forked from PeterKDunn/SRM-Textbook
-
Notifications
You must be signed in to change notification settings - Fork 0
/
01-intro.Rmd
executable file
·482 lines (353 loc) · 19.2 KB
/
01-intro.Rmd
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
\mainmatter
# Research: An introduction {#Intro}
::: {.objectivesBox .objectives data-latex="{iconmonstr-target-4-240.png}"}
In this chapter, you will learn to:
* identify quantitative and qualitative research.
* identify the steps in the quantitative research process.
:::
## How do we know what we know? {#HowDoWeKnow}
People once believed that all life regularly and commonly
`r if (knitr::is_latex_output()) {
'arose spontaneously'
} else {
'[arose spontaneously](http://en.wikipedia.org/wiki/Spontaneous_generation)'
}`
from non-living matter.
*Recipes* even existed to create life; for example, van Helmont [@data:VanHelmont:Transformations; @data:Latour:PasteurPouchet] gave this
`r if (knitr::is_latex_output()) {
'recipe'
} else {
'[recipe](https://www.mnn.com/green-tech/research-innovations/stories/how-to-make-a-mouse-the-bizarre-recipes-borne-of-spontaneous)'
}`
for making a mouse:
<div style="float:right; width: 222x; border: 1px; padding:10px"><img src="Illustrations/animal-1238239_1920.jpg" width="200px"/></div>
<!-- \begin{wrapfigure}{R}{.25\textwidth} -->
<!-- \begin{center} -->
<!-- \includegraphics[width=.20\textwidth]{Illustrations/animal-1238239_1920.jpg} -->
<!-- \end{center} -->
<!-- \end{wrapfigure} -->
> If a soiled shirt is placed in the opening of a vessel containing grains of wheat, the reaction of the leaven in the shirt with fumes from the wheat will, after approximately twenty-one days, transform the wheat into mice.
This idea was called 'spontaneous generation'.
We now know this isn't true... but where did the idea come from?
Through *observation*.
Spontaneous generation was proposed to explain what had been *observed*: following the above process *did* produce mice.
However, this hypothesis ('possible explanation') was rejected when evidence was found that contradicted the hypothesis.
So, a new hypothesis was proposed and tested, based on further *evidence*.
Briefly, this is the *evidence-based, scientific process*.
More recently, the dangers of smoking were still being debated into the 1990s:
<div style="float:right; width: 222x; border: 1px; padding:10px">
Researcy </div>
<!-- \begin{wrapfigure}{R}{.25\textwidth} -->
<!-- \begin{center} -->
<!-- \includegraphics[width=.20\textwidth]{Illustrations/cigarette-1270516_640.jpg} -->
<!-- \end{center} -->
<!-- \end{wrapfigure} -->
> ... a causal role for smoking [has] not been proved beyond reasonable doubt.
>
> --- @eysenck1991were, p. 429
All scientific knowledge emerges in a similar way: Observations lead to hypotheses, which are tested against the *evidence*.
If the evidence *contradicts* the hypothesis, the hypothesis is rejected; if the evidence *supports* the hypothesis, the hypothesis is *temporarily* accepted since contradictory evidence may emerge in the future.
Hypotheses not contradicted by large amounts of evidence, over a long time, are sometimes called *laws* or *theories* (such as the 'law of conservation of energy').
Importantly, theories and laws can be disproven if contradictory evidence emerges.
Knowledge in all scientific disciplines is accumulated using a similar process.
<!--
* How do we know the gestation length for *Gilbert's Potoroo* [@stead2010monitoring]?
* How do we know that paracetamol eases pain [@weil2007paracetamol]?
* How do we know that exercise is good for us [@curfman1993exercise]?
* How do we know if
`r if (knitr::is_latex_output()) {
'permeable pavement technology'
} else {
'[permeable pavement technology](https://en.wikipedia.org/wiki/Permeable_paving)'
}`
is effective in reducing runoff [@mullaney2014practical]?
-->
## Evidence-based research {#EvidenceBasedResearch}
Every discipline changes, develops, improves, and adapts---usually through *research*.
Your discipline is not the same as it was 10 years ago; it will change in the next 10 years.
Scientists, engineers and health practitioners need to know how to understand and adapt to this change.
To remain current in your discipline, understanding research is vital, even if you will not be conducting research yourself.
You still need to know the language, tools, concepts and ideas of research: you need to know how to critique research.
research is the foundation of science.
<iframe src="https://learningapps.org/watch?v=ppa1cz56222" style="border:0px;width:100%;height:500px" allowfullscreen="true" webkitallowfullscreen="true" mozallowfullscreen="true"></iframe>
Scientific research systematically answers questions using *data*; that is, science seeks *evidence-based answers*.
'Evidence-based research' refers to research conclusions based on *evidence*, rather than hunches, feelings, intuition, hopes, or tradition.
The *evidence* comes from analysing the collected *data*.
::: {.definition #Data name="Data"}
Data refers to information (observations or measurements) obtained from a study, such as numbers, labels, recordings, videos, text, etc.
:::
::: {.definition #DataSet name="Dataset"}
A *dataset* refers to an *organised* and *structured* collection of data from a study.
:::
Research involves designing how to collect data, and analysing that data; this book covers both parts.
<!-- (https://www.usgs.gov/faqs/what-are-differences-between-data-a-dataset-and-a-database?qt-news_science_products=0#qt-news_science_products) -->
## Example: research in action {#Research-in-Action}
During 1988/1989, an unusually high number of cases of the *Legionella longbeachae* infection were observed in South Australia.
The researchers wanted to identify the source of the infection to prevent further infections.
The researchers noticed that many of those infected were gardeners who had recently handled potting mix, so they hypothesised that the infection was somehow associated with using potting mix.
They designed a study to test this hypothesis, then collected data from 100 people (25 *with* the infection, and 75 similar people *without* the infection).
<div style="float:right; width: 222x; border: 1px; padding:10px">
<img src="Illustrations/pexels-lukas-296230.jpg" width="200px"/>
</div>
The researchers described and summarised their data, then analysed the data to reach an evidence-based conclusion: potting mix was partially, but not solely, responsible for the increase in infection numbers.
The researchers communicated their recommendations to
`r if (knitr::is_latex_output()) {
'reduce the risks of people contracting the infection.'
} else {
'[reduce the risks of people contracting the infection](https://www.worksafe.qld.gov.au/safety-and-prevention/hazards/hazardous-exposures/biological-hazards/legionella-risks-from-work-with-potting-mix-and-compost).'
}`
::: {.thinkBox .think data-latex="{iconmonstr-light-bulb-2-240.png}"}
In this book, we learn about each of the *six steps of research*.
`r if( knitr::is_html_output() ) {
"Arrange these steps into the usual order:"
}`
`r if( knitr::is_latex_output() ) {
"Try arranging these steps into the usual order, using these terms:
*Design*; \\enskip
*Summarise and describe*; \\enskip
*Analyse*; \\enskip
*Ask*; \\enskip
*Collect*; \\enskip
*Report*."
}`
**Step 1**:
`r if( knitr::is_html_output() ) {mcq(
c("Design",
"Summarise and describe",
"Analyse",
answer = "Ask",
"Collect",
"Report") )
} else {
"_________"}` the question
**Step 2**:
`r if( knitr::is_html_output() ) {mcq(
c(answer = "Design",
"Summarise and describe",
"Analyse",
"Ask",
"Collect",
"Report") )
} else {
"_________"}` the study
**Step 3**:
`r if( knitr::is_html_output() ) {mcq(
c("Design",
"Summarise and describe",
"Analyse",
"Ask",
answer = "Collect",
"Report") )
} else {
"_________"}` the data
**Step 4**:
`r if( knitr::is_html_output() ) {mcq(
c("Design",
answer = "Summarise and describe",
"Analyse",
"Ask",
"Collect",
"Report") )
} else {
"_________"}` the data
**Step 5**:
`r if( knitr::is_html_output() ) {mcq(
c("Design",
"Summarise and describe",
answer = "Analyse",
"Ask",
"Collect",
"Report") )
} else {
"_________"}` the data
**Step 6**:
`r if( knitr::is_html_output() ) {mcq(
c("Design",
"Summarise and describe",
"Analyse",
"Ask",
"Collect",
answer = "Report") )
} else {
"_________"}` the results
:::
## Types of research {#TypesOfResearch}
Research can be broadly classified as *qualitative* or *quantitative* research, which are different yet complementary (Table\ \@ref(tab:TypesOfResearch)).
Both methods have advantages and disadvantages, and are often used together (called *mixed methods* research).
The decision to use qualitative, quantitative or mixed methods approaches should depend on the aims of the research, not the skills or knowledge of the researchers conducting the research.
```{r}
ResearchTypes <- array( dim = c(10, 3) )
ResearchTypes[1, ] <- c("",
"![](./Pics/iconmonstr-pen-3-240.png){#id .class height=80px width=80px}",
"![](./Pics/iconmonstr-calculator-9-240.png){#id .class height=80px width=80px}")
ResearchTypes[2, ] <- c("Aspect",
"Qualitative",
"Quantitative")
ResearchTypes[3, ] <- c("What",
"Feelings, opinions",
"Measured or observed data")
ResearchTypes[4, ] <- c("Why",
"Suggest hypotheses",
"Tests hypotheses")
ResearchTypes[5, ] <- c("Conclusions",
"Detailed",
"General")
ResearchTypes[6, ] <- c("Data",
"Words, pictures, ...",
"Numbers, measurements, ...")
ResearchTypes[7, ] <- c("Size",
"Usually small samples are studied",
"Often large samples are studied")
ResearchTypes[8, ] <- c("Time",
"Time-consuming",
"More efficient")
ResearchTypes[9, ] <- c("Applicability",
"Rarely generalisable",
"Sometimes generalisable")
ResearchTypes[10, ] <- c("Examples",
"Interviews, focus groups, diaries",
"Experiments, surveys, measurements")
ResearchTypes <- ResearchTypes[, c(2, 1, 3) ] ### Reorder
```
```{r TypesOfResearch}
if( knitr::is_latex_output() ) {
kable(ResearchTypes[3:10,],
col.names = c("Qualitative",
"Aspect",
"Quantitative"),
format = "latex",
align = c("r", "c", "l"),
linesep = c("","","","\\addlinespace"),
longtable = FALSE,
caption = "Comparing qualitative and quantitative research",
booktabs = TRUE) %>%
kable_styling(font_size = 10) %>%
column_spec(column = 2,
bold = TRUE) %>%
row_spec(row = 0,
bold = TRUE)
}
if( knitr::is_html_output() ) {
kable(ResearchTypes,
format = "html",
align = c("r", "c", "l"),
longtable = FALSE,
caption = "Comparing qualitative and quantitative research",
booktabs = TRUE) %>%
column_spec(column = 2,
bold = TRUE) %>%
row_spec(row = 2,
bold = TRUE)
}
```
Briefly, *qualitative research* leads to a deeper understanding, usually from a very narrowly-defined group.
Meanings, motivations, opinions or themes often emerge from qualitative research.
In contrast, *quantitative research* summarises and analyses data usually from large groups, using *numerical* methods, such as averages and percentages.
In quantitative research, typically information about a larger group of interest (a *population*) is found from a subset of the population (a *sample*).
<!--
> ... *quantitative* data gets you the numbers to... [support] the broad general points of your research.
> *Qualitative* data brings you the details and the depth to understand their full implications.
>
> --- [SurveyMonkey website](https://www.surveymonkey.com/mp/quantitative-vs-qualitative-research/), October 2019; emphasis added.
-->
::: {.definition #QualitativeResearch name="Quantitative research"}
*Quantitative research* summarises and analyses data using numerical methods, such as producing averages and percentages.
:::
::: {.importantBox .important data-latex="{iconmonstr-warning-8-240.png}"}
This book is about *quantitative* research.
:::
<div style="float:right; width: 222x; border: 1px; padding:10px">
<img src="Illustrations/pexels-rathaphon-nanthapreecha-3846205.jpg" width="200px"/>
</div>
::: {.example #BroadTypesOfResearch name="Types of research"}
Suppose we wish to learn about the perceived benefits and barriers for adopting electric vehicles (EVs).
A *qualitative research study* might use two small *focus groups*: one group comprising people who *have* purchased EV, and another group comprising people who *haven't*.
The researchers ask each group about the reasons for their purchase.
A *quantitative research* study might survey a large number of buyers of EVs and buyers of non-EVs, and ask the buyers' age, sex, and questions about the reasons for their purchase.
The survey responses could be analysed by numerically summarising, and comparing, the responses for buyers and non-buyers of EVs.
A *mixed methods* study may combine both of the above.
:::
## The steps in research {#SixStepsOfResearch}
The research process ideally follows the process in Fig.\ \@ref(fig:SixSteps), but this is not always possible or practical.
The process is not always linear: researchers may jump from step to step as necessary.
Nonetheless, each step is important.
```{r SixSteps, fig.cap="", fig.align="center", fig.width=3, out.width="35%", fig.cap="The six basic steps in research. The steps are not always performed linearly."}
SixSteps()
```
All steps are discussed in this book:
* **Asking** the research question: Chap.\ \@ref(RQs).
Research begins by asking a question.
* **Designing** the study: Chaps.\ \@ref(ResearchDesign) to\ \@ref(Interpretation).
In evidence-based research, the question is answered using data.
A study must be designed to obtain that data, which includes determining who or what to study, how to find them, what information to obtain, and ensuring data are obtained ethically.
* **Collecting** the data: Chap.\ \@ref(CollectingDataProcedures).
The data collection process must be clearly documented.
* **Describing** and **summarising** the data: Chaps.\ \@ref(DescribingVars) to\ \@ref(NumericalQual).
Before analysis, the data must be described and summarised.
A computer is useful for this step.
* **Analysing** the data: Chaps.\ \@ref(MakingDecisions) to\ \@ref(Regression).
Analysis refers to determining how the data answer the research question, and depends on the type of data and the research question.
A computer is useful for this step.
* **Reporting** the results: Chaps.\ \@ref(Reading) and\ \@ref(WritingResearch).
Communicating the results appropriately, accurately and ethically is important, includes reporting any limitations of the study.
## Using software in research {#Software-In-Research}
Using spreadsheets in research for storing and analysing data requires care; extremely expensive and dangerous errors have been made due to using spreadsheets [@altarawneh2017pilot].
These problems emerge for different reasons:
<div style="float:right; width: 222x; border: 1px; padding:10px">
<img src="Illustrations/students-1807505_640.jpg" width="200px"/>
</div>
* Spreadsheets may *automatically alter data* (for example, reformatting entries as dates), even when not appropriate [@ziemann2016gene].
* Spreadsheets may include *formulas with errors* [@panko1998hitting] that are *difficult to locate* and hence fix [@panko2016we; @Retraction:London:Excel].
* Spreadsheets *do not leave a record* of how the data have been analysed or prepared; for example, formulas can be difficult to understand and parse.
Keeping a record of the analysis, preparation of variables, and other operations with the data is good scientific practice (*reproducible research*; see Sect.\ \@ref(ReproducibleResearch)) [@simons2019reproducible]).
Reproducibility ensures, among other advantages, that results can be checked by the researchers and verified by others.
* Excel has *bugs* [@keeling2004numerical; @melard2014accuracy] even in basic operations [@berger2007nonstandard; @hargreaves2010polynomial].
Sometimes these errors are made worse after attempts to fix them [@mccullough2002accuracy].
Problems with using spreadsheets, as with any software package, are often due to human error, but *spreadsheets make the errors hard to find and fix*.
Spreadsheets are useful for data collection and manipulation, but are not designed for scientific analysis.
Be careful using spreadsheets for research and analysis.
*Statistical* software (such as jamovi, Python, R, SAS, SPSS, etc.) helps avoid many of these problems.
Statistical software:
* is designed for large datasets.
* encourages reproducible research (Sect. \@ref(ReproducibleResearch)).
* allows high-precision formatting and graphics.
* is powerful; with some programming skills, almost anything is possible.
* is designed for performing statistical analyses and working with data.
::: {.softwareBox .software data-latex="{iconmonstr-laptop-4-240.png}"}
In this book, output from the statistical software packages jamovi [@Software:jamovi] and SPSS [@Software:SPSS] is sometimes shown.
:::
:::{.example}
@zeeberg2004mistaken found Excel was corrupting data files used in genetic analysis (p. 1):
> The date conversions affect at least 30 gene names; the floating-point conversions affect at least 2,000 [...]
> These conversions are irreversible; the original gene names cannot be recovered.
:::
## Quick review questions {#Chap1-QuickReview}
::: {.webex-check .webex-box}
Which of the following are likely to be answered using *quantitative* or *qualitative* research studies?
1. What percentage of the population experiences minor side-effects from this medication?\tightlist
`r if( knitr::is_html_output() ) {mcq( c("Qualitative", answer="Quantitative") )}`
1. What is the average number of roof-top solar panels installed on domestic properties?
`r if( knitr::is_html_output() ) {mcq( c("Qualitative", answer="Quantitative") )}`
1. Why do people opt to purchase an electric car?
`r if( knitr::is_html_output() ) {mcq( c( answer="Qualitative", "Quantitative") )}`
:::
## Exercises {#IntroExercises}
Selected answers are available in Sect.\ \@ref(IntroAnswer).
::: {.exercise #RQsTypeTourniquet}
Consider the research question: "Which of three different junctional tourniquets are quickest, on average, to apply?"
Is this RQ likely to be answered using a *quantitative* or *qualitative* research study?
:::
::: {.exercise #RQsTypeMangroves}
Consider the research question: "Why do people dump rubbish in mangroves?"
Is this RQ likely to be answered using a *quantitative* or *qualitative* research study?
:::
<!-- QUICK REVIEW ANSWERS -->
`r if (knitr::is_html_output()) '<!--'`
::: {.EOCanswerBox .EOCanswer data-latex="{iconmonstr-check-mark-14-240.png}"}
**Answers to in-chapter questions:**
- \textbf{\textit{Quick Revision:}}
**1.** Quantitative
**2.** Quantitative
**3.** Qualitative.
:::
`r if (knitr::is_html_output()) '-->'`