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
|
context("check if inconsistencies/errors are correctly classified")
# test if the following cases are correctly identified as errors --------------
# check classification of regular errors in all types of tests
test_that("simple errors are classified as such", {
txt1 <- "t(28) = 2.20, p = .03"
txt2 <- "F(2, 28) = 2.20, p = .15"
txt3 <- "r(28) = .22, p = .26"
txt4 <- "chi2(28) = 22.20, p = .79"
txt5 <- " z = 2.20, p = .04"
txt6 <- "Q(28) = 22.20, p = .79"
expect_true(statcheck(txt1, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt2, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt3, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt4, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt5, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt6, messages = FALSE)[[VAR_ERROR]])
})
# classify p-values of negative test statistics correctly
test_that("p-values of negative tests are correctly classified", {
txt1 <- " Z = -2.42, p = 0.016" # no error
txt2 <- "t(28) = -2.20, p = .03" # error
expect_false(statcheck(txt1, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt2, messages = FALSE)[[VAR_ERROR]])
})
# classify inexactly reported p-values correctly
test_that("inexactly reported p-values are correctly classified",{
txt1 <- "t(28) = 2.20, ns"
txt2 <- "t(28) = 2.20, p > .05"
txt3 <- "t(28) = 2.0, p < .05"
expect_true(statcheck(txt1, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt2, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt3, messages = FALSE)[[VAR_ERROR]])
})
# also classify decision errors as errors
test_that("decision errors are also classified as errors",{
txt1 <- "t(28) = 1.20, p = .03"
txt2 <- "t(28) = 2.20, p = .30"
expect_true(statcheck(txt1, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt2, messages = FALSE)[[VAR_ERROR]])
})
# test if the following cases are correctly identified as correct -------------
# correct rounding
test_that("correctly rounded p-values are not considered errors", {
txt1 <- "t(28) = 2, p = .02"
txt2 <- "t(28) = 2, p = .14"
txt3 <- "t(28) = 2.2, p = .03" # rounded lower bound p-value
txt4 <- "t(28) = 2.2, p = .04"
txt5 <- "t(28) = 2.20, p = .036"
txt6 <- "t(28) = 2.20, p = .037"
expect_false(statcheck(txt1, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt2, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt3, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt4, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt5, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt6, messages = FALSE)[[VAR_ERROR]])
})
# test if different arguments concerning errors work --------------------------
# OneTailedTests: assume all tests are one-tailed
test_that("OneTailedTests considers everything as one-tailed", {
txt1 <- "t(28) = 2.20, p = .02"
txt2 <- "t(28) = 2.20, p = .04"
txt3 <- "this test is one-tailed: t(28) = 2.20, p = .02, but this one is not: t(28) = 2.20, p = .04"
expect_false(statcheck(txt1, messages = FALSE, OneTailedTests = TRUE)[[VAR_ERROR]])
expect_true(statcheck(txt2, messages = FALSE, OneTailedTests = TRUE)[[VAR_ERROR]])
expect_equal(statcheck(txt3, messages = FALSE, OneTailedTests = TRUE)[[VAR_ERROR]], c(FALSE, TRUE))
})
# OneTailedTxt: automated detection of one-tailed test in text
test_that("automated one-tailed test detection works", {
txt1 <- "t(28) = 2.20, p = .018"
txt2 <- "t(28) = 2.20, p = .01, one-tailed"
txt3 <- "t(28) = 2.20, p = .018, one-tailed"
txt4 <- "t(28) = 2.20, p = .018, one-sided"
txt5 <- "t(28) = 2.20, p = .018, directional"
# don't correct for one-tailed testing here
expect_true(statcheck(txt1, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt1, messages = FALSE, OneTailedTxt = TRUE)[[VAR_ERROR]])
expect_true(statcheck(txt2, messages = FALSE, OneTailedTxt = TRUE)[[VAR_ERROR]])
expect_true(statcheck(txt3, messages = FALSE)[[VAR_ERROR]])
# correct for one-tailed testing here
expect_false(statcheck(txt3, messages = FALSE, OneTailedTxt = TRUE)[[VAR_ERROR]])
expect_false(statcheck(txt4, messages = FALSE, OneTailedTxt = TRUE)[[VAR_ERROR]])
expect_false(statcheck(txt5, messages = FALSE, OneTailedTxt = TRUE)[[VAR_ERROR]])
# check that p-values were corrected in these cases
p_1tail <- pt(2.20, 28, lower.tail = FALSE)
expect_equal(statcheck(c(txt3, txt4, txt5), messages = FALSE,
OneTailedTxt = TRUE)[[VAR_COMPUTED_P]], rep(p_1tail, 3))
})
# pZeroError: check if p = .000 is counted as an inconsistency or not
test_that("you can adapt whether p = .000 is counted as inconsistent or not", {
txt1 <- "t(28) = 22.20, p = .000"
txt2 <- "t(28) = 22.20, p < .000" # this is always an Error
expect_true(statcheck(txt1, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt1, messages = FALSE, pZeroError = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt2, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt2, messages = FALSE, pZeroError = FALSE)[[VAR_ERROR]])
})
# test classifications of (in)exact test statistcs and (in)exact p-values ----
# test statistics exactly reported
test_that("cases where t = ... are correctly classified", {
# calculate range of correct p-values
lowp <- pt(2.25, 28, lower.tail = FALSE)*2
upp <- pt(2.15, 28, lower.tail = FALSE)*2
# correct
txt1 <- "t(28) = 2.2, p = .036" # correct
txt2 <- "t(28) = 2.2, p < .08" # correct
txt3 <- "t(28) = 2.2, p > .02" # correct
# error
txt4 <- paste("t(28) = 2.2, p >", upp) # error
txt5 <- paste("t(28) = 2.2, p <", lowp) # error
txt6 <- "t(28) = 2.2, p = .08" # error
txt7 <- "t(28) = 2.2, p = .02" # error
txt8 <- "t(28) = 2.2, p > .08" # error
txt9 <- "t(28) = 2.2, p < .02" # error
expect_false(statcheck(txt1, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt2, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt3, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt4, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt5, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt6, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt7, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt8, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt9, messages = FALSE)[[VAR_ERROR]])
})
# test statistic reported as <
test_that("cases where t < ... are correctly classified", {
# calculate range of correct p-values
lowp <- pt(2.25, 28, lower.tail = FALSE)*2
upp <- pt(2.15, 28, lower.tail = FALSE)*2
# correct
txt1 <- paste("t(28) < 2.20, p >", upp)
txt2 <- "t(28) < 2.2, p = .08"
txt3 <- "t(28) < 2.2, p > .08"
txt4 <- "t(28) < 2.2, p < .08"
txt5 <- "t(28) < 2.2, p > .02"
# error
txt6 <- paste("t(28) < 2.2, p =", lowp)
txt7 <- paste("t(28) < 2.2, p <", lowp)
txt8 <- "t(28) < 2.2, p < .02"
txt9 <- "t(28) < 2.2, p = .02"
expect_false(statcheck(txt1, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt2, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt3, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt4, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt5, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt6, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt7, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt8, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt9, messages = FALSE)[[VAR_ERROR]])
})
# test statistic reported as >
test_that("cases where t > ... are correctly classified", {
# calculate range of correct p-values
lowp <- pt(2.25, 28, lower.tail = FALSE)*2
upp <- pt(2.15, 28, lower.tail = FALSE)*2
# correct
txt1 <- paste("t(28) > 2.20, p <", upp)
txt2 <- "t(28) > 2.2, p = .02"
txt3 <- "t(28) > 2.2, p > .02"
txt4 <- "t(28) > 2.2, p < .02"
txt5 <- "t(28) > 2.2, p < .08"
# error
txt6 <- paste("t(28) > 2.2, p =", upp)
txt7 <- paste("t(28) > 2.2, p >", upp)
txt8 <- "t(28) > 2.2, p > .08"
txt9 <- "t(28) > 2.2, p = .08"
expect_false(statcheck(txt1, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt2, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt3, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt4, messages = FALSE)[[VAR_ERROR]])
expect_false(statcheck(txt5, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt6, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt7, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt8, messages = FALSE)[[VAR_ERROR]])
expect_true(statcheck(txt9, messages = FALSE)[[VAR_ERROR]])
})
|
