexA.html#exercise-34.4"><i class="fa fa-check"></i><b>A.9.4</b> Exercise 34.4</a></li>
<li class="chapter" data-level="A.9.5" data-path="appendexA.html"><a href="appendexA.html#exercise-33.5"><i class="fa fa-check"></i><b>A.9.5</b> Exercise 33.5</a></li>
</ul></li>
</ul></li>
<li class="chapter" data-level="B" data-path="uncertainty_derivation.html"><a href="uncertainty_derivation.html"><i class="fa fa-check"></i><b>B</b> Uncertainty derivation</a>
<ul>
<li class="chapter" data-level="B.1" data-path="uncertainty_derivation.html"><a href="uncertainty_derivation.html#propagation-of-error-for-addition-and-subtraction"><i class="fa fa-check"></i><b>B.1</b> Propagation of error for addition and subtraction</a></li>
<li class="chapter" data-level="B.2" data-path="uncertainty_derivation.html"><a href="uncertainty_derivation.html#propagation-of-error-for-multiplication-and-division"><i class="fa fa-check"></i><b>B.2</b> Propagation of error for multiplication and division</a></li>
</ul></li>
<li class="chapter" data-level="" data-path="references.html"><a href="references.html"><i class="fa fa-check"></i>References</a></li>
<li class="divider"></li>
<li><a href="https://github.com/rstudio/bookdown" target="blank">Published with bookdown</a></li>
</ul>
</nav>
</div>
<div class="book-body">
<div class="body-inner">
<div class="book-header" role="navigation">
<h1>
<i class="fa fa-circle-o-notch fa-spin"></i><a href="./">Fundamental statistical concepts and techniques in the biological and environmental sciences: With jamovi</a>
</h1>
</div>
<div class="page-wrapper" tabindex="-1" role="main">
<div class="page-inner">
<section class="normal" id="section-">
<div id="Chapter_35" class="section level1 hasAnchor" number="35">
<h1><span class="header-section-number">Chapter 35</span> Randomisation<a href="Chapter_35.html#Chapter_35" class="anchor-section" aria-label="Anchor link to header"></a></h1>
<p>Since introducing statistical hypothesis testing in <a href="Chapter_21.html#Chapter_21">Chapter 21</a>, this book has steadily introduced statistical tests that can be run for different data types.
In all of these statistical tests, the general idea is the same.
We calculate some test statistic, then compare this test statistic to a predetermined null distribution to find the probability (i.e., p-value) of getting a test statistic as or more extreme than the one we calculated if the null hypothesis were true.
This chapter introduces a different approach.
Instead of using a pre-determined null distribution, we will build the null distribution using our data.
This approach is not one that is often introduced in introductory statistics texts.
It is included here for two reasons.
First, randomisation presents a different way of thinking statistically without introducing an entirely different philosophical or methodological approach such as likelihood <span class="citation">(<a href="#ref-Edwards1972" role="doc-biblioref">Edwards, 1972</a>)</span> or Bayesian statistics <span class="citation">(<a href="#ref-Lee1997" role="doc-biblioref">Lee, 1997</a>)</span>.
Second, it helps reinforce the concept of what null hypothesis testing is and what p-values are.
Before explaining the randomisation approach, it is useful to summarise the parametric hypothesis tests introduced in earlier chapters.</p>
<div id="summary-of-parametric-hypothesis-testing" class="section level2 hasAnchor" number="35.1">
<h2><span class="header-section-number">35.1</span> Summary of parametric hypothesis testing<a href="Chapter_35.html#summary-of-parametric-hypothesis-testing" class="anchor-section" aria-label="Anchor link to header"></a></h2>
<p>For the parametric tests introduced in previous chapters, null distributions included the t-distribution, F-distribution, and <span class="math inline">\(\chi^{2}\)</span> distribution.
For the t-tests in <a href="Chapter_22.html#Chapter_22">Chapter 22</a>, the test statistic was the t-statistic, which we compared to a t-distribution.
The <a href="Chapter_22.html#one-sample-t-test">one-sample t-test</a> compared the mean of some variable (<span class="math inline">\(\bar{x}\)</span>) to a specific number (<span class="math inline">\(\mu_{0}\)</span>), the <a href="Chapter_22.html#independent-samples-t-test">independent samples t-test</a> compared two group means, and the <a href="#paired-sample-t-test">paired samples t-test</a> compared the mean difference between two paired groups.
All of these tests used the t-distribution and calculated some value of t.
Very low or high values of t at the extreme ends of the t-distribution are unlikely if the null hypothesis is true, so, in a two-tailed test, these are associated with low p-values that lead us to reject the null hypothesis.</p>
<p>For the analysis of variance (ANOVA) tests of <a href="Chapter_24.html#Chapter_24">Chapter 24</a>, the relevant test statistic was the <span class="math inline">\(F\)</span> statistic, with the F-distribution being the null distribution expected if two variances are equal.
The <a href="Chapter_24.html#one-way-anova">one-way ANOVA</a> used the within and among group variances of two or more groups to test the null hypothesis that all group means are equal.
The two-way ANOVA of <a href="Chapter_27.html#Chapter_27">Chapter 27</a> extended the framework of the one-way ANOVA, allowing for a second variable of groups.
This made it possible to simultaneously test whether or not the means of two different group types were the same, and whether or not there was an interaction between group types.
All of these ANOVA tests calculated some value of <span class="math inline">\(F\)</span> and compared it to the F distribution with appropriate degrees of freedom.
Sufficiently high <span class="math inline">\(F\)</span> values were associated with a low p-value and therefore the rejection of the null hypothesis.</p>
<p>The Chi-square tests introduced in <a href="Chapter_29.html#Chapter_29">Chapter 29</a> were used to test the frequencies of categorical observations and determine if they matched some expected frequencies (<a href="#chi-squared-goodness-of-fit">Chi-square goodness of fit</a> test) or were associated in some way with the frequencies of another variable (<a href="#chi-squared-test-of-association">Chi-square test of association</a>).
In these tests, the <span class="math inline">\(\chi^{2}\)</span> statistic was used and compared to a null <span class="math inline">\(\chi^{2}\)</span> distribution with an appropriate degrees of freedom.
High <span class="math inline">\(\chi^{2}\)</span> values were associated with low p-values and the rejection of the null hypothesis.</p>
<p>For testing the significance of correlation coefficients (see <a href="Chapter_30.html#Chapter_30">Chapter 30</a>) and linear regression coefficients (see <a href="Chapter_32.html#Chapter_32">Chapter 32</a>), a t-distribution was used.
And an F distribution was used to test for the overall significance of linear regression models.</p>
<p>For these tests, the approach to hypothesis testing was therefore always to use the t-distribution, F distribution, or <span class="math inline">\(\chi^{2}\)</span> distribution in some way.
These distributions are more formally defined in mathematical statistics <span class="citation">(<a href="#ref-Miller2004" role="doc-biblioref">Miller & Miller, 2004</a>)</span>, a field of study that uses mathematics to derive the probability distributions that arise from an outcome of random events (e.g., the coin-flipping of <a href="#an-instructive-example">Section 15.1</a>).
The reason that we use these distributions in statistical hypothesis testing is that they are often quite good at describing the outcomes that we expect when we collect a sample from a population.
But this is not always the case.
Recall that sometimes the assumptions of a particular statistical test were not met.
In this case, a non-parametric alternative was introduced.
The non-parametric test used the ranks of data instead of the actual values (e.g., <a href="Chapter_22.html#wilcoxon-test">Wilcoxon</a>, <a href="Chapter_22.html#mann-whitney-u-test">Mann-Whitney U</a>, <a href="Chapter_26.html#Chapter_26">Kruskal-Wallis H</a>, and <a href="#spearman-rank-correlation-coefficient">Spearman’s rank correlation coefficient</a> tests).
Randomisation uses a different approach.</p>
</div>
<div id="randomisation-approach" class="section level2 hasAnchor" number="35.2">
<h2><span class="header-section-number">35.2</span> Randomisation approach<a href="Chapter_35.html#randomisation-approach" class="anchor-section" aria-label="Anchor link to header"></a></h2>
<p>Randomisation takes a different approach to null hypothesis testing.
Instead of assuming a theoretical null distribution against which we compare our test statistic, we ask, ‘if the ordering of the data we collected was actually random, then what is the probability of getting a test statistic as or more extreme than the one that we actually did’.
Rather than using a null distribution derived from mathematical statistics, we will build the null distribution by randomising our data in some useful way <span class="citation">(<a href="#ref-Manly2007" role="doc-biblioref">Manly, 2007</a>)</span>.
Conceptually, this is often easier to understand because randomisation approaches make it easier to see why the null distribution exists and what it is doing.
Unfortunately, these methods are more challenging to implement in practice because using them usually requires knowing a bit of coding.
The best way to get started is with an instructive example.</p>
</div>
<div id="randomisation-for-hypothesis-testing" class="section level2 hasAnchor" number="35.3">
<h2><span class="header-section-number">35.3</span> Randomisation for hypothesis testing<a href="Chapter_35.html#randomisation-for-hypothesis-testing" class="anchor-section" aria-label="Anchor link to header"></a></h2>
<p>As in several previous chapters, the dataset used here is inspired by the many species of wasps that lay their eggs in the flowers of the Sonoran Desert rock fig (<em>Ficus petiolaris</em>).
This tree is distributed throughout the Baja peninsula, and in parts of mainland Mexico.
Fig trees and the wasps that develop inside of them have a fascinating ecology, but for now we will just focus on the morphologies of two closely related species as an example.
The fig wasps in Figure 35.1 are two unnamed species of the genus <em>Idarnes</em>, which we can refer to simply as ‘Short-ovipositor 1’ (SO1) and ‘Short-ovipositor 2’ (SO2).</p>
<div class="figure"><span style="display:block;" id="fig:unnamed-chunk-156"></span>
<img src="img/fig_wasps.jpg" alt="A two panel figure is shown, with two black wasps under a microscope slide shown side by side, each of which has an ovipositor about twice its body length." width="100%" />
<p class="caption">
Figure 35.1: Two fig wasp species, roughly 3 mm in length, labelled ‘SO1’ and ‘SO2’. Image modified from Duthie, Abbott, and Nason (2015).
</p>
</div>
<p>The reason that these two species are called ‘SO1’ and ‘SO2’ is that there is actually another species that lays its eggs in <em>F. petiolaris</em> flowers, one with an ovipositor that is at least twice as long as those in Figure 35.1 <span class="citation">(<a href="#ref-Duthie2015b" role="doc-biblioref">Duthie et al., 2015</a>)</span>.</p>
<p>Suppose that we have some data on the lengths of the ovipositors from each species.
We might want to know whether the mean ovipositor length differs between the two species. Figure 35.2 shows histograms of ovipositor lengths collected from 32 fig wasps, 17 SO1s and 15 SO2s.</p>
<div class="figure"><span style="display:block;" id="fig:unnamed-chunk-157"></span>
<img src="bookdown-demo_files/figure-html/unnamed-chunk-157-1.png" alt="A two panel figure showing a histogram in each panel, the left with a histogram of SO1 ovipositor length and the right with another SO2 ovipositor length. Histograms are roughly normally distributed" width="672" />
<p class="caption">
Figure 35.2: Ovipositor length distributions for two unnamed species of fig wasps, SO1 (A) and SO2 (B), collected from Baja, Mexico.
</p>
</div>
<p>To test whether or not mean ovipositor length is different between these two fig wasps, our standard approach would be to use an independent samples t-test (see <a href="Chapter_22.html#independent-samples-t-test">Section 22.2</a>).
The null hypothesis would be that the two means are the same, and the alternative (two-sided) hypothesis would be that the two means are not the same.
We would need to check the assumption that the data are normally distributed, and that both samples have similar variances.
Assuming that the assumption of normality is not violated (in which case we would need to consider a Mann-Whitney U test), and that both groups had similar variances (if not, we would use Welch’s t-test), we could proceed with calculating our t-statistic for pooled sample variance,</p>
<p><span class="math display">\[t_{\bar{y}_{\mathrm{SO1}} - \bar{y}_{\mathrm{SO2}}} = \frac{\bar{y}_{\mathrm{SO1}} - \bar{y}_{\mathrm{SO2}}}{s_{p}}.\]</span></p>
<p>The <span class="math inline">\(s_{p}\)</span> is just being used as a short-hand to indicate the pooled standard deviation.
For the two species of fig wasps, <span class="math inline">\(\bar{y}_{\mathrm{SO1}} =\)</span> 3.0301176, <span class="math inline">\(\bar{y}_{\mathrm{SO2}} =\)</span> 2.8448667, and <span class="math inline">\(s_{p} =\)</span> 0.0765801.
We can therefore calculate <span class="math inline">\(t_{\bar{y}_{\mathrm{SO1}} - \bar{y}_{\mathrm{SO2}}}\)</span>,</p>
<p><span class="math display">\[t_{\bar{y}_{\mathrm{SO1}} - \bar{y}_{\mathrm{SO2}}} = \frac{3.03 - 2.845}{0.077}.\]</span></p>
<p>After we calculate our t-statistic as <span class="math inline">\(t_{\bar{y}_{\mathrm{SO1}} - \bar{y}_{\mathrm{SO2}}} =\)</span> 2.419, we would use the t-distribution to find the p-value (or, rather, get jamovi to do this for us).
Figure 35.3 shows the t-distribution for 30 degrees of freedom with an arrow pointing at the value of <span class="math inline">\(t_{\bar{y}_{\mathrm{SO1}} - \bar{y}_{\mathrm{SO2}}} =\)</span> 2.419.</p>
<div class="figure"><span style="display:block;" id="fig:unnamed-chunk-158"></span>
<img src="bookdown-demo_files/figure-html/unnamed-chunk-158-1.png" alt="A bell curve of the t-distribution is shown with an arrow pointing downward on the x-axis value around 2.419." width="672" />
<p class="caption">
Figure 35.3: A t-distribution is shown with a calculated t-statistic of 2.419 indicated with a downward arrow.
</p>
</div>
<p>If the null hypothesis is true, and <span class="math inline">\(\bar{y}_{\mathrm{SO1}}\)</span> and <span class="math inline">\(\bar{y}_{\mathrm{SO2}}\)</span> are actually sampled from a population with the same mean, then the probability of randomly sampling a value more extreme than 2.419 (i.e., greater than 2.419 or less than -2.419) is <span class="math inline">\(P =\)</span> 0.0218352.
We therefore reject the null hypothesis because <span class="math inline">\(P < 0.05\)</span>, and we conclude that the mean ovipositor lengths of each species are not the same.</p>
<p>Randomisation takes a different approach to the same problem.
Instead of using the t-distribution in Figure 35.3, we will build our own null distribution using the fig wasp ovipositor length data.
We do it by randomising group identity in the dataset.
The logic is that if there really is no difference between group means, then we should be able to randomly shuffle group identities (species) and get a difference between means that is not far off the one we actually get from the data.
In other words, what would the difference between group means be if we just mixed up all of the species (so some SO1s become SO2s, some SO2s become SO1s, some stay the same), then calculated the difference between means of the mixed up groups?
If we just do this once, then we cannot learn much.
But if we randomly shuffle the groups many, many times (say at least 9999), then we could see what the difference between group means would look like just by chance, i.e., if ovipositor length really was not different between SO1 and SO2.
We could then compare our actual difference between mean ovipositor lengths to this null distribution in which the difference between groups means really is random (it has to be, we randomised the groups ourselves!).
The idea is easiest to see using an interactive application<a href="#fn82" class="footnote-ref" id="fnref82"><sup>82</sup></a> that builds a null distribution of differences between the mean of SO1 and SO2 and compares it to the observed difference of <span class="math inline">\(\bar{y}_{\mathrm{SO1}} - \bar{y}_{\mathrm{SO2}} =\)</span> 0.185.</p>
<p>With modern computing power, we do not need to do this randomisation manually.
A desktop computer can easily reshuffle the species identities and calculate a difference between means thousands of times in less than a second.
The histogram below shows the distribution of the difference between species mean ovipositor length if we were to randomly reshuffle groups 99999 times.</p>
<div class="figure"><span style="display:block;" id="fig:unnamed-chunk-159"></span>
<img src="bookdown-demo_files/figure-html/unnamed-chunk-159-1.png" alt="A histogram with 120 narrow bars that form a bell shaped distribution around a mean of 0. An arrow points out a value of mean SO1 minus mean SO2" width="672" />
<p class="caption">
Figure 35.4: A distribution of the difference between mean species ovipositor lengths in two different species of fig wasps when species identity is randomly shuffled. The true difference between sampled mean SO1 and SO2 ovipositor lengths is pointed out with the black arrow at 0.185. Ovipositor lengths were collected from wasps in Baja, Mexico in 2010.
</p>
</div>
<p>Given this distribution of randomised mean differences between species ovipositor lengths, the observed difference of 0.185 appears to be fairly extreme.
We can quantify how extreme by figuring out the proportion of mean differences in the above histogram that are more extreme than our observed difference (i.e., greater than 0.185 or less than -0.185).
It turns out that only 2085 out of 99999 random mean differences between ovipositor lengths were as or more extreme than our observed difference of 0.185.
To express this as a probability, we can simply take the number of differences as or more extreme than our observed difference (including the observed one itself), divided by the total number of differences (again, including the observed one),</p>
<p><span class="math display">\[P = \frac{2085 + 1}{99999 + 1}.\]</span></p>
<p>When we calculate the above, we get a value of 0.02086.
Notice that the calculated value is assigned with a ‘<span class="math inline">\(P\)</span>’.
This is because the value <strong>is a p-value</strong> (technically, an unbiased estimate of a p-value). Consider how close it is to the value of <span class="math inline">\(P =\)</span> 0.0218352 that we got from the traditional t-test.
Conceptually, we are doing the same thing in both cases; we are comparing a test statistic to a null distribution to see how extreme the differences between groups really are.</p>
</div>
<div id="randomisation-assumptions" class="section level2 hasAnchor" number="35.4">
<h2><span class="header-section-number">35.4</span> Randomisation assumptions<a href="Chapter_35.html#randomisation-assumptions" class="anchor-section" aria-label="Anchor link to header"></a></h2>
<p>Recall from <a href="Chapter_22.html#assumptions-of-t-tests">Section 22.4</a> the assumptions underlying t-tests, which include (1) data are continuous, (2) sample observations are a random sample from the population, and (3) sample means are normally distributed around the true mean.
With randomisation, we do not need to assume 2 or 3.
The randomisation approach is still valid even if the data are not normally distributed.
Samples can have different variances.
The observations do not even need to be independent.
The validity of the randomisation approach even when standard assumptions do not apply can be quite useful.</p>
<p>The downside of the randomisation approach is that the statistical inferences that we make are limited to our sample, not the broader population (recall the difference between a sample and population from <a href="Chapter_4.html#Chapter_4">Chapter 4</a>).
Because the randomisation method does not assume that the data are a random sample from the population of interest (as is the case for the traditional t-test), we cannot formally make an inference about the difference between populations from which the sample was made.
This is not necessarily a problem in practice.
It is only relevant in terms of the formal assumptions of the model.
Once we run our randomisation test, it might be entirely reasonable to argue verbally that the results of our randomisation test can generalise to our population of interest.
In other words, we can argue that the difference between groups in the sample reflects a difference in the populations from which the sample came <span class="citation">(<a href="#ref-Ernst2004" role="doc-biblioref">Ernst, 2004</a>; <a href="#ref-Ludbrook1998" role="doc-biblioref">Ludbrook & Dudley, 1998</a>)</span>.</p>
</div>
<div id="bootstrapping" class="section level2 hasAnchor" number="35.5">
<h2><span class="header-section-number">35.5</span> Bootstrapping<a href="Chapter_35.html#bootstrapping" class="anchor-section" aria-label="Anchor link to header"></a></h2>
<p>We can also use randomisation to calculate confidence intervals for a variable of interest. Remember from <a href="Chapter_18.html#Chapter_18">Chapter 18</a> the traditional way to calculate lower and upper confidence intervals using a normal distribution,</p>
<p><span class="math display">\[LCI = \bar{x} - (z \times SE),\]</span></p>
<p><span class="math display">\[UCI = \bar{x} + (z \times SE).\]</span></p>
<p>Recall from <a href="Chapter_19.html#Chapter_19">Chapter 19</a> that a t-score is substituted for a z-score to account for a finite population size.</p>
<p>Suppose we want to calculate 95% confidence intervals for the ovipositor lengths of SO1 (Figure 35.2A).
There are 17 ovipositor lengths for SO1.</p>
<p>3.256, 3.133, 3.071, 2.299, 2.995, 2.929, 3.291, 2.658, 3.406, 2.976, 2.817, 3.133, 3.000, 3.027, 3.178, 3.133, 3.210</p>
<p>To get the 95% confidence interval using the method from <a href="Chapter_18.html#Chapter_18">Chapter 18</a>, we would calculate the mean <span class="math inline">\(\bar{x}_{SO1} =\)</span> 3.03, standard deviation <span class="math inline">\(s_{\mathrm{SO1}} =\)</span> 0.26, and the t-score for <span class="math inline">\(df = N - 1\)</span> (where <span class="math inline">\(N = 17\)</span>), which is <span class="math inline">\(t = 2.120\)</span>.
Keeping in mind the formula for standard error <span class="math inline">\(s/\sqrt{N}\)</span>, we can calculate,</p>
<p><span class="math display">\[LCI = 3.03 - \left(2.120 \times \frac{0.26}{\sqrt{17}}\right),\]</span></p>
<p><span class="math display">\[UCI = 3.03 + \left(2.120 \times \frac{0.26}{\sqrt{17}}\right).\]</span></p>
<p>Calculating the above gives us values of LCI = 2.896 and UCI = 3.164.</p>
<p>Again, randomisation uses a different approach to get an estimate of the same confidence intervals.
Instead of calculating the standard error and multiplying it by a z-score or t-score to encompass a particular interval of probability density, we can instead resample the data we have with replacement many times, calculating the mean each time we resample.
The general idea is that this process of resampling approximates what would happen if we were to go back and resample new data from our original population many times, thereby giving us the distribution of means from all of these hypothetical resampling events <span class="citation">(<a href="#ref-Manly2007" role="doc-biblioref">Manly, 2007</a>)</span>.
To calculate our 95% confidence intervals, we then only need to rank the calculated means and find the mean closest to the lowest 2.5% and the highest 97.5%.</p>
<p>Remember from <a href="Chapter_15.html#sampling-with-and-without-replacement">Section 15.3</a> that the phrase ‘resampling with replacement’ just means that we are going to randomly sample some values, but not remove them from the pool of possible values after they are sampled.
If we resample the numbers above with replacement, we might therefore sample some values two or more times by chance, and other values might not be sampled at all.
The numbers below resample from the above with replacement.</p>
<p>2.299, 3.027, 2.929, 3.256, 3.071, 2.658, 3.133, 3.027, 2.976, 2.817, 3.406, 3.210, 2.976, 3.133, 3.133, 3.178, 2.299</p>
<p>Notice that some values appear more than once in the data above, while other values that were present in the original dataset are no longer present after resampling with replacement. Consequently, these new resampled data have a different mean than the original.
The mean of the original 17 SO1 ovipositor length values was 3.03, and the mean of the values resampled above is 2.972.
We can resample another set of numbers to get a new mean.</p>
<p>3.133, 3.027, 2.658, 3.291, 3.133, 2.299, 2.976, 2.658, 2.929, 3.133, 2.929, 2.995, 3.071, 2.299, 3.000, 3.133, 3.133</p>
<p>The mean of the above sample is 2.929.
We can continue doing this until we have a high number of random samples and means. Figure 35.5 shows the distribution of means if we repeat this process of resampling with replacement and calculating the mean 10000 times.</p>
<div class="figure"><span style="display:block;" id="fig:unnamed-chunk-162"></span>
<img src="bookdown-demo_files/figure-html/unnamed-chunk-162-1.png" alt="A histogram is shown with a mostly bell shape but skewing slightly to the left; x-axis values range from about 2.8 to 3.2." width="672" />
<p class="caption">
Figure 35.5: A distribution of bootstrapped values of ovipositor lengths from 17 measurements of a fig wasp species collected from Baja, Mexico. A total of 10000 bootstrapped means are shown, and arrows indicate the location of the 2.5% (LCI) and 97.5% (UCI) ranked boostrapped mean values.
</p>
</div>
<p>The arrows show the locations of the 2.5% and 97.5% ranked values of the bootstrapped means.
Note that it does not matter if the above distribution is not normal (it appears a bit skewed).
The bootstrap still works.
The values using the randomisation approach are LCI = 2.897 and UCI = 3.14.
These values are quite similar to those calculated with the traditional approach because we are doing the same thing, conceptually.
But instead of finding confidence intervals of the sample means around the true mean using the t-distribution, we are actually simulating the process of resampling from the population and calculating sample means many times.</p>
</div>
<div id="randomisation-conclusions" class="section level2 hasAnchor" number="35.6">
<h2><span class="header-section-number">35.6</span> Randomisation conclusions<a href="Chapter_35.html#randomisation-conclusions" class="anchor-section" aria-label="Anchor link to header"></a></h2>
<p>The examples demonstrated in this chapter are just a small set of what is possible using randomisation approaches.
Randomisation and bootstrapping are highly flexible tools that can be used in a variety of ways <span class="citation">(<a href="#ref-Manly2007" role="doc-biblioref">Manly, 2007</a>)</span>.
They could be used as substitutes (or complements) to any of the hypothesis tests that we have used in this book.
For example, to use a randomisation approach for testing whether or not a correlation coefficient is significantly different from 0, we could randomly shuffle the values of one variable 9999 times.
After each shuffle, we would calculate the Pearson product moment correlation coefficient, then use the 9999 randomised correlations as a null distribution to compare against the observed correlation coefficient.
This would give us a plot like the one shown in Figure 35.4, but showing the null distribution of the correlation coefficient instead of the difference between group means.
Similar approaches could be used for ANOVA or linear regression <span class="citation">(<a href="#ref-Manly2007" role="doc-biblioref">Manly, 2007</a>)</span>.</p>
</div>
</div>
<h3>References<a href="references.html#references" class="anchor-section" aria-label="Anchor link to header"></a></h3>
<div id="refs" class="references csl-bib-body hanging-indent" line-spacing="2">
<div id="ref-Duthie2015b" class="csl-entry">
Duthie, A. B., Abbott, K. C., & Nason, J. D. (2015). <span class="nocase">Trade-offs and coexistence in fluctuating environments: evidence for a key dispersal-fecundity trade-off in five nonpollinating fig wasps</span>. <em>American Naturalist</em>, <em>186</em>(1), 151–158. <a href="https://doi.org/10.1086/681621">https://doi.org/10.1086/681621</a>
</div>
<div id="ref-Edwards1972" class="csl-entry">
Edwards, A. W. F. (1972). <em><span class="nocase">Likelihood: An account of the statistical concept of likelihood and its application to scientific inference</span></em> (p. 235). Cambridge University Press, Cambridge, UK.
</div>
<div id="ref-Ernst2004" class="csl-entry">
Ernst, M. D. (2004). <span class="nocase">Permutation methods: A basis for exact inference</span>. <em>Statistical Science</em>, <em>19</em>(4), 676–685. <a href="https://doi.org/10.1214/088342304000000396">https://doi.org/10.1214/088342304000000396</a>
</div>
<div id="ref-Lee1997" class="csl-entry">
Lee, P. (1997). <em>Bayesian statistics: An introduction, 344 pp</em>. Edward Arnold, London.
</div>
<div id="ref-Ludbrook1998" class="csl-entry">
Ludbrook, J., & Dudley, H. (1998). Why permutation tests are superior to t and <span>F</span> tests in biomedical research. <em>American Statistician</em>, <em>52</em>(2), 127–132.
</div>
<div id="ref-Manly2007" class="csl-entry">
Manly, B. F. J. (2007). <em><span class="nocase">Randomization, Bootstrap and Monte Carlo Methods in Biology</span></em> (3rd ed.). Chapman & Hall/CRC, Boca Raton, USA.
</div>
<div id="ref-Miller2004" class="csl-entry">
Miller, I., & Miller, M. (2004). <em><span class="nocase">John E. Freund’s mathematical statistics</span></em> (7th ed., p. 614). Pearson Prentice Hall, Upper Saddle River, New Jersey, USA.
</div>
</div>
<div class="footnotes">
<hr />
<ol start="82">
<li id="fn82"><p><a href="https://bradduthie.github.io/stats/app/randomisation/">https://bradduthie.github.io/stats/app/randomisation/</a><a href="Chapter_35.html#fnref82" class="footnote-back">↩︎</a></p></li>
</ol>
</div>
</section>
</div>
</div>
</div>
<a href="Chapter_34.html" class="navigation navigation-prev " aria-label="Previous page"><i class="fa fa-angle-left"></i></a>
<a href="appendexA.html" class="navigation navigation-next " aria-label="Next page"><i class="fa fa-angle-right"></i></a>
</div>
</div>
<script src="libs/gitbook-2.6.7/js/app.min.js"></script>
<script src="libs/gitbook-2.6.7/js/clipboard.min.js"></script>
<script src="libs/gitbook-2.6.7/js/plugin-search.js"></script>
<script src="libs/gitbook-2.6.7/js/plugin-sharing.js"></script>
<script src="libs/gitbook-2.6.7/js/plugin-fontsettings.js"></script>
<script src="libs/gitbook-2.6.7/js/plugin-bookdown.js"></script>
<script src="libs/gitbook-2.6.7/js/jquery.highlight.js"></script>
<script src="libs/gitbook-2.6.7/js/plugin-clipboard.js"></script>
<script>
gitbook.require(["gitbook"], function(gitbook) {
gitbook.start({
"sharing": {
"github": false,
"facebook": true,
"twitter": true,
"linkedin": false,
"weibo": false,
"instapaper": false,
"vk": false,
"whatsapp": false,
"all": ["facebook", "twitter", "linkedin", "weibo", "instapaper"]
},
"fontsettings": {
"theme": "white",
"family": "sans",
"size": 2
},
"edit": {
"link": "https://github.com/rstudio/bookdown-demo/edit/master/09-Regression.Rmd",
"text": "Edit"
},
"history": {
"link": null,
"text": null
},
"view": {
"link": null,
"text": null
},
"download": null,
"search": {
"engine": "fuse",
"options": null
},
"toc": {
"collapse": "subsection"
}
});
});
</script>
<!-- dynamically load mathjax for compatibility with self-contained -->
<script>
(function () {
var script = document.createElement("script");
script.type = "text/javascript";
var src = "true";
if (src === "" || src === "true") src = "https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.9/latest.js?config=TeX-MML-AM_CHTML";
if (location.protocol !== "file:")
if (/^https?:/.test(src))
src = src.replace(/^https?:/, '');
script.src = src;
document.getElementsByTagName("head")[0].appendChild(script);
})();
</script>
</body>
</html>